\\n> Where did you find the precise escaping rules for markdown? I do not think that it should be invented...\\n> \\n> \\n> I have looked at RST, according to this reference:\\n> \\n> \\thttp://docutils.sourceforge.net/docs/ref/rst/restructuredtext.html#grid-tables\\n> \\n> The good news is that you do not need to handle a special | case because you would only produce clean tables.\\n> \\n> I've tested UTF-8 with plane 1 (你好!) and plane 2 (𠀡) and the alignment seems to worked well, incredible!\\n> \\n>>> My main interest on this was in rst. I`m using markdown only in github issues and my knowldge about md is poor.\\n> \\n> Then maybe only do RST?!\\n> \\n> It looks much simpler anyway, and if you do MARKDOWN the support needs to be clean.\\n> \\n> About the code:\\n> \\n> I'm still at odds with the code which needs to test for markdown to call for different functions in multiple places. If you keep md and in order to avoid that, I would suggest to extend the pg_wcs* functions with a list of caracters which may have different sizes with additionnal args, say:\\n> \\n> pg_wcssize(// same args, plus:\\n> char * escaped_chars, // will require escaping\\n> int escape_len, // how many chars added when escaping\\n> int nllen // len of newline if substituted\\n> );\\n> \\n> So that pg_wcssize(..., \\\"\\\\r\\\", 1, 1) would behave as before (\\\\n and \\\\t are rather special cases), and the various constants could be held in the format description so the whole thing would be parametric.\\n> \\n> Same approach with format.\\n> \\n> That would allow to simplify the code significantly and to share it between MARKDOWN and others. Also, the multiple \\\"else if\\\" list would be simplified by using strchr or the escaped_chars string.\\n\\nThis patch was moved into the current Commitfest marked “Waiting for author”\\nwith the above review. Have you had a chance to work on it addressing the\\nreview comments such that we can expect a new version within this CF?\\n\\ncheers ./daniel\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Wed, 13 Sep 2017 01:14:03 +0200\",\n \"msg_from\": \"Daniel Gustafsson <daniel@yesql.se>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"I have`t any new version recently, because i was waiting for new version.\\nBut I will have some time on this in next months. How many time I have to\\nnext freeze?\\n\\nThanks Je;\\n\\n2017-09-13 1:14 GMT+02:00 Daniel Gustafsson <daniel@yesql.se>:\\n\\n> > On 08 May 2017, at 12:02, Fabien COELHO <fabien.coelho@mines-\\n> paristech.fr> wrote:\\n> >\\n> > Hello Jan,\\n> >\\n> > Please give a number to submitted patches. I think that this was v3.\\n> >\\n> > The patch does NOT fix various issues I pointed out in my previous\\n> review:\\n> >\\n> > - tabs introduced in \\\"doc/src/sgml/ref/psql-ref.sgml\\\"\\n> > - too long help line in \\\"src/bin/psql/help.c\\\"\\n> > - spurious space after a comma in \\\"src/fe_utils/print.c\\\"\\n> > and possibly elsewhere.\\n> >\\n> > On Sun, 23 Apr 2017, Jan Michálek wrote:\\n> >\\n> >>>> Markdown include characters/sequences which are interpreted as\\n> markers:\\n> >>>> _Italic_, **Bold**, *** => horizontal rules, > block quote... `inline\\n> >>>> code`... If they are interpreted within a table cell then probably\\n> they\\n> >>>> should be escaped somehow.\\n> >>\\n> >> I have treated \\\"_*|<>\\\"\\n> >\\n> > Probably not enough, see below. Note the escaping chars should also be\\n> escaped.\\n> >\\n> >>> I`m able to sanitize characters, but complex sequences will be\\n> problem. I\\n> >>> will look on this, but I don`t know, if I`m able to do this.\\n> >\\n> > I do not know whether only those are necessary. Have you checked?\\n> Guessing is probably not the right approach.\\n> >\\n> >\\n> > Concerning MARKDOWN, and according to the following source about github\\n> markdown implementation:\\n> >\\n> > https://enterprise.github.com/downloads/en/markdown-cheatsheet.pdf\\n> >\\n> > The following characters may need to be backslash escaped, although it\\n> does not cover HTML stuff.\\n> >\\n> > \\\\ backslash\\n> > ` backtick\\n> > * asterisk\\n> > _ underscore\\n> > {} curly braces\\n> > [] square brackets\\n> > () parentheses\\n> > # hash mark\\n> > + plus sign\\n> > - minus sign (hyphen)\\n> > . dot\\n> > ! exclamation\\n> >\\n> > Another source https://genius.com/3057216 suggests (* # / ( ) [ ] < >),\\n> > which should protect HTML.\\n> >\\n> > However, the escaping seems to be the backslash character, NOT using\\n> html encoding < as done in your version.\\n> >\\n> > Where did you find the precise escaping rules for markdown? I do not\\n> think that it should be invented...\\n> >\\n> >\\n> > I have looked at RST, according to this reference:\\n> >\\n> > http://docutils.sourceforge.net/docs/ref/rst/\\n> restructuredtext.html#grid-tables\\n> >\\n> > The good news is that you do not need to handle a special | case because\\n> you would only produce clean tables.\\n> >\\n> > I've tested UTF-8 with plane 1 (你好!) and plane 2 (𠀡) and the alignment\\n> seems to worked well, incredible!\\n> >\\n> >>> My main interest on this was in rst. I`m using markdown only in github\\n> issues and my knowldge about md is poor.\\n> >\\n> > Then maybe only do RST?!\\n> >\\n> > It looks much simpler anyway, and if you do MARKDOWN the support needs\\n> to be clean.\\n> >\\n> > About the code:\\n> >\\n> > I'm still at odds with the code which needs to test for markdown to call\\n> for different functions in multiple places. If you keep md and in order to\\n> avoid that, I would suggest to extend the pg_wcs* functions with a list of\\n> caracters which may have different sizes with additionnal args, say:\\n> >\\n> > pg_wcssize(// same args, plus:\\n> > char * escaped_chars, // will require escaping\\n> > int escape_len, // how many chars added when escaping\\n> > int nllen // len of newline if substituted\\n> > );\\n> >\\n> > So that pg_wcssize(..., \\\"\\\\r\\\", 1, 1) would behave as before (\\\\n and \\\\t\\n> are rather special cases), and the various constants could be held in the\\n> format description so the whole thing would be parametric.\\n> >\\n> > Same approach with format.\\n> >\\n> > That would allow to simplify the code significantly and to share it\\n> between MARKDOWN and others. Also, the multiple \\\"else if\\\" list would be\\n> simplified by using strchr or the escaped_chars string.\\n>\\n> This patch was moved into the current Commitfest marked “Waiting for\\n> author”\\n> with the above review. Have you had a chance to work on it addressing the\\n> review comments such that we can expect a new version within this CF?\\n>\\n> cheers ./daniel\\n\\n\\n\\n\\n-- \\nJelen\\nStarší čeledín datovýho chlíva\\n\\nI have`t any new version recently, because i was waiting for new version. But I will have some time on this in next months. How many time I have to next freeze?Thanks Je;2017-09-13 1:14 GMT+02:00 Daniel Gustafsson <daniel@yesql.se>:> On 08 May 2017, at 12:02, Fabien COELHO <fabien.coelho@mines-paristech.fr> wrote:\\n>\\n> Hello Jan,\\n>\\n> Please give a number to submitted patches. I think that this was v3.\\n>\\n> The patch does NOT fix various issues I pointed out in my previous review:\\n>\\n> - tabs introduced in \\\"doc/src/sgml/ref/psql-ref.sgml\\\"\\n> - too long help line in \\\"src/bin/psql/help.c\\\"\\n> - spurious space after a comma in \\\"src/fe_utils/print.c\\\"\\n> and possibly elsewhere.\\n>\\n> On Sun, 23 Apr 2017, Jan Michálek wrote:\\n>\\n>>>> Markdown include characters/sequences which are interpreted as markers:\\n>>>> _Italic_, **Bold**, *** => horizontal rules, > block quote... `inline\\n>>>> code`... If they are interpreted within a table cell then probably they\\n>>>> should be escaped somehow.\\n>>\\n>> I have treated \\\"_*|<>\\\"\\n>\\n> Probably not enough, see below. Note the escaping chars should also be escaped.\\n>\\n>>> I`m able to sanitize characters, but complex sequences will be problem. I\\n>>> will look on this, but I don`t know, if I`m able to do this.\\n>\\n> I do not know whether only those are necessary. Have you checked? Guessing is probably not the right approach.\\n>\\n>\\n> Concerning MARKDOWN, and according to the following source about github markdown implementation:\\n>\\n> https://enterprise.github.com/downloads/en/markdown-cheatsheet.pdf\\n>\\n> The following characters may need to be backslash escaped, although it does not cover HTML stuff.\\n>\\n> \\\\ backslash\\n> ` backtick\\n> * asterisk\\n> _ underscore\\n> {} curly braces\\n> [] square brackets\\n> () parentheses\\n> # hash mark\\n> + plus sign\\n> - minus sign (hyphen)\\n> . dot\\n> ! exclamation\\n>\\n> Another source https://genius.com/3057216 suggests (* # / ( ) [ ] < >),\\n> which should protect HTML.\\n>\\n> However, the escaping seems to be the backslash character, NOT using html encoding < as done in your version.\\n>\\n> Where did you find the precise escaping rules for markdown? I do not think that it should be invented...\\n>\\n>\\n> I have looked at RST, according to this reference:\\n>\\n> http://docutils.sourceforge.net/docs/ref/rst/restructuredtext.html#grid-tables\\n>\\n> The good news is that you do not need to handle a special | case because you would only produce clean tables.\\n>\\n> I've tested UTF-8 with plane 1 (你好!) and plane 2 (𠀡) and the alignment seems to worked well, incredible!\\n>\\n>>> My main interest on this was in rst. I`m using markdown only in github issues and my knowldge about md is poor.\\n>\\n> Then maybe only do RST?!\\n>\\n> It looks much simpler anyway, and if you do MARKDOWN the support needs to be clean.\\n>\\n> About the code:\\n>\\n> I'm still at odds with the code which needs to test for markdown to call for different functions in multiple places. If you keep md and in order to avoid that, I would suggest to extend the pg_wcs* functions with a list of caracters which may have different sizes with additionnal args, say:\\n>\\n> pg_wcssize(// same args, plus:\\n> char * escaped_chars, // will require escaping\\n> int escape_len, // how many chars added when escaping\\n> int nllen // len of newline if substituted\\n> );\\n>\\n> So that pg_wcssize(..., \\\"\\\\r\\\", 1, 1) would behave as before (\\\\n and \\\\t are rather special cases), and the various constants could be held in the format description so the whole thing would be parametric.\\n>\\n> Same approach with format.\\n>\\n> That would allow to simplify the code significantly and to share it between MARKDOWN and others. Also, the multiple \\\"else if\\\" list would be simplified by using strchr or the escaped_chars string.\\n\\nThis patch was moved into the current Commitfest marked “Waiting for author”\\nwith the above review. Have you had a chance to work on it addressing the\\nreview comments such that we can expect a new version within this CF?\\n\\ncheers ./daniel-- JelenStarší čeledín datovýho chlíva\",\n \"msg_date\": \"Mon, 18 Sep 2017 15:31:44 +0200\",\n \"msg_from\": \"=?UTF-8?Q?Jan_Mich=C3=A1lek?= <godzilalalala@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"> On 18 Sep 2017, at 15:31, Jan Michálek <godzilalalala@gmail.com> wrote:\\n> \\n> I have`t any new version recently, because i was waiting for new version. But I will have some time on this in next months. How many time I have to next freeze?\\n\\nThe next commitfest runs from 2017-11-01 to 2017-11-30, I will move your patch\\nthere.\\n\\ncheers ./daniel\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Mon, 18 Sep 2017 15:33:24 +0200\",\n \"msg_from\": \"Daniel Gustafsson <daniel@yesql.se>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"Thanks\\n\\n2017-09-18 15:33 GMT+02:00 Daniel Gustafsson <daniel@yesql.se>:\\n\\n> > On 18 Sep 2017, at 15:31, Jan Michálek <godzilalalala@gmail.com> wrote:\\n> >\\n> > I have`t any new version recently, because i was waiting for new\\n> version. But I will have some time on this in next months. How many time I\\n> have to next freeze?\\n>\\n> The next commitfest runs from 2017-11-01 to 2017-11-30, I will move your\\n> patch\\n> there.\\n>\\n> cheers ./daniel\\n\\n\\n\\n\\n-- \\nJelen\\nStarší čeledín datovýho chlíva\\n\\nThanks2017-09-18 15:33 GMT+02:00 Daniel Gustafsson <daniel@yesql.se>:> On 18 Sep 2017, at 15:31, Jan Michálek <godzilalalala@gmail.com> wrote:\\n>\\n> I have`t any new version recently, because i was waiting for new version. But I will have some time on this in next months. How many time I have to next freeze?\\n\\nThe next commitfest runs from 2017-11-01 to 2017-11-30, I will move your patch\\nthere.\\n\\ncheers ./daniel-- JelenStarší čeledín datovýho chlíva\",\n \"msg_date\": \"Mon, 18 Sep 2017 15:39:51 +0200\",\n \"msg_from\": \"=?UTF-8?Q?Jan_Mich=C3=A1lek?= <godzilalalala@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"On Mon, Sep 18, 2017 at 10:33 PM, Daniel Gustafsson <daniel@yesql.se> wrote:\\n>> On 18 Sep 2017, at 15:31, Jan Michálek <godzilalalala@gmail.com> wrote:\\n>>\\n>> I have`t any new version recently, because i was waiting for new version. But I will have some time on this in next months. How many time I have to next freeze?\\n>\\n> The next commitfest runs from 2017-11-01 to 2017-11-30, I will move your patch\\n> there.\\n\\nIt's been 1 month and a half since this last status, and no updates\\nhave showed up. I am marking this patch as returned with feedback. If\\nyou are able to reproduce a new version, please make sure to register\\nit in the commit fest for review.\\n-- \\nMichael\\n\\n\",\n \"msg_date\": \"Tue, 28 Nov 2017 10:43:10 +0900\",\n \"msg_from\": \"Michael Paquier <michael.paquier@gmail.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"Sorry, I haven`t any new progres.\\nCan you move my patch to next (or it is problem)? I have to solve some\\npersonal issues. I want to finish this, but im not able to promise, that it\\nwill be in next months.\\n\\nJe;\\n\\n2017-11-28 2:43 GMT+01:00 Michael Paquier <michael.paquier@gmail.com>:\\n\\n> On Mon, Sep 18, 2017 at 10:33 PM, Daniel Gustafsson <daniel@yesql.se>\\n> wrote:\\n> >> On 18 Sep 2017, at 15:31, Jan Michálek <godzilalalala@gmail.com> wrote:\\n> >>\\n> >> I have`t any new version recently, because i was waiting for new\\n> version. But I will have some time on this in next months. How many time I\\n> have to next freeze?\\n> >\\n> > The next commitfest runs from 2017-11-01 to 2017-11-30, I will move your\\n> patch\\n> > there.\\n>\\n> It's been 1 month and a half since this last status, and no updates\\n> have showed up. I am marking this patch as returned with feedback. If\\n> you are able to reproduce a new version, please make sure to register\\n> it in the commit fest for review.\\n> --\\n> Michael\\n>\\n\\n\\n\\n-- \\nJelen\\nStarší čeledín datovýho chlíva\\n\\nSorry, I haven`t any new progres.Can you move my patch to next (or it is problem)? I have to solve some personal issues. I want to finish this, but im not able to promise, that it will be in next months.Je;2017-11-28 2:43 GMT+01:00 Michael Paquier <michael.paquier@gmail.com>:On Mon, Sep 18, 2017 at 10:33 PM, Daniel Gustafsson <daniel@yesql.se> wrote:\\n>> On 18 Sep 2017, at 15:31, Jan Michálek <godzilalalala@gmail.com> wrote:\\n>>\\n>> I have`t any new version recently, because i was waiting for new version. But I will have some time on this in next months. How many time I have to next freeze?\\n>\\n> The next commitfest runs from 2017-11-01 to 2017-11-30, I will move your patch\\n> there.\\n\\nIt's been 1 month and a half since this last status, and no updates\\nhave showed up. I am marking this patch as returned with feedback. If\\nyou are able to reproduce a new version, please make sure to register\\nit in the commit fest for review.\\n--\\nMichael\\n-- JelenStarší čeledín datovýho chlíva\",\n \"msg_date\": \"Tue, 28 Nov 2017 12:01:45 +0100\",\n \"msg_from\": \"=?UTF-8?Q?Jan_Mich=C3=A1lek?= <godzilalalala@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: [HACKERS] Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"\\n> On 28 Nov 2017, at 12:01, Jan Michálek <godzilalalala@gmail.com> wrote:\\n> \\n> Sorry, I haven`t any new progres.\\n> Can you move my patch to next (or it is problem)? I have to solve some personal issues. I want to finish this, but im not able to promise, that it will be in next months.\\n\\nNo problems, it has been returned with feedback which doesn’t mean “no thanks,\\nit means “re-register it in the then open commitfest when you have had time to\\nfix a new version.\\n\\nWhen you have had time to hack on it, just add it again and it will be up for\\nreview.\\n\\ncheers ./daniel\\n\",\n \"msg_date\": \"Tue, 28 Nov 2017 12:07:53 +0100\",\n \"msg_from\": \"Daniel Gustafsson <daniel@yesql.se>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"Thanks\\nMy english is little bit poor and my understanding of formal part of commit\\nprocess is only partial.\\n\\n2017-11-28 12:07 GMT+01:00 Daniel Gustafsson <daniel@yesql.se>:\\n\\n>\\n> > On 28 Nov 2017, at 12:01, Jan Michálek <godzilalalala@gmail.com> wrote:\\n> >\\n> > Sorry, I haven`t any new progres.\\n> > Can you move my patch to next (or it is problem)? I have to solve some\\n> personal issues. I want to finish this, but im not able to promise, that it\\n> will be in next months.\\n>\\n> No problems, it has been returned with feedback which doesn’t mean “no\\n> thanks,\\n> it means “re-register it in the then open commitfest when you have had\\n> time to\\n> fix a new version.\\n>\\n> When you have had time to hack on it, just add it again and it will be up\\n> for\\n> review.\\n>\\n> cheers ./daniel\\n\\n\\n\\n\\n-- \\nJelen\\nStarší čeledín datovýho chlíva\\n\\nThanksMy english is little bit poor and my understanding of formal part of commit process is only partial.2017-11-28 12:07 GMT+01:00 Daniel Gustafsson <daniel@yesql.se>:\\n> On 28 Nov 2017, at 12:01, Jan Michálek <godzilalalala@gmail.com> wrote:\\n>\\n> Sorry, I haven`t any new progres.\\n> Can you move my patch to next (or it is problem)? I have to solve some personal issues. I want to finish this, but im not able to promise, that it will be in next months.\\n\\nNo problems, it has been returned with feedback which doesn’t mean “no thanks,\\nit means “re-register it in the then open commitfest when you have had time to\\nfix a new version.\\n\\nWhen you have had time to hack on it, just add it again and it will be up for\\nreview.\\n\\ncheers ./daniel-- JelenStarší čeledín datovýho chlíva\",\n \"msg_date\": \"Tue, 28 Nov 2017 13:20:51 +0100\",\n \"msg_from\": \"=?UTF-8?Q?Jan_Mich=C3=A1lek?= <godzilalalala@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: [HACKERS] Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"On Tue, Nov 28, 2017 at 9:20 PM, Jan Michálek <godzilalalala@gmail.com> wrote:\\n> Thanks\\n\\nPlease avoid top-posting. This breaks the thread logic.\\n\\n> My english is little bit poor and my understanding of formal part of commit\\n> process is only partial.\\n\\nDon't worry, we'll all here to learn. If you are able to hack a new\\nversion, please feel free to post a new patch and register a new entry\\nin the commit fest application.\\n-- \\nMichael\\n\\n\",\n \"msg_date\": \"Tue, 28 Nov 2017 21:32:15 +0900\",\n \"msg_from\": \"Michael Paquier <michael.paquier@gmail.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Other formats in pset like markdown, rst, mediawiki\"\n },\n {\n \"msg_contents\": \"2017-11-28 13:32 GMT+01:00 Michael Paquier <michael.paquier@gmail.com>:\\n\\n> On Tue, Nov 28, 2017 at 9:20 PM, Jan Michálek <godzilalalala@gmail.com>\\n> wrote:\\n> > Thanks\\n>\\n> Please avoid top-posting. This breaks the thread logic.\\n>\\n\\nOh, sorry\\n\\n\\n>\\n> > My english is little bit poor and my understanding of formal part of\\n> commit\\n> > process is only partial.\\n>\\n> Don't worry, we'll all here to learn. If you are able to hack a new\\n> version, please feel free to post a new patch and register a new entry\\n> in the commit fest application.\\n> --\\n> Michael\\n>\\n\\n\\n\\n-- \\nJelen\\nStarší čeledín datovýho chlíva\\n\\n2017-11-28 13:32 GMT+01:00 Michael Paquier <michael.paquier@gmail.com>:On Tue, Nov 28, 2017 at 9:20 PM, Jan Michálek <godzilalalala@gmail.com> wrote:\\n> Thanks\\n\\nPlease avoid top-posting. This breaks the thread logic.Oh, sorry \\n\\n> My english is little bit poor and my understanding of formal part of commit\\n> process is only partial.\\n\\nDon't worry, we'll all here to learn. If you are able to hack a new\\nversion, please feel free to post a new patch and register a new entry\\nin the commit fest application.\\n--\\nMichael\\n-- JelenStarší čeledín datovýho chlíva\",\n \"msg_date\": \"Tue, 28 Nov 2017 13:36:22 +0100\",\n \"msg_from\": \"=?UTF-8?Q?Jan_Mich=C3=A1lek?= <godzilalalala@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: [HACKERS] Other formats in pset like markdown, rst, mediawiki\"\n }\n]","isConversational":false}}},{"rowIdx":43997,"cells":{"threads":{"kind":"list like","value":[{"msg_contents":"Hi\n\nCurrently is not possible to control sort columns for \\d* commands. Usually\nschema and table name is used. Really often task is collect the most big\nobjects in database. \"\\dt+, \\di+\" shows necessary information, but not in\npractical order.\n\nInstead introduction some additional flags to backslash commands, I propose\na special psql variable that can be used for specification of order used\nwhen some plus command is used.\n\nsome like\n\nset EXTENDED_DESCRIBE_SORT size_desc\n\\dt+\n\\l+\n\\di+\n\nPossible variants: schema_table, table_schema, size_desc, size_asc\n\nComments, notes?\n\nRegards\n\nPavel\n\nHiCurrently is not possible to control sort columns for \\d* commands. Usually schema and table name is used. Really often task is collect the most big objects in database. \"\\dt+, \\di+\" shows necessary information, but not in practical order.Instead introduction some additional flags to backslash commands, I propose a special psql variable that can be used for specification of order used when some plus command is used.some likeset EXTENDED_DESCRIBE_SORT size_desc\\dt+\\l+\\di+Possible variants: schema_table, table_schema, size_desc, size_ascComments, notes?RegardsPavel","msg_date":"Thu, 23 Feb 2017 12:17:31 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"proposal - psql: possibility to specify sort for describe commands,\n when size is printed"},{"msg_contents":"Hi\n\n2017-02-23 12:17 GMT+01:00 Pavel Stehule <pavel.stehule@gmail.com>:\n\n> Hi\n>\n> Currently is not possible to control sort columns for \\d* commands.\n> Usually schema and table name is used. Really often task is collect the\n> most big objects in database. \"\\dt+, \\di+\" shows necessary information, but\n> not in practical order.\n>\n> Instead introduction some additional flags to backslash commands, I\n> propose a special psql variable that can be used for specification of order\n> used when some plus command is used.\n>\n> some like\n>\n> set EXTENDED_DESCRIBE_SORT size_desc\n> \\dt+\n> \\l+\n> \\di+\n>\n> Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> Comments, notes?\n>\n\nhere is a patch\n\nRegards\n\nPavel\n\n\n>\n> Regards\n>\n> Pavel\n>\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers","msg_date":"Fri, 24 Feb 2017 22:32:47 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: proposal - psql: possibility to specify sort for describe\n commands, when size is printed"},{"msg_contents":"On 2/24/17 16:32, Pavel Stehule wrote:\n> set EXTENDED_DESCRIBE_SORT size_desc\n> \\dt+\n> \\l+\n> \\di+\n> \n> Possible variants: schema_table, table_schema, size_desc, size_asc\n\nI can see this being useful, but I think it needs to be organized a\nlittle better.\n\nSort key and sort direction should be separate settings.\n\nI'm not sure why we need to have separate settings to sort by schema\nname and table name. But if we do, then we should support that for all\nobject types. I think maybe that's something we shouldn't get into\nright now.\n\nSo I would have one setting for sort key = {name|size} and on for sort\ndirection = {asc|desc}.\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Fri, 10 Mar 2017 09:10:08 -0500","msg_from":"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> > \n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n> \n> I can see this being useful, but I think it needs to be organized a\n> little better.\n> \n> Sort key and sort direction should be separate settings.\n> \n> I'm not sure why we need to have separate settings to sort by schema\n> name and table name. But if we do, then we should support that for all\n> object types. I think maybe that's something we shouldn't get into\n> right now.\n> \n> So I would have one setting for sort key = {name|size} and on for sort\n> direction = {asc|desc}.\n\nPerhaps I'm trying to be overly cute here, but why not let the user\nsimply provide a bit of SQL to be put at the end of the query?\n\nThat is, something like:\n\n\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\n\nThanks!\n\nStephen","msg_date":"Fri, 10 Mar 2017 09:16:45 -0500","msg_from":"Stephen Frost <sfrost@snowman.net>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Fri, Mar 10, 2017 at 5:10 PM, Peter Eisentraut <\npeter.eisentraut@2ndquadrant.com> wrote:\n\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> >\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> I can see this being useful, but I think it needs to be organized a\n> little better.\n>\n> Sort key and sort direction should be separate settings.\n>\n\nI agree.\n\nI'm not sure why we need to have separate settings to sort by schema\n> name and table name.\n\n\nI think sorting by schema name, object name makes sense for people, who\nhave objects of same name in different schemas.\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Fri, Mar 10, 2017 at 5:10 PM, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:On 2/24/17 16:32, Pavel Stehule wrote:\n> set EXTENDED_DESCRIBE_SORT size_desc\n> \\dt+\n> \\l+\n> \\di+\n>\n> Possible variants: schema_table, table_schema, size_desc, size_asc\n\nI can see this being useful, but I think it needs to be organized a\nlittle better.\n\nSort key and sort direction should be separate settings. I agree.\nI'm not sure why we need to have separate settings to sort by schema\nname and table name.I think sorting by schema name, object name makes sense for people, who have objects of same name in different schemas.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Fri, 10 Mar 2017 17:45:43 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-03-10 15:10 GMT+01:00 Peter Eisentraut <\npeter.eisentraut@2ndquadrant.com>:\n\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> >\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> I can see this being useful, but I think it needs to be organized a\n> little better.\n>\n> Sort key and sort direction should be separate settings.\n>\n\nok\n\nmaybe\n\nPREFERRED_SORT_COLUMNS\nand PREFERRED_SORT_DIRECTION ?\n\n>\n> I'm not sure why we need to have separate settings to sort by schema\n> name and table name. But if we do, then we should support that for all\n> object types. I think maybe that's something we shouldn't get into\n> right now.\n>\n\nIt can be useful, when you repeat one table name in more schema - usually,\nwhere schema is related per one customer, project, ...\n\nRegards\n\nPavel\n\n\n>\n> So I would have one setting for sort key = {name|size} and on for sort\n> direction = {asc|desc}.\n>\n> --\n> Peter Eisentraut http://www.2ndQuadrant.com/\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>\n\n2017-03-10 15:10 GMT+01:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 2/24/17 16:32, Pavel Stehule wrote:\n> set EXTENDED_DESCRIBE_SORT size_desc\n> \\dt+\n> \\l+\n> \\di+\n>\n> Possible variants: schema_table, table_schema, size_desc, size_asc\n\nI can see this being useful, but I think it needs to be organized a\nlittle better.\n\nSort key and sort direction should be separate settings.okmaybePREFERRED_SORT_COLUMNSand PREFERRED_SORT_DIRECTION ? \n\nI'm not sure why we need to have separate settings to sort by schema\nname and table name. But if we do, then we should support that for all\nobject types. I think maybe that's something we shouldn't get into\nright now.It can be useful, when you repeat one table name in more schema - usually, where schema is related per one customer, project, ...RegardsPavel \n\nSo I would have one setting for sort key = {name|size} and on for sort\ndirection = {asc|desc}.\n\n--\nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services","msg_date":"Fri, 10 Mar 2017 15:57:38 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:\n\n> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> > On 2/24/17 16:32, Pavel Stehule wrote:\n> > > set EXTENDED_DESCRIBE_SORT size_desc\n> > > \\dt+\n> > > \\l+\n> > > \\di+\n> > >\n> > > Possible variants: schema_table, table_schema, size_desc, size_asc\n> >\n> > I can see this being useful, but I think it needs to be organized a\n> > little better.\n> >\n> > Sort key and sort direction should be separate settings.\n> >\n> > I'm not sure why we need to have separate settings to sort by schema\n> > name and table name. But if we do, then we should support that for all\n> > object types. I think maybe that's something we shouldn't get into\n> > right now.\n> >\n> > So I would have one setting for sort key = {name|size} and on for sort\n> > direction = {asc|desc}.\n>\n> Perhaps I'm trying to be overly cute here, but why not let the user\n> simply provide a bit of SQL to be put at the end of the query?\n>\n> That is, something like:\n>\n> \\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\n>\n\nI think that's the question of usability. After all, one can manually type\ncorresponding SQL instead of \\d* commands. However, it's quite cumbersome\nto do this every time.\nI found quite useful to being able to switch between different sortings\nquickly. For instance, after seeing tables sorted by name, user can\nrequire them sorted by size descending, then sorted by size ascending,\netc...\nTherefore, I find user-defined SQL clause to be cumbersome. Even psql\nvariable itself seems to be cumbersome for me.\nI would propose to add sorting as second optional argument to \\d*\ncommands. Any thoughts?\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> >\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> I can see this being useful, but I think it needs to be organized a\n> little better.\n>\n> Sort key and sort direction should be separate settings.\n>\n> I'm not sure why we need to have separate settings to sort by schema\n> name and table name. But if we do, then we should support that for all\n> object types. I think maybe that's something we shouldn't get into\n> right now.\n>\n> So I would have one setting for sort key = {name|size} and on for sort\n> direction = {asc|desc}.\n\nPerhaps I'm trying to be overly cute here, but why not let the user\nsimply provide a bit of SQL to be put at the end of the query?\n\nThat is, something like:\n\n\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\d* commands. Any thoughts?------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Fri, 10 Mar 2017 18:00:14 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-03-10 15:16 GMT+01:00 Stephen Frost <sfrost@snowman.net>:\n\n> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> > On 2/24/17 16:32, Pavel Stehule wrote:\n> > > set EXTENDED_DESCRIBE_SORT size_desc\n> > > \\dt+\n> > > \\l+\n> > > \\di+\n> > >\n> > > Possible variants: schema_table, table_schema, size_desc, size_asc\n> >\n> > I can see this being useful, but I think it needs to be organized a\n> > little better.\n> >\n> > Sort key and sort direction should be separate settings.\n> >\n> > I'm not sure why we need to have separate settings to sort by schema\n> > name and table name. But if we do, then we should support that for all\n> > object types. I think maybe that's something we shouldn't get into\n> > right now.\n> >\n> > So I would have one setting for sort key = {name|size} and on for sort\n> > direction = {asc|desc}.\n>\n> Perhaps I'm trying to be overly cute here, but why not let the user\n> simply provide a bit of SQL to be put at the end of the query?\n>\n> That is, something like:\n>\n> \\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\n>\n\nFor example - the size is displayed in pretty form - raw form is not\ndisplayed - so simple ORDER BY clause is not possible.\n\nBut setting LIMIT is not bad idea - although it is probably much more\ncomplex for implementation.\n\n\\pset DESCRIBE_LIMIT 100\n\\pset EXTENDED_DESCRIBE_LIMIT 100\n\ncan be implemented as next step\n\nRegards\n\nPavel\n\n\n\n>\n> Thanks!\n>\n> Stephen\n>\n\n2017-03-10 15:16 GMT+01:00 Stephen Frost <sfrost@snowman.net>:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> >\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> I can see this being useful, but I think it needs to be organized a\n> little better.\n>\n> Sort key and sort direction should be separate settings.\n>\n> I'm not sure why we need to have separate settings to sort by schema\n> name and table name. But if we do, then we should support that for all\n> object types. I think maybe that's something we shouldn't get into\n> right now.\n>\n> So I would have one setting for sort key = {name|size} and on for sort\n> direction = {asc|desc}.\n\nPerhaps I'm trying to be overly cute here, but why not let the user\nsimply provide a bit of SQL to be put at the end of the query?\n\nThat is, something like:\n\n\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'For example - the size is displayed in pretty form - raw form is not displayed - so simple ORDER BY clause is not possible.But setting LIMIT is not bad idea - although it is probably much more complex for implementation.\\pset DESCRIBE_LIMIT 100\\pset EXTENDED_DESCRIBE_LIMIT 100can be implemented as next stepRegardsPavel \n\nThanks!\n\nStephen","msg_date":"Fri, 10 Mar 2017 16:01:13 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On 3/10/17 09:57, Pavel Stehule wrote:\n> PREFERRED_SORT_COLUMNS\n> and PREFERRED_SORT_DIRECTION ? \n\nI think the name \"preferred\" implies that it will be ignored if it's not\nfound or something like that, but I don't think that's what you are\nimplementing.\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Fri, 10 Mar 2017 10:05:04 -0500","msg_from":"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n\n> On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:\n>\n>> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n>> > On 2/24/17 16:32, Pavel Stehule wrote:\n>> > > set EXTENDED_DESCRIBE_SORT size_desc\n>> > > \\dt+\n>> > > \\l+\n>> > > \\di+\n>> > >\n>> > > Possible variants: schema_table, table_schema, size_desc, size_asc\n>> >\n>> > I can see this being useful, but I think it needs to be organized a\n>> > little better.\n>> >\n>> > Sort key and sort direction should be separate settings.\n>> >\n>> > I'm not sure why we need to have separate settings to sort by schema\n>> > name and table name. But if we do, then we should support that for all\n>> > object types. I think maybe that's something we shouldn't get into\n>> > right now.\n>> >\n>> > So I would have one setting for sort key = {name|size} and on for sort\n>> > direction = {asc|desc}.\n>>\n>> Perhaps I'm trying to be overly cute here, but why not let the user\n>> simply provide a bit of SQL to be put at the end of the query?\n>>\n>> That is, something like:\n>>\n>> \\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\n>>\n>\n> I think that's the question of usability. After all, one can manually\n> type corresponding SQL instead of \\d* commands. However, it's quite\n> cumbersome to do this every time.\n> I found quite useful to being able to switch between different sortings\n> quickly. For instance, after seeing tables sorted by name, user can\n> require them sorted by size descending, then sorted by size ascending,\n> etc...\n> Therefore, I find user-defined SQL clause to be cumbersome. Even psql\n> variable itself seems to be cumbersome for me.\n> I would propose to add sorting as second optional argument to \\d*\n> commands. Any thoughts?\n>\n\nThis proposal was here already - maybe two years ago. The psql command\nparser doesn't allow any complex syntax - more - the more parameters in one\npsql commands is hard to remember, hard to read.\n\nWith my proposal, and patch I would to cover following use case. It is real\ncase. Anytime when we used \\dt+ in psql we needed sort by size desc. When\nwe should to see a size, then the top is interesting. This case is not\nabsolute, but very often, so I would to create some simple way, how to do\nsome parametrization (really simple).\n\nPavel\n\n\n\n>\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n\n2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> >\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> I can see this being useful, but I think it needs to be organized a\n> little better.\n>\n> Sort key and sort direction should be separate settings.\n>\n> I'm not sure why we need to have separate settings to sort by schema\n> name and table name. But if we do, then we should support that for all\n> object types. I think maybe that's something we shouldn't get into\n> right now.\n>\n> So I would have one setting for sort key = {name|size} and on for sort\n> direction = {asc|desc}.\n\nPerhaps I'm trying to be overly cute here, but why not let the user\nsimply provide a bit of SQL to be put at the end of the query?\n\nThat is, something like:\n\n\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\d* commands. Any thoughts?This proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.With my proposal, and patch I would to cover following use case. It is real case. Anytime when we used \\dt+ in psql we needed sort by size desc. When we should to see a size, then the top is interesting. This case is not absolute, but very often, so I would to create some simple way, how to do some parametrization (really simple).Pavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Fri, 10 Mar 2017 16:06:52 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-03-10 16:05 GMT+01:00 Peter Eisentraut <\npeter.eisentraut@2ndquadrant.com>:\n\n> On 3/10/17 09:57, Pavel Stehule wrote:\n> > PREFERRED_SORT_COLUMNS\n> > and PREFERRED_SORT_DIRECTION ?\n>\n> I think the name \"preferred\" implies that it will be ignored if it's not\n> found or something like that, but I don't think that's what you are\n> implementing.\n>\n\nok if it will be used only for verbose describe commands , then the name\nEXTENDED_DESCRIBE_SORT_COLUMNS, and EXTENDED_DESCRIBE_SORT_DIRECTION.\n\nPavel\n\n\n>\n> --\n> Peter Eisentraut http://www.2ndQuadrant.com/\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>\n\n2017-03-10 16:05 GMT+01:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 3/10/17 09:57, Pavel Stehule wrote:\n> PREFERRED_SORT_COLUMNS\n> and PREFERRED_SORT_DIRECTION ?\n\nI think the name \"preferred\" implies that it will be ignored if it's not\nfound or something like that, but I don't think that's what you are\nimplementing.ok if it will be used only for verbose describe commands , then the name EXTENDED_DESCRIBE_SORT_COLUMNS, and EXTENDED_DESCRIBE_SORT_DIRECTION.Pavel \n\n--\nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services","msg_date":"Fri, 10 Mar 2017 16:08:58 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-03-10 15:45 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n\n> On Fri, Mar 10, 2017 at 5:10 PM, Peter Eisentraut <\n> peter.eisentraut@2ndquadrant.com> wrote:\n>\n>> On 2/24/17 16:32, Pavel Stehule wrote:\n>> > set EXTENDED_DESCRIBE_SORT size_desc\n>> > \\dt+\n>> > \\l+\n>> > \\di+\n>> >\n>> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>>\n>> I can see this being useful, but I think it needs to be organized a\n>> little better.\n>>\n>> Sort key and sort direction should be separate settings.\n>>\n>\n> I agree.\n>\n> I'm not sure why we need to have separate settings to sort by schema\n>> name and table name.\n>\n>\n> I think sorting by schema name, object name makes sense for people, who\n> have objects of same name in different schemas.\n>\n\nI am sending a updated version with separated sort direction in special\nvariable\n\nThere is a question. Has desc direction sense for columns like schema or\ntable name?\n\nUsing desc, asc for size is natural. But for tablename?\n\nRegards\n\nPavel\n\n\n\n>\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n>\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers","msg_date":"Sat, 11 Mar 2017 13:06:13 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"Hi Alexander,\n\nOn 3/11/17 7:06 AM, Pavel Stehule wrote:\n\n> I am sending a updated version with separated sort direction in special\n> variable\n>\n> There is a question. Has desc direction sense for columns like schema or\n> table name?\n>\n> Using desc, asc for size is natural. But for tablename?\n\nDo you know when you'll have a chance to review the updated patch?\n\nThanks,\n-- \n-David\ndavid@pgmasters.net\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Tue, 21 Mar 2017 18:26:59 -0400","msg_from":"David Steele <david@pgmasters.net>","msg_from_op":false,"msg_subject":"Re: proposal - psql: possibility to specify sort for describe\n commands, when size is printed"},{"msg_contents":"On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com>\nwrote:\n\n> 2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n>\n>> On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net>\n>> wrote:\n>>\n>>> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n>>> > On 2/24/17 16:32, Pavel Stehule wrote:\n>>> > > set EXTENDED_DESCRIBE_SORT size_desc\n>>> > > \\dt+\n>>> > > \\l+\n>>> > > \\di+\n>>> > >\n>>> > > Possible variants: schema_table, table_schema, size_desc,\n>>> size_asc\n>>> >\n>>> > I can see this being useful, but I think it needs to be organized a\n>>> > little better.\n>>> >\n>>> > Sort key and sort direction should be separate settings.\n>>> >\n>>> > I'm not sure why we need to have separate settings to sort by schema\n>>> > name and table name. But if we do, then we should support that for all\n>>> > object types. I think maybe that's something we shouldn't get into\n>>> > right now.\n>>> >\n>>> > So I would have one setting for sort key = {name|size} and on for sort\n>>> > direction = {asc|desc}.\n>>>\n>>> Perhaps I'm trying to be overly cute here, but why not let the user\n>>> simply provide a bit of SQL to be put at the end of the query?\n>>>\n>>> That is, something like:\n>>>\n>>> \\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\n>>>\n>>\n>> I think that's the question of usability. After all, one can manually\n>> type corresponding SQL instead of \\d* commands. However, it's quite\n>> cumbersome to do this every time.\n>> I found quite useful to being able to switch between different sortings\n>> quickly. For instance, after seeing tables sorted by name, user can\n>> require them sorted by size descending, then sorted by size ascending,\n>> etc...\n>> Therefore, I find user-defined SQL clause to be cumbersome. Even psql\n>> variable itself seems to be cumbersome for me.\n>> I would propose to add sorting as second optional argument to \\d*\n>> commands. Any thoughts?\n>>\n>\n> This proposal was here already - maybe two years ago. The psql command\n> parser doesn't allow any complex syntax - more - the more parameters in one\n> psql commands is hard to remember, hard to read.\n>\n\nCould you please provide a link to this discussion. Probably working with\nmultiple parameters in psql commands require some rework, but that's\ndefinitely doable.\n\n\n> With my proposal, and patch I would to cover following use case. It is\n> real case. Anytime when we used \\dt+ in psql we needed sort by size desc.\n> When we should to see a size, then the top is interesting. This case is not\n> absolute, but very often, so I would to create some simple way, how to do\n> some parametrization (really simple).\n>\n\nWe could combine both approaches: add parameters to psql commands and add\npsql DEFAULT_(SORT_COLUMNS|DIRECTION|LIMIT) parameters.\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> >\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> I can see this being useful, but I think it needs to be organized a\n> little better.\n>\n> Sort key and sort direction should be separate settings.\n>\n> I'm not sure why we need to have separate settings to sort by schema\n> name and table name. But if we do, then we should support that for all\n> object types. I think maybe that's something we shouldn't get into\n> right now.\n>\n> So I would have one setting for sort key = {name|size} and on for sort\n> direction = {asc|desc}.\n\nPerhaps I'm trying to be overly cute here, but why not let the user\nsimply provide a bit of SQL to be put at the end of the query?\n\nThat is, something like:\n\n\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\d* commands. Any thoughts?This proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.Could you please provide a link to this discussion. Probably working with multiple parameters in psql commands require some rework, but that's definitely doable. With my proposal, and patch I would to cover following use case. It is real case. Anytime when we used \\dt+ in psql we needed sort by size desc. When we should to see a size, then the top is interesting. This case is not absolute, but very often, so I would to create some simple way, how to do some parametrization (really simple).We could combine both approaches: add parameters to psql commands and add psql DEFAULT_(SORT_COLUMNS|DIRECTION|LIMIT) parameters.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Mon, 27 Mar 2017 14:59:32 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-03-27 13:59 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n\n> On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com>\n> wrote:\n>\n>> 2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>\n>> :\n>>\n>>> On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net>\n>>> wrote:\n>>>\n>>>> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n>>>> > On 2/24/17 16:32, Pavel Stehule wrote:\n>>>> > > set EXTENDED_DESCRIBE_SORT size_desc\n>>>> > > \\dt+\n>>>> > > \\l+\n>>>> > > \\di+\n>>>> > >\n>>>> > > Possible variants: schema_table, table_schema, size_desc,\n>>>> size_asc\n>>>> >\n>>>> > I can see this being useful, but I think it needs to be organized a\n>>>> > little better.\n>>>> >\n>>>> > Sort key and sort direction should be separate settings.\n>>>> >\n>>>> > I'm not sure why we need to have separate settings to sort by schema\n>>>> > name and table name. But if we do, then we should support that for\n>>>> all\n>>>> > object types. I think maybe that's something we shouldn't get into\n>>>> > right now.\n>>>> >\n>>>> > So I would have one setting for sort key = {name|size} and on for sort\n>>>> > direction = {asc|desc}.\n>>>>\n>>>> Perhaps I'm trying to be overly cute here, but why not let the user\n>>>> simply provide a bit of SQL to be put at the end of the query?\n>>>>\n>>>> That is, something like:\n>>>>\n>>>> \\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\n>>>>\n>>>\n>>> I think that's the question of usability. After all, one can manually\n>>> type corresponding SQL instead of \\d* commands. However, it's quite\n>>> cumbersome to do this every time.\n>>> I found quite useful to being able to switch between different sortings\n>>> quickly. For instance, after seeing tables sorted by name, user can\n>>> require them sorted by size descending, then sorted by size ascending,\n>>> etc...\n>>> Therefore, I find user-defined SQL clause to be cumbersome. Even psql\n>>> variable itself seems to be cumbersome for me.\n>>> I would propose to add sorting as second optional argument to \\d*\n>>> commands. Any thoughts?\n>>>\n>>\n>> This proposal was here already - maybe two years ago. The psql command\n>> parser doesn't allow any complex syntax - more - the more parameters in one\n>> psql commands is hard to remember, hard to read.\n>>\n>\n> Could you please provide a link to this discussion. Probably working with\n> multiple parameters in psql commands require some rework, but that's\n> definitely doable.\n>\n\nhttp://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/proposal-psql-show-longest-tables/oldest\nhttps://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.com\n\n\n>\n>> With my proposal, and patch I would to cover following use case. It is\n>> real case. Anytime when we used \\dt+ in psql we needed sort by size desc.\n>> When we should to see a size, then the top is interesting. This case is not\n>> absolute, but very often, so I would to create some simple way, how to do\n>> some parametrization (really simple).\n>>\n>\n> We could combine both approaches: add parameters to psql commands and add\n> psql DEFAULT_(SORT_COLUMNS|DIRECTION|LIMIT) parameters.\n>\n\nIt is possible - This moment is my interest concentrated to psql settings -\nthe unpractical order in \\dt+ irritate me :). I understand so it depends on\nuse-case. I worked in OLAP and still I have lot of customers with\nperformance incidents - the first task - show most big tables, most big\nindexes.\n\nRegards\n\nPavel\n\n\n\n\n\n>\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n>\n\n2017-03-27 13:59 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> >\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> I can see this being useful, but I think it needs to be organized a\n> little better.\n>\n> Sort key and sort direction should be separate settings.\n>\n> I'm not sure why we need to have separate settings to sort by schema\n> name and table name. But if we do, then we should support that for all\n> object types. I think maybe that's something we shouldn't get into\n> right now.\n>\n> So I would have one setting for sort key = {name|size} and on for sort\n> direction = {asc|desc}.\n\nPerhaps I'm trying to be overly cute here, but why not let the user\nsimply provide a bit of SQL to be put at the end of the query?\n\nThat is, something like:\n\n\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\d* commands. Any thoughts?This proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.Could you please provide a link to this discussion. Probably working with multiple parameters in psql commands require some rework, but that's definitely doable.http://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/proposal-psql-show-longest-tables/oldesthttps://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.com With my proposal, and patch I would to cover following use case. It is real case. Anytime when we used \\dt+ in psql we needed sort by size desc. When we should to see a size, then the top is interesting. This case is not absolute, but very often, so I would to create some simple way, how to do some parametrization (really simple).We could combine both approaches: add parameters to psql commands and add psql DEFAULT_(SORT_COLUMNS|DIRECTION|LIMIT) parameters.It is possible - This moment is my interest concentrated to psql settings - the unpractical order in \\dt+ irritate me :). I understand so it depends on use-case. I worked in OLAP and still I have lot of customers with performance incidents - the first task - show most big tables, most big indexes.RegardsPavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Tue, 28 Mar 2017 09:12:26 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Tue, Mar 28, 2017 at 10:12 AM, Pavel Stehule <pavel.stehule@gmail.com>\nwrote:\n\n> 2017-03-27 13:59 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n>\n>> On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com>\n>> wrote:\n>>\n>>> 2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru\n>>> >:\n>>>\n>>>> On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net>\n>>>> wrote:\n>>>>\n>>>>> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n>>>>> > On 2/24/17 16:32, Pavel Stehule wrote:\n>>>>> > > set EXTENDED_DESCRIBE_SORT size_desc\n>>>>> > > \\dt+\n>>>>> > > \\l+\n>>>>> > > \\di+\n>>>>> > >\n>>>>> > > Possible variants: schema_table, table_schema, size_desc,\n>>>>> size_asc\n>>>>> >\n>>>>> > I can see this being useful, but I think it needs to be organized a\n>>>>> > little better.\n>>>>> >\n>>>>> > Sort key and sort direction should be separate settings.\n>>>>> >\n>>>>> > I'm not sure why we need to have separate settings to sort by schema\n>>>>> > name and table name. But if we do, then we should support that for\n>>>>> all\n>>>>> > object types. I think maybe that's something we shouldn't get into\n>>>>> > right now.\n>>>>> >\n>>>>> > So I would have one setting for sort key = {name|size} and on for\n>>>>> sort\n>>>>> > direction = {asc|desc}.\n>>>>>\n>>>>> Perhaps I'm trying to be overly cute here, but why not let the user\n>>>>> simply provide a bit of SQL to be put at the end of the query?\n>>>>>\n>>>>> That is, something like:\n>>>>>\n>>>>> \\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\n>>>>>\n>>>>\n>>>> I think that's the question of usability. After all, one can manually\n>>>> type corresponding SQL instead of \\d* commands. However, it's quite\n>>>> cumbersome to do this every time.\n>>>> I found quite useful to being able to switch between different sortings\n>>>> quickly. For instance, after seeing tables sorted by name, user can\n>>>> require them sorted by size descending, then sorted by size ascending,\n>>>> etc...\n>>>> Therefore, I find user-defined SQL clause to be cumbersome. Even psql\n>>>> variable itself seems to be cumbersome for me.\n>>>> I would propose to add sorting as second optional argument to \\d*\n>>>> commands. Any thoughts?\n>>>>\n>>>\n>>> This proposal was here already - maybe two years ago. The psql command\n>>> parser doesn't allow any complex syntax - more - the more parameters in one\n>>> psql commands is hard to remember, hard to read.\n>>>\n>>\n>> Could you please provide a link to this discussion. Probably working\n>> with multiple parameters in psql commands require some rework, but that's\n>> definitely doable.\n>>\n>\n> http://grokbase.com/t/postgresql/pgsql-hackers/\n> 137nt5p6s0/proposal-psql-show-longest-tables/oldest\n> https://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRT\n> iUQJQHwNJ8ecN2W@mail.gmail.com\n>\n\nI took a look to these threads, but I didn't find place where difficulties\nof adding extra arguments to psql commands are pointed.\nCould you, please, point particular messages about it?\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Tue, Mar 28, 2017 at 10:12 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-03-27 13:59 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\n> On 2/24/17 16:32, Pavel Stehule wrote:\n> > set EXTENDED_DESCRIBE_SORT size_desc\n> > \\dt+\n> > \\l+\n> > \\di+\n> >\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\n>\n> I can see this being useful, but I think it needs to be organized a\n> little better.\n>\n> Sort key and sort direction should be separate settings.\n>\n> I'm not sure why we need to have separate settings to sort by schema\n> name and table name. But if we do, then we should support that for all\n> object types. I think maybe that's something we shouldn't get into\n> right now.\n>\n> So I would have one setting for sort key = {name|size} and on for sort\n> direction = {asc|desc}.\n\nPerhaps I'm trying to be overly cute here, but why not let the user\nsimply provide a bit of SQL to be put at the end of the query?\n\nThat is, something like:\n\n\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\d* commands. Any thoughts?This proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.Could you please provide a link to this discussion. Probably working with multiple parameters in psql commands require some rework, but that's definitely doable.http://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/proposal-psql-show-longest-tables/oldesthttps://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.comI took a look to these threads, but I didn't find place where difficulties of adding extra arguments to psql commands are pointed.Could you, please, point particular messages about it?------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Thu, 30 Mar 2017 18:09:51 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":">\n>\n>>>>\n>>>> This proposal was here already - maybe two years ago. The psql command\n>>>> parser doesn't allow any complex syntax - more - the more parameters in one\n>>>> psql commands is hard to remember, hard to read.\n>>>>\n>>>\n>>> Could you please provide a link to this discussion. Probably working\n>>> with multiple parameters in psql commands require some rework, but that's\n>>> definitely doable.\n>>>\n>>\n>> http://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/\n>> proposal-psql-show-longest-tables/oldest\n>> https://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvq\n>> PE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.com\n>>\n>\n> I took a look to these threads, but I didn't find place where difficulties\n> of adding extra arguments to psql commands are pointed.\n> Could you, please, point particular messages about it?\n>\n\nI am sorry - maybe my memory doesn't serve well\n\nPavel\n\n\n>\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n>\n\nThis proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.Could you please provide a link to this discussion. Probably working with multiple parameters in psql commands require some rework, but that's definitely doable.http://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/proposal-psql-show-longest-tables/oldesthttps://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.comI took a look to these threads, but I didn't find place where difficulties of adding extra arguments to psql commands are pointed.Could you, please, point particular messages about it?I am sorry - maybe my memory doesn't serve well Pavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Thu, 30 Mar 2017 17:49:46 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"Hi,\n\nOn 2017-03-11 13:06:13 +0100, Pavel Stehule wrote:\n> 2017-03-10 15:45 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n> \n> > On Fri, Mar 10, 2017 at 5:10 PM, Peter Eisentraut <\n> > peter.eisentraut@2ndquadrant.com> wrote:\n> >\n> >> On 2/24/17 16:32, Pavel Stehule wrote:\n> >> > set EXTENDED_DESCRIBE_SORT size_desc\n> >> > \\dt+\n> >> > \\l+\n> >> > \\di+\n> >> >\n> >> > Possible variants: schema_table, table_schema, size_desc, size_asc\n> >>\n> >> I can see this being useful, but I think it needs to be organized a\n> >> little better.\n> >>\n> >> Sort key and sort direction should be separate settings.\n> >>\n> >\n> > I agree.\n> >\n> > I'm not sure why we need to have separate settings to sort by schema\n> >> name and table name.\n> >\n> >\n> > I think sorting by schema name, object name makes sense for people, who\n> > have objects of same name in different schemas.\n> >\n> \n> I am sending a updated version with separated sort direction in special\n> variable\n> \n> There is a question. Has desc direction sense for columns like schema or\n> table name?\n> \n> Using desc, asc for size is natural. But for tablename?\n\nI think it's pretty clear that we don't have sufficient agreement on the\ndesign, not to speak of an implementation for an agreed upon design, to\nget this into v10. The patch also has been submitted late in the v10\ncycle, and has received attention. I'm therefore moving it to the next\ncommitfest.\n\nRegards,\n\nAndres\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Wed, 5 Apr 2017 16:10:20 -0700","msg_from":"Andres Freund <andres@anarazel.de>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On 3/11/17 07:06, Pavel Stehule wrote:\n> I am sending a updated version with separated sort direction in special\n> variable\n\nThis patch also needs a rebase.\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Mon, 14 Aug 2017 22:37:35 -0400","msg_from":"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"Hi\n\n2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com\n>:\n\n> On 3/11/17 07:06, Pavel Stehule wrote:\n> > I am sending a updated version with separated sort direction in special\n> > variable\n>\n> This patch also needs a rebase.\n>\n\nI am sending rebased patch\n\nRegards\n\nPavel\n\n\n>\n> --\n> Peter Eisentraut http://www.2ndQuadrant.com/\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers","msg_date":"Wed, 16 Aug 2017 14:06:44 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n\n> Hi\n>\n> 2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.\n> com>:\n>\n>> On 3/11/17 07:06, Pavel Stehule wrote:\n>> > I am sending a updated version with separated sort direction in special\n>> > variable\n>>\n>> This patch also needs a rebase.\n>>\n>\n> I am sending rebased patch\n>\n\nrebased again + fix obsolete help\n\nRegards\n\nPavel\n\n\n\n>\n> Regards\n>\n> Pavel\n>\n>\n>>\n>> --\n>> Peter Eisentraut http://www.2ndQuadrant.com/\n>> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>>\n>\n>\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers","msg_date":"Fri, 8 Sep 2017 06:13:47 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com>\nwrote:\n\n> 2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n>\n>> Hi\n>>\n>> 2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.\n>> com>:\n>>\n>>> On 3/11/17 07:06, Pavel Stehule wrote:\n>>> > I am sending a updated version with separated sort direction in special\n>>> > variable\n>>>\n>>> This patch also needs a rebase.\n>>>\n>>\n>> I am sending rebased patch\n>>\n>\n> rebased again + fix obsolete help\n>\n\nFor me, patch applies cleanly, builds and passed regression tests.\nHowever, patch misses regression tests covering added functionality.\nPatch is definitely harmless, i.e. it doesn't affect anybody who doesn't\nuse new functionality.\nBut I still would prefer ordering to be options of \\d* commands while psql\nvariables be defaults for those options...\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:Hi2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 3/11/17 07:06, Pavel Stehule wrote:\n> I am sending a updated version with separated sort direction in special\n> variable\n\nThis patch also needs a rebase.I am sending rebased patchrebased again + fix obsolete helpFor me, patch applies cleanly, builds and passed regression tests.However, patch misses regression tests covering added functionality.Patch is definitely harmless, i.e. it doesn't affect anybody who doesn't use new functionality.But I still would prefer ordering to be options of \\d* commands while psql variables be defaults for those options...------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Tue, 19 Sep 2017 17:14:20 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"Hi\n\n2017-09-19 16:14 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n\n> On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com>\n> wrote:\n>\n>> 2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n>>\n>>> Hi\n>>>\n>>> 2017-08-15 4:37 GMT+02:00 Peter Eisentraut <\n>>> peter.eisentraut@2ndquadrant.com>:\n>>>\n>>>> On 3/11/17 07:06, Pavel Stehule wrote:\n>>>> > I am sending a updated version with separated sort direction in\n>>>> special\n>>>> > variable\n>>>>\n>>>> This patch also needs a rebase.\n>>>>\n>>>\n>>> I am sending rebased patch\n>>>\n>>\n>> rebased again + fix obsolete help\n>>\n>\n> For me, patch applies cleanly, builds and passed regression tests.\n> However, patch misses regression tests covering added functionality.\n>\n\nI am not sure if there are any tests related to output of \\dt+ commands -\nthere result is platform depend.\n\n\n> Patch is definitely harmless, i.e. it doesn't affect anybody who doesn't\n> use new functionality.\n> But I still would prefer ordering to be options of \\d* commands while psql\n> variables be defaults for those options...\n>\n\nI understand\n\na) I don't think so commands like \\dt++ (or similar) is good idea - these\ncommands should be simple how it is possible\n\nb) this patch doesn't block any other design - more it opens the door\nbecause the executive part will be implemented and users can have a\nexperience with with different output sorts - so if people will need more\nquick change of result sort, then the work in this area will continue.\n\nRegards\n\nPavel\n\n\n\n\n\n\n\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n\nHi2017-09-19 16:14 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:Hi2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 3/11/17 07:06, Pavel Stehule wrote:\n> I am sending a updated version with separated sort direction in special\n> variable\n\nThis patch also needs a rebase.I am sending rebased patchrebased again + fix obsolete helpFor me, patch applies cleanly, builds and passed regression tests.However, patch misses regression tests covering added functionality.I am not sure if there are any tests related to output of \\dt+ commands - there result is platform depend. Patch is definitely harmless, i.e. it doesn't affect anybody who doesn't use new functionality.But I still would prefer ordering to be options of \\d* commands while psql variables be defaults for those options...I understand a) I don't think so commands like \\dt++ (or similar) is good idea - these commands should be simple how it is possibleb) this patch doesn't block any other design - more it opens the door because the executive part will be implemented and users can have a experience with with different output sorts - so if people will need more quick change of result sort, then the work in this area will continue.RegardsPavel------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Tue, 19 Sep 2017 18:54:03 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Tue, Sep 19, 2017 at 7:54 PM, Pavel Stehule <pavel.stehule@gmail.com>\nwrote:\n\n> 2017-09-19 16:14 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n>\n>> On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com>\n>> wrote:\n>>\n>>> 2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n>>>\n>>>> Hi\n>>>>\n>>>> 2017-08-15 4:37 GMT+02:00 Peter Eisentraut <\n>>>> peter.eisentraut@2ndquadrant.com>:\n>>>>\n>>>>> On 3/11/17 07:06, Pavel Stehule wrote:\n>>>>> > I am sending a updated version with separated sort direction in\n>>>>> special\n>>>>> > variable\n>>>>>\n>>>>> This patch also needs a rebase.\n>>>>>\n>>>>\n>>>> I am sending rebased patch\n>>>>\n>>>\n>>> rebased again + fix obsolete help\n>>>\n>>\n>> For me, patch applies cleanly, builds and passed regression tests.\n>> However, patch misses regression tests covering added functionality.\n>>\n>\n> I am not sure if there are any tests related to output of \\dt+ commands -\n> there result is platform depend.\n>\n\nBTW, why isn't order by name_schema available for \\dt? If it's available\nwe could at least cover this case by plain regression tests.\n\\dt+ could be covered by TAP tests, but it isn't yet. I think one day we\nshould add them. However, I don't think we should force you to write them\nin order to push this simple patch.\n\nPatch is definitely harmless, i.e. it doesn't affect anybody who doesn't\n>> use new functionality.\n>> But I still would prefer ordering to be options of \\d* commands while\n>> psql variables be defaults for those options...\n>>\n>\n> I understand\n>\n> a) I don't think so commands like \\dt++ (or similar) is good idea - these\n> commands should be simple how it is possible\n>\n\nI don't particularly like \\dt++, but second argument is probably an option.\n\n\n> b) this patch doesn't block any other design - more it opens the door\n> because the executive part will be implemented and users can have a\n> experience with with different output sorts - so if people will need more\n> quick change of result sort, then the work in this area will continue.\n>\n\nOK. As reviewer, I'm not going to block this patch if you see its\nfunctionality limited by just psql variables.\nI think you should add support of name_schema \\dt and some regression tests\nfor this case, before I can mark this as \"ready for committer\".\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Tue, Sep 19, 2017 at 7:54 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-09-19 16:14 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:Hi2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 3/11/17 07:06, Pavel Stehule wrote:\n> I am sending a updated version with separated sort direction in special\n> variable\n\nThis patch also needs a rebase.I am sending rebased patchrebased again + fix obsolete helpFor me, patch applies cleanly, builds and passed regression tests.However, patch misses regression tests covering added functionality.I am not sure if there are any tests related to output of \\dt+ commands - there result is platform depend. BTW, why isn't order by name_schema available for \\dt? If it's available we could at least cover this case by plain regression tests.\\dt+ could be covered by TAP tests, but it isn't yet. I think one day we should add them. However, I don't think we should force you to write them in order to push this simple patch.Patch is definitely harmless, i.e. it doesn't affect anybody who doesn't use new functionality.But I still would prefer ordering to be options of \\d* commands while psql variables be defaults for those options...I understand a) I don't think so commands like \\dt++ (or similar) is good idea - these commands should be simple how it is possibleI don't particularly like \\dt++, but second argument is probably an option. b) this patch doesn't block any other design - more it opens the door because the executive part will be implemented and users can have a experience with with different output sorts - so if people will need more quick change of result sort, then the work in this area will continue.OK. As reviewer, I'm not going to block this patch if you see its functionality limited by just psql variables.I think you should add support of name_schema \\dt and some regression tests for this case, before I can mark this as \"ready for committer\".------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Tue, 19 Sep 2017 23:03:04 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On 9/19/17 12:54, Pavel Stehule wrote:\n> However, patch misses regression tests covering added functionality.\n> \n> I am not sure if there are any tests related to output of \\dt+ commands\n> - there result is platform depend.\n\nHow so?\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Wed, 20 Sep 2017 18:52:40 -0400","msg_from":"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On 9/8/17 00:13, Pavel Stehule wrote:\n> I am sending rebased patch\n> \n> rebased again + fix obsolete help\n\nWhy are the variables called VERBOSE_SORT_* ? What is verbose about them?\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Wed, 20 Sep 2017 18:53:31 -0400","msg_from":"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Thu, Sep 21, 2017 at 1:52 AM, Peter Eisentraut <\npeter.eisentraut@2ndquadrant.com> wrote:\n\n> On 9/19/17 12:54, Pavel Stehule wrote:\n> > However, patch misses regression tests covering added functionality.\n> >\n> > I am not sure if there are any tests related to output of \\dt+ commands\n> > - there result is platform depend.\n>\n> How so?\n\n\n\\dt+ reports relation sizes whose are platform depended.\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Thu, Sep 21, 2017 at 1:52 AM, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:On 9/19/17 12:54, Pavel Stehule wrote:\n> However, patch misses regression tests covering added functionality.\n>\n> I am not sure if there are any tests related to output of \\dt+ commands\n> - there result is platform depend.\n\nHow so?\\dt+ reports relation sizes whose are platform depended.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Thu, 21 Sep 2017 11:16:52 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Thu, Sep 21, 2017 at 1:53 AM, Peter Eisentraut <\npeter.eisentraut@2ndquadrant.com> wrote:\n\n> On 9/8/17 00:13, Pavel Stehule wrote:\n> > I am sending rebased patch\n> >\n> > rebased again + fix obsolete help\n>\n> Why are the variables called VERBOSE_SORT_* ? What is verbose about them?\n\n\nI assume Pavel called them so, because they are working only for \"verbose\"\nmode of command. I.e. they are working for \\dt+ not \\dt.\nHowever, in \\dt 2 of 3 sorting modes might work: schema_name and\nname_schema. Thus, I think it worths enabling these variables for \"non\nverbose\" mode of commands too.\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Thu, Sep 21, 2017 at 1:53 AM, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:On 9/8/17 00:13, Pavel Stehule wrote:\n> I am sending rebased patch\n>\n> rebased again + fix obsolete help\n\nWhy are the variables called VERBOSE_SORT_* ? What is verbose about them?I assume Pavel called them so, because they are working only for \"verbose\" mode of command. I.e. they are working for \\dt+ not \\dt.However, in \\dt 2 of 3 sorting modes might work: schema_name and name_schema. Thus, I think it worths enabling these variables for \"non verbose\" mode of commands too.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Thu, 21 Sep 2017 11:19:30 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-09-21 10:19 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n\n> On Thu, Sep 21, 2017 at 1:53 AM, Peter Eisentraut <\n> peter.eisentraut@2ndquadrant.com> wrote:\n>\n>> On 9/8/17 00:13, Pavel Stehule wrote:\n>> > I am sending rebased patch\n>> >\n>> > rebased again + fix obsolete help\n>>\n>> Why are the variables called VERBOSE_SORT_* ? What is verbose about them?\n>\n>\n> I assume Pavel called them so, because they are working only for \"verbose\"\n> mode of command. I.e. they are working for \\dt+ not \\dt.\n> However, in \\dt 2 of 3 sorting modes might work: schema_name and\n> name_schema. Thus, I think it worths enabling these variables for \"non\n> verbose\" mode of commands too.\n>\n\nyes. It was designed for + commands only. Can be enhanced to all commands -\nthen VERBOSE prefix should be removed - not sure if it is necessary. For me\ninteresting different order than default is only in verbose mode.\n\n\n\n\n\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n\n2017-09-21 10:19 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Thu, Sep 21, 2017 at 1:53 AM, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:On 9/8/17 00:13, Pavel Stehule wrote:\n> I am sending rebased patch\n>\n> rebased again + fix obsolete help\n\nWhy are the variables called VERBOSE_SORT_* ? What is verbose about them?I assume Pavel called them so, because they are working only for \"verbose\" mode of command. I.e. they are working for \\dt+ not \\dt.However, in \\dt 2 of 3 sorting modes might work: schema_name and name_schema. Thus, I think it worths enabling these variables for \"non verbose\" mode of commands too.yes. It was designed for + commands only. Can be enhanced to all commands - then VERBOSE prefix should be removed - not sure if it is necessary. For me interesting different order than default is only in verbose mode. ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Thu, 21 Sep 2017 10:27:32 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On 9/21/17 04:27, Pavel Stehule wrote:\n> yes. It was designed for + commands only. Can be enhanced to all\n> commands - then VERBOSE prefix should be removed - not sure if it is\n> necessary. For me interesting different order than default is only in\n> verbose mode.\n\nI see where you are coming from, but there is no association in the\nexisting UI that equates \"+\" to the word \"verbose\". I think just\nremoving the verbose prefix and applying the sorting behavior in all\ncases should be easier to explain and implement.\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Thu, 21 Sep 2017 09:30:13 -0400","msg_from":"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-09-21 15:30 GMT+02:00 Peter Eisentraut <\npeter.eisentraut@2ndquadrant.com>:\n\n> On 9/21/17 04:27, Pavel Stehule wrote:\n> > yes. It was designed for + commands only. Can be enhanced to all\n> > commands - then VERBOSE prefix should be removed - not sure if it is\n> > necessary. For me interesting different order than default is only in\n> > verbose mode.\n>\n> I see where you are coming from, but there is no association in the\n> existing UI that equates \"+\" to the word \"verbose\". I think just\n> removing the verbose prefix and applying the sorting behavior in all\n> cases should be easier to explain and implement.\n>\n\nI though about it - but I am not sure if one kind of these variables is\npractical.\n\nif I don't need a size, then sort by schema, name is ok (I didn't need any\nelse ever). With only one kind of these variables, this setting is common -\nwhat is not practical.\n\nI need sort by size in verbose mode (where size is visible) in 100% - so it\nwill be saved to psqlrc. And when size will be invisible, then sort by size\nis not practical, and can be messy (because size is not visible).\n\nSo I don't think so removing VERBOSE prefix is a good idea - or we should\nto do different design (have not a idea how)\n\nRegards\n\nPavel\n\n\n> --\n> Peter Eisentraut http://www.2ndQuadrant.com/\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>\n\n2017-09-21 15:30 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/21/17 04:27, Pavel Stehule wrote:\n> yes. It was designed for + commands only. Can be enhanced to all\n> commands - then VERBOSE prefix should be removed - not sure if it is\n> necessary. For me interesting different order than default is only in\n> verbose mode.\n\nI see where you are coming from, but there is no association in the\nexisting UI that equates \"+\" to the word \"verbose\". I think just\nremoving the verbose prefix and applying the sorting behavior in all\ncases should be easier to explain and implement.I though about it - but I am not sure if one kind of these variables is practical.if I don't need a size, then sort by schema, name is ok (I didn't need any else ever). With only one kind of these variables, this setting is common - what is not practical.I need sort by size in verbose mode (where size is visible) in 100% - so it will be saved to psqlrc. And when size will be invisible, then sort by size is not practical, and can be messy (because size is not visible). So I don't think so removing VERBOSE prefix is a good idea - or we should to do different design (have not a idea how)RegardsPavel\n\n--\nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services","msg_date":"Thu, 21 Sep 2017 19:54:01 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On 9/21/17 13:54, Pavel Stehule wrote:\n> I see where you are coming from, but there is no association in the\n> existing UI that equates \"+\" to the word \"verbose\". I think just\n> removing the verbose prefix and applying the sorting behavior in all\n> cases should be easier to explain and implement.\n> \n> I though about it - but I am not sure if one kind of these variables is\n> practical.\n> \n> if I don't need a size, then sort by schema, name is ok (I didn't need\n> any else ever). With only one kind of these variables, this setting is\n> common - what is not practical.\n\nBut you are proposing also to add a variable configuring the sort\ndirection. It would be weird that \\dX+ observed the sort direction but\n\\dX did not.\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Thu, 21 Sep 2017 14:20:56 -0400","msg_from":"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-09-21 20:20 GMT+02:00 Peter Eisentraut <\npeter.eisentraut@2ndquadrant.com>:\n\n> On 9/21/17 13:54, Pavel Stehule wrote:\n> > I see where you are coming from, but there is no association in the\n> > existing UI that equates \"+\" to the word \"verbose\". I think just\n> > removing the verbose prefix and applying the sorting behavior in all\n> > cases should be easier to explain and implement.\n> >\n> > I though about it - but I am not sure if one kind of these variables is\n> > practical.\n> >\n> > if I don't need a size, then sort by schema, name is ok (I didn't need\n> > any else ever). With only one kind of these variables, this setting is\n> > common - what is not practical.\n>\n> But you are proposing also to add a variable configuring the sort\n> direction. It would be weird that \\dX+ observed the sort direction but\n> \\dX did not.\n>\n\nyes and no.\n\nschema_name, name_schema or SORT_DIRECTION has sense for both type of\ncommands.\n\nsize sort has sense only for \\dX+ command.\n\nI am thinking about solution and the most clean I see two distinct\nvariables:\n\nSORT_COLUMNS and VERBOSE_SORT_COLUMNS\n\nwhen VERBOSE_SORT_COLUMNS will be undefined, then SORT_COLUMNS is used for\n\\dX+ command too.\n\nIs it acceptable?\n\n\n\n\n> --\n> Peter Eisentraut http://www.2ndQuadrant.com/\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>\n\n2017-09-21 20:20 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/21/17 13:54, Pavel Stehule wrote:\n> I see where you are coming from, but there is no association in the\n> existing UI that equates \"+\" to the word \"verbose\". I think just\n> removing the verbose prefix and applying the sorting behavior in all\n> cases should be easier to explain and implement.\n>\n> I though about it - but I am not sure if one kind of these variables is\n> practical.\n>\n> if I don't need a size, then sort by schema, name is ok (I didn't need\n> any else ever). With only one kind of these variables, this setting is\n> common - what is not practical.\n\nBut you are proposing also to add a variable configuring the sort\ndirection. It would be weird that \\dX+ observed the sort direction but\n\\dX did not.yes and no. schema_name, name_schema or SORT_DIRECTION has sense for both type of commands.size sort has sense only for \\dX+ command.I am thinking about solution and the most clean I see two distinct variables:SORT_COLUMNS and VERBOSE_SORT_COLUMNS when VERBOSE_SORT_COLUMNS will be undefined, then SORT_COLUMNS is used for \\dX+ command too.Is it acceptable?\n\n--\nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services","msg_date":"Thu, 21 Sep 2017 20:30:19 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-09-21 20:30 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n\n>\n>\n> 2017-09-21 20:20 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.\n> com>:\n>\n>> On 9/21/17 13:54, Pavel Stehule wrote:\n>> > I see where you are coming from, but there is no association in the\n>> > existing UI that equates \"+\" to the word \"verbose\". I think just\n>> > removing the verbose prefix and applying the sorting behavior in all\n>> > cases should be easier to explain and implement.\n>> >\n>> > I though about it - but I am not sure if one kind of these variables is\n>> > practical.\n>> >\n>> > if I don't need a size, then sort by schema, name is ok (I didn't need\n>> > any else ever). With only one kind of these variables, this setting is\n>> > common - what is not practical.\n>>\n>> But you are proposing also to add a variable configuring the sort\n>> direction. It would be weird that \\dX+ observed the sort direction but\n>> \\dX did not.\n>>\n>\n> yes and no.\n>\n> schema_name, name_schema or SORT_DIRECTION has sense for both type of\n> commands.\n>\n> size sort has sense only for \\dX+ command.\n>\n> I am thinking about solution and the most clean I see two distinct\n> variables:\n>\n> SORT_COLUMNS and VERBOSE_SORT_COLUMNS\n>\n> when VERBOSE_SORT_COLUMNS will be undefined, then SORT_COLUMNS is used for\n> \\dX+ command too.\n>\n> Is it acceptable?\n>\n\nI though more about it, and I am thinking so this direction is not good.\n\nExample: somebody set SORT_COLUMNS to schema_name value. This is nonsense\nfor \\l command\n\nNow, I am thinking so more correct and practical design is based on special\nmode, activated by variable\n\nPREFER_SIZE_SORT .. (off, asc, desc)\n\nThis has sense for wide group of commands that can show size. And when size\nis not visible, then this option is not active.\n\nWhat do you think about this proposal?\n\nRegards\n\nPavel\n\n\n\n>\n>\n>\n>> --\n>> Peter Eisentraut http://www.2ndQuadrant.com/\n>> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>>\n>\n>\n\n2017-09-21 20:30 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:2017-09-21 20:20 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/21/17 13:54, Pavel Stehule wrote:\n> I see where you are coming from, but there is no association in the\n> existing UI that equates \"+\" to the word \"verbose\". I think just\n> removing the verbose prefix and applying the sorting behavior in all\n> cases should be easier to explain and implement.\n>\n> I though about it - but I am not sure if one kind of these variables is\n> practical.\n>\n> if I don't need a size, then sort by schema, name is ok (I didn't need\n> any else ever). With only one kind of these variables, this setting is\n> common - what is not practical.\n\nBut you are proposing also to add a variable configuring the sort\ndirection. It would be weird that \\dX+ observed the sort direction but\n\\dX did not.yes and no. schema_name, name_schema or SORT_DIRECTION has sense for both type of commands.size sort has sense only for \\dX+ command.I am thinking about solution and the most clean I see two distinct variables:SORT_COLUMNS and VERBOSE_SORT_COLUMNS when VERBOSE_SORT_COLUMNS will be undefined, then SORT_COLUMNS is used for \\dX+ command too.Is it acceptable?I though more about it, and I am thinking so this direction is not good. Example: somebody set SORT_COLUMNS to schema_name value. This is nonsense for \\l commandNow, I am thinking so more correct and practical design is based on special mode, activated by variablePREFER_SIZE_SORT .. (off, asc, desc)This has sense for wide group of commands that can show size. And when size is not visible, then this option is not active.What do you think about this proposal?RegardsPavel \n\n--\nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services","msg_date":"Fri, 22 Sep 2017 15:16:35 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On 9/22/17 09:16, Pavel Stehule wrote:\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\n> nonsense for \\l command\n> \n> Now, I am thinking so more correct and practical design is based on\n> special mode, activated by variable\n> \n> PREFER_SIZE_SORT .. (off, asc, desc)\n> \n> This has sense for wide group of commands that can show size. And when\n> size is not visible, then this option is not active.\n\nMaybe this shouldn't be a variable at all. It's not like you'll set\nthis as a global preference. You probably want it for one command only.\n So a per-command option might make more sense.\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Fri, 22 Sep 2017 15:12:29 -0400","msg_from":"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-09-22 21:12 GMT+02:00 Peter Eisentraut <\npeter.eisentraut@2ndquadrant.com>:\n\n> On 9/22/17 09:16, Pavel Stehule wrote:\n> > Example: somebody set SORT_COLUMNS to schema_name value. This is\n> > nonsense for \\l command\n> >\n> > Now, I am thinking so more correct and practical design is based on\n> > special mode, activated by variable\n> >\n> > PREFER_SIZE_SORT .. (off, asc, desc)\n> >\n> > This has sense for wide group of commands that can show size. And when\n> > size is not visible, then this option is not active.\n>\n> Maybe this shouldn't be a variable at all. It's not like you'll set\n> this as a global preference. You probably want it for one command only.\n> So a per-command option might make more sense.\n>\n\nSure, I cannot to know, what users will do. But, when I need to see a size\nof objects, then I prefer the sort by size desc every time. If I need to\nfind some object, then I can to use a searching in pager. So in my case,\nthis settings will be in psqlrc. In GoodData we used years own\ncustomization - the order by size was hardcoded and nobody reported me any\nissue.\n\nAlexander proposed some per command option, but current syntax of psql\ncommands don't allows some simple parametrization. If it can be user\nfriendly, then it should be short. From implementation perspective, it\nshould be simply parsed. It should be intuitive too - too much symbols\ntogether is not good idea.\n\nMaybe some prefix design - but it is not design for common people (although\nthese people don't use psql usually)\n\n'\\sort size \\dt ?\n\n\\dt:sort_by_size\n\\dt+:sort_by_size ?\n\nI don't see any good design in this direction\n\nRegards\n\nPavel\n\n\n\n\n\n\n\n>\n> --\n> Peter Eisentraut http://www.2ndQuadrant.com/\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>\n\n2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/22/17 09:16, Pavel Stehule wrote:\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\n> nonsense for \\l command\n>\n> Now, I am thinking so more correct and practical design is based on\n> special mode, activated by variable\n>\n> PREFER_SIZE_SORT .. (off, asc, desc)\n>\n> This has sense for wide group of commands that can show size. And when\n> size is not visible, then this option is not active.\n\nMaybe this shouldn't be a variable at all. It's not like you'll set\nthis as a global preference. You probably want it for one command only.\n So a per-command option might make more sense.Sure, I cannot to know, what users will do. But, when I need to see a size of objects, then I prefer the sort by size desc every time. If I need to find some object, then I can to use a searching in pager. So in my case, this settings will be in psqlrc. In GoodData we used years own customization - the order by size was hardcoded and nobody reported me any issue. Alexander proposed some per command option, but current syntax of psql commands don't allows some simple parametrization. If it can be user friendly, then it should be short. From implementation perspective, it should be simply parsed. It should be intuitive too - too much symbols together is not good idea.Maybe some prefix design - but it is not design for common people (although these people don't use psql usually)'\\sort size \\dt ?\\dt:sort_by_size \\dt+:sort_by_size ?I don't see any good design in this direction RegardsPavel \n\n--\nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services","msg_date":"Fri, 22 Sep 2017 21:31:12 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"Hi\n\n2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n\n>\n>\n> 2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.\n> com>:\n>\n>> On 9/22/17 09:16, Pavel Stehule wrote:\n>> > Example: somebody set SORT_COLUMNS to schema_name value. This is\n>> > nonsense for \\l command\n>> >\n>> > Now, I am thinking so more correct and practical design is based on\n>> > special mode, activated by variable\n>> >\n>> > PREFER_SIZE_SORT .. (off, asc, desc)\n>> >\n>> > This has sense for wide group of commands that can show size. And when\n>> > size is not visible, then this option is not active.\n>>\n>> Maybe this shouldn't be a variable at all. It's not like you'll set\n>> this as a global preference. You probably want it for one command only.\n>> So a per-command option might make more sense.\n>>\n>\n> Sure, I cannot to know, what users will do. But, when I need to see a size\n> of objects, then I prefer the sort by size desc every time. If I need to\n> find some object, then I can to use a searching in pager. So in my case,\n> this settings will be in psqlrc. In GoodData we used years own\n> customization - the order by size was hardcoded and nobody reported me any\n> issue.\n>\n> Alexander proposed some per command option, but current syntax of psql\n> commands don't allows some simple parametrization. If it can be user\n> friendly, then it should be short. From implementation perspective, it\n> should be simply parsed. It should be intuitive too - too much symbols\n> together is not good idea.\n>\n> Maybe some prefix design - but it is not design for common people\n> (although these people don't use psql usually)\n>\n> '\\sort size \\dt ?\n>\n> \\dt:sort_by_size\n> \\dt+:sort_by_size ?\n>\n> I don't see any good design in this direction\n>\n>\nI though about Alexander proposal, and I am thinking so it can be probably\nbest if we respect psql design. I implemented two command suffixes\n(supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n\nso list of tables can be sorted with commands:\n\n\\dt+sd (in this case, the order is not strict), so command\n\\dtsd+ is working too (same \\disd+ or \\di+sd)\n\nThese two chars are acceptable. Same principle is used for \\l command\n\n\\lsd+ or \\l+sd\n\nWhat do you think about it?\n\nRegards\n\nPavel\n\n\n\n> Regards\n>\n> Pavel\n>\n>\n>\n>\n>\n>\n>\n>>\n>> --\n>> Peter Eisentraut http://www.2ndQuadrant.com/\n>> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n>>\n>\n>\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers","msg_date":"Sat, 28 Oct 2017 14:46:46 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com>\nwrote:\n\n> 2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n>\n>>\n>>\n>> 2017-09-22 21:12 GMT+02:00 Peter Eisentraut <\n>> peter.eisentraut@2ndquadrant.com>:\n>>\n>>> On 9/22/17 09:16, Pavel Stehule wrote:\n>>> > Example: somebody set SORT_COLUMNS to schema_name value. This is\n>>> > nonsense for \\l command\n>>> >\n>>> > Now, I am thinking so more correct and practical design is based on\n>>> > special mode, activated by variable\n>>> >\n>>> > PREFER_SIZE_SORT .. (off, asc, desc)\n>>> >\n>>> > This has sense for wide group of commands that can show size. And when\n>>> > size is not visible, then this option is not active.\n>>>\n>>> Maybe this shouldn't be a variable at all. It's not like you'll set\n>>> this as a global preference. You probably want it for one command only.\n>>> So a per-command option might make more sense.\n>>>\n>>\n>> Sure, I cannot to know, what users will do. But, when I need to see a\n>> size of objects, then I prefer the sort by size desc every time. If I need\n>> to find some object, then I can to use a searching in pager. So in my case,\n>> this settings will be in psqlrc. In GoodData we used years own\n>> customization - the order by size was hardcoded and nobody reported me any\n>> issue.\n>>\n>> Alexander proposed some per command option, but current syntax of psql\n>> commands don't allows some simple parametrization. If it can be user\n>> friendly, then it should be short. From implementation perspective, it\n>> should be simply parsed. It should be intuitive too - too much symbols\n>> together is not good idea.\n>>\n>> Maybe some prefix design - but it is not design for common people\n>> (although these people don't use psql usually)\n>>\n>> '\\sort size \\dt ?\n>>\n>> \\dt:sort_by_size\n>> \\dt+:sort_by_size ?\n>>\n>> I don't see any good design in this direction\n>>\n>>\n> I though about Alexander proposal, and I am thinking so it can be probably\n> best if we respect psql design. I implemented two command suffixes\n> (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n>\n> so list of tables can be sorted with commands:\n>\n> \\dt+sd (in this case, the order is not strict), so command\n> \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>\n> These two chars are acceptable. Same principle is used for \\l command\n>\n> \\lsd+ or \\l+sd\n>\n> What do you think about it?\n>\n\nI think \\lsd+ command would be another postgres meme :)\nBTW, are you going to provide an ability to sort by name, schema?\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/22/17 09:16, Pavel Stehule wrote:\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\n> nonsense for \\l command\n>\n> Now, I am thinking so more correct and practical design is based on\n> special mode, activated by variable\n>\n> PREFER_SIZE_SORT .. (off, asc, desc)\n>\n> This has sense for wide group of commands that can show size. And when\n> size is not visible, then this option is not active.\n\nMaybe this shouldn't be a variable at all. It's not like you'll set\nthis as a global preference. You probably want it for one command only.\n So a per-command option might make more sense.Sure, I cannot to know, what users will do. But, when I need to see a size of objects, then I prefer the sort by size desc every time. If I need to find some object, then I can to use a searching in pager. So in my case, this settings will be in psqlrc. In GoodData we used years own customization - the order by size was hardcoded and nobody reported me any issue. Alexander proposed some per command option, but current syntax of psql commands don't allows some simple parametrization. If it can be user friendly, then it should be short. From implementation perspective, it should be simply parsed. It should be intuitive too - too much symbols together is not good idea.Maybe some prefix design - but it is not design for common people (although these people don't use psql usually)'\\sort size \\dt ?\\dt:sort_by_size \\dt+:sort_by_size ?I don't see any good design in this direction I though about Alexander proposal, and I am thinking so it can be probably best if we respect psql design. I implemented two command suffixes (supported only when it has sense) \"s\" sorted by size and \"d\" as descentso list of tables can be sorted with commands:\\dt+sd (in this case, the order is not strict), so command\\dtsd+ is working too (same \\disd+ or \\di+sd)These two chars are acceptable. Same principle is used for \\l command\\lsd+ or \\l+sdWhat do you think about it?I think \\lsd+ command would be another postgres meme :)BTW, are you going to provide an ability to sort by name, schema?------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Sun, 29 Oct 2017 00:35:56 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-10-28 23:35 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n\n> On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com>\n> wrote:\n>\n>> 2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n>>\n>>>\n>>>\n>>> 2017-09-22 21:12 GMT+02:00 Peter Eisentraut <\n>>> peter.eisentraut@2ndquadrant.com>:\n>>>\n>>>> On 9/22/17 09:16, Pavel Stehule wrote:\n>>>> > Example: somebody set SORT_COLUMNS to schema_name value. This is\n>>>> > nonsense for \\l command\n>>>> >\n>>>> > Now, I am thinking so more correct and practical design is based on\n>>>> > special mode, activated by variable\n>>>> >\n>>>> > PREFER_SIZE_SORT .. (off, asc, desc)\n>>>> >\n>>>> > This has sense for wide group of commands that can show size. And when\n>>>> > size is not visible, then this option is not active.\n>>>>\n>>>> Maybe this shouldn't be a variable at all. It's not like you'll set\n>>>> this as a global preference. You probably want it for one command only.\n>>>> So a per-command option might make more sense.\n>>>>\n>>>\n>>> Sure, I cannot to know, what users will do. But, when I need to see a\n>>> size of objects, then I prefer the sort by size desc every time. If I need\n>>> to find some object, then I can to use a searching in pager. So in my case,\n>>> this settings will be in psqlrc. In GoodData we used years own\n>>> customization - the order by size was hardcoded and nobody reported me any\n>>> issue.\n>>>\n>>> Alexander proposed some per command option, but current syntax of psql\n>>> commands don't allows some simple parametrization. If it can be user\n>>> friendly, then it should be short. From implementation perspective, it\n>>> should be simply parsed. It should be intuitive too - too much symbols\n>>> together is not good idea.\n>>>\n>>> Maybe some prefix design - but it is not design for common people\n>>> (although these people don't use psql usually)\n>>>\n>>> '\\sort size \\dt ?\n>>>\n>>> \\dt:sort_by_size\n>>> \\dt+:sort_by_size ?\n>>>\n>>> I don't see any good design in this direction\n>>>\n>>>\n>> I though about Alexander proposal, and I am thinking so it can be\n>> probably best if we respect psql design. I implemented two command suffixes\n>> (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n>>\n>> so list of tables can be sorted with commands:\n>>\n>> \\dt+sd (in this case, the order is not strict), so command\n>> \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>>\n>> These two chars are acceptable. Same principle is used for \\l command\n>>\n>> \\lsd+ or \\l+sd\n>>\n>> What do you think about it?\n>>\n>\n> I think \\lsd+ command would be another postgres meme :)\n> BTW, are you going to provide an ability to sort by name, schema?\n>\n\nIt has sense only for tables - probably only \\dtn \"n\" like name\n\n\n\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n\n2017-10-28 23:35 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/22/17 09:16, Pavel Stehule wrote:\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\n> nonsense for \\l command\n>\n> Now, I am thinking so more correct and practical design is based on\n> special mode, activated by variable\n>\n> PREFER_SIZE_SORT .. (off, asc, desc)\n>\n> This has sense for wide group of commands that can show size. And when\n> size is not visible, then this option is not active.\n\nMaybe this shouldn't be a variable at all. It's not like you'll set\nthis as a global preference. You probably want it for one command only.\n So a per-command option might make more sense.Sure, I cannot to know, what users will do. But, when I need to see a size of objects, then I prefer the sort by size desc every time. If I need to find some object, then I can to use a searching in pager. So in my case, this settings will be in psqlrc. In GoodData we used years own customization - the order by size was hardcoded and nobody reported me any issue. Alexander proposed some per command option, but current syntax of psql commands don't allows some simple parametrization. If it can be user friendly, then it should be short. From implementation perspective, it should be simply parsed. It should be intuitive too - too much symbols together is not good idea.Maybe some prefix design - but it is not design for common people (although these people don't use psql usually)'\\sort size \\dt ?\\dt:sort_by_size \\dt+:sort_by_size ?I don't see any good design in this direction I though about Alexander proposal, and I am thinking so it can be probably best if we respect psql design. I implemented two command suffixes (supported only when it has sense) \"s\" sorted by size and \"d\" as descentso list of tables can be sorted with commands:\\dt+sd (in this case, the order is not strict), so command\\dtsd+ is working too (same \\disd+ or \\di+sd)These two chars are acceptable. Same principle is used for \\l command\\lsd+ or \\l+sdWhat do you think about it?I think \\lsd+ command would be another postgres meme :)BTW, are you going to provide an ability to sort by name, schema?It has sense only for tables - probably only \\dtn \"n\" like name ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Sat, 28 Oct 2017 23:47:27 +0200","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Sun, Oct 29, 2017 at 12:47 AM, Pavel Stehule <pavel.stehule@gmail.com>\nwrote:\n\n> 2017-10-28 23:35 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n>\n>> On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com>\n>> wrote:\n>>\n>>> 2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\n>>>\n>>>>\n>>>>\n>>>> 2017-09-22 21:12 GMT+02:00 Peter Eisentraut <\n>>>> peter.eisentraut@2ndquadrant.com>:\n>>>>\n>>>>> On 9/22/17 09:16, Pavel Stehule wrote:\n>>>>> > Example: somebody set SORT_COLUMNS to schema_name value. This is\n>>>>> > nonsense for \\l command\n>>>>> >\n>>>>> > Now, I am thinking so more correct and practical design is based on\n>>>>> > special mode, activated by variable\n>>>>> >\n>>>>> > PREFER_SIZE_SORT .. (off, asc, desc)\n>>>>> >\n>>>>> > This has sense for wide group of commands that can show size. And\n>>>>> when\n>>>>> > size is not visible, then this option is not active.\n>>>>>\n>>>>> Maybe this shouldn't be a variable at all. It's not like you'll set\n>>>>> this as a global preference. You probably want it for one command\n>>>>> only.\n>>>>> So a per-command option might make more sense.\n>>>>>\n>>>>\n>>>> Sure, I cannot to know, what users will do. But, when I need to see a\n>>>> size of objects, then I prefer the sort by size desc every time. If I need\n>>>> to find some object, then I can to use a searching in pager. So in my case,\n>>>> this settings will be in psqlrc. In GoodData we used years own\n>>>> customization - the order by size was hardcoded and nobody reported me any\n>>>> issue.\n>>>>\n>>>> Alexander proposed some per command option, but current syntax of psql\n>>>> commands don't allows some simple parametrization. If it can be user\n>>>> friendly, then it should be short. From implementation perspective, it\n>>>> should be simply parsed. It should be intuitive too - too much symbols\n>>>> together is not good idea.\n>>>>\n>>>> Maybe some prefix design - but it is not design for common people\n>>>> (although these people don't use psql usually)\n>>>>\n>>>> '\\sort size \\dt ?\n>>>>\n>>>> \\dt:sort_by_size\n>>>> \\dt+:sort_by_size ?\n>>>>\n>>>> I don't see any good design in this direction\n>>>>\n>>>>\n>>> I though about Alexander proposal, and I am thinking so it can be\n>>> probably best if we respect psql design. I implemented two command suffixes\n>>> (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n>>>\n>>> so list of tables can be sorted with commands:\n>>>\n>>> \\dt+sd (in this case, the order is not strict), so command\n>>> \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>>>\n>>> These two chars are acceptable. Same principle is used for \\l command\n>>>\n>>> \\lsd+ or \\l+sd\n>>>\n>>> What do you think about it?\n>>>\n>>\n>> I think \\lsd+ command would be another postgres meme :)\n>> BTW, are you going to provide an ability to sort by name, schema?\n>>\n>\n> It has sense only for tables - probably only \\dtn \"n\" like name\n>\n\nIn general, this approach looks good for me.\nRegarding current state of patch, I'd like to see new options documented.\nAlso, it would be better to replace \"bool sort_size\" with enum assuming\nthere could be other sorting orders in future.\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Sun, Oct 29, 2017 at 12:47 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-10-28 23:35 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/22/17 09:16, Pavel Stehule wrote:\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\n> nonsense for \\l command\n>\n> Now, I am thinking so more correct and practical design is based on\n> special mode, activated by variable\n>\n> PREFER_SIZE_SORT .. (off, asc, desc)\n>\n> This has sense for wide group of commands that can show size. And when\n> size is not visible, then this option is not active.\n\nMaybe this shouldn't be a variable at all. It's not like you'll set\nthis as a global preference. You probably want it for one command only.\n So a per-command option might make more sense.Sure, I cannot to know, what users will do. But, when I need to see a size of objects, then I prefer the sort by size desc every time. If I need to find some object, then I can to use a searching in pager. So in my case, this settings will be in psqlrc. In GoodData we used years own customization - the order by size was hardcoded and nobody reported me any issue. Alexander proposed some per command option, but current syntax of psql commands don't allows some simple parametrization. If it can be user friendly, then it should be short. From implementation perspective, it should be simply parsed. It should be intuitive too - too much symbols together is not good idea.Maybe some prefix design - but it is not design for common people (although these people don't use psql usually)'\\sort size \\dt ?\\dt:sort_by_size \\dt+:sort_by_size ?I don't see any good design in this direction I though about Alexander proposal, and I am thinking so it can be probably best if we respect psql design. I implemented two command suffixes (supported only when it has sense) \"s\" sorted by size and \"d\" as descentso list of tables can be sorted with commands:\\dt+sd (in this case, the order is not strict), so command\\dtsd+ is working too (same \\disd+ or \\di+sd)These two chars are acceptable. Same principle is used for \\l command\\lsd+ or \\l+sdWhat do you think about it?I think \\lsd+ command would be another postgres meme :)BTW, are you going to provide an ability to sort by name, schema?It has sense only for tables - probably only \\dtn \"n\" like nameIn general, this approach looks good for me.Regarding current state of patch, I'd like to see new options documented. Also, it would be better to replace \"bool sort_size\" with enum assuming there could be other sorting orders in future.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Tue, 31 Oct 2017 11:26:52 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"Hi\n\n\nIn general, this approach looks good for me.\n> Regarding current state of patch, I'd like to see new options documented.\n> Also, it would be better to replace \"bool sort_size\" with enum assuming\n> there could be other sorting orders in future.\n>\n\nI am sending updated patch with some basic doc\n\nRegards\n\nPavel\n\n>\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers","msg_date":"Wed, 1 Nov 2017 12:44:31 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com> wrote:\n> I though about Alexander proposal, and I am thinking so it can be probably\n> best if we respect psql design. I implemented two command suffixes\n> (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n>\n> so list of tables can be sorted with commands:\n>\n> \\dt+sd (in this case, the order is not strict), so command\n> \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>\n> These two chars are acceptable. Same principle is used for \\l command\n>\n> \\lsd+ or \\l+sd\n>\n> What do you think about it?\n>\n\nI really hate that syntax. This is going to turn into an\nincomprehensible mess, and isn't easily extended to support other\noptions.\n\nI agree with people who have said they would prefer this to be\navailable as a per-command option rather than as a variable that you\nhave to set, but it needs a clearer syntax. I actually like Stephen's\nidea of using a user-defined SQL snippet, because that's a familiar\nsyntax to people, and it avoids adding an ever-increasing number of\noptions to these commands. Instead, the syntax could simply be:\n\n\\d[S+] [ pattern ] [ ( auxiliary sql ) ]\n\n(and similar for the other commands)\n\nThe auxiliary SQL could be pretty much anything to allow user-defined\nordering and filtering.\n\nI think parsing the optional auxiliary SQL snippet in parentheses at\nthe end should be quite straightforward, provided that psql makes no\nattempt to actually parse the SQL contained in the parentheses -- it\nshould just add it to the SQL it sends to the backend (like \\copy\ndoes). For example, for \\d+, instead of generating\n\nSELECT n.nspname as \"Schema\",\n c.relname as \"Name\",\n CASE c.relkind WHEN 'r' THEN 'table' WHEN 'v' THEN 'view' WHEN 'm'\nTHEN 'materialized view' WHEN 'i' THEN 'index' WHEN 'S' THEN\n'sequence' WHEN 's' THEN 'special' WHEN 'f' THEN 'foreign table' WHEN\n'p' THEN 'table' END as \"Type\",\n pg_catalog.pg_get_userbyid(c.relowner) as \"Owner\",\n pg_catalog.pg_size_pretty(pg_catalog.pg_table_size(c.oid)) as \"Size\",\n pg_catalog.obj_description(c.oid, 'pg_class') as \"Description\"\nFROM pg_catalog.pg_class c\n LEFT JOIN pg_catalog.pg_namespace n ON n.oid = c.relnamespace\nWHERE c.relkind IN ('r','p','v','m','S','f','')\n AND n.nspname <> 'pg_catalog'\n AND n.nspname <> 'information_schema'\n AND n.nspname !~ '^pg_toast'\n AND pg_catalog.pg_table_is_visible(c.oid)\nORDER BY 1,2;\n\nwe could generate\n\nSELECT schema as \"Schema\",\n name as \"Name\",\n type as \"Type\",\n owner as \"Owner\",\n pg_catalog.pg_size_pretty(size) as \"Size\",\n description as \"Description\"\nFROM (\nSELECT n.nspname as schema,\n c.relname as name,\n CASE c.relkind WHEN 'r' THEN 'table' WHEN 'v' THEN 'view' WHEN 'm'\nTHEN 'materialized view' WHEN 'i' THEN 'index' WHEN 'S' THEN\n'sequence' WHEN 's' THEN 'special' WHEN 'f' THEN 'foreign table' WHEN\n'p' THEN 'table' END as type,\n pg_catalog.pg_get_userbyid(c.relowner) as owner,\n pg_catalog.pg_table_size(c.oid) as size,\n pg_catalog.obj_description(c.oid, 'pg_class') as description\nFROM pg_catalog.pg_class c\n LEFT JOIN pg_catalog.pg_namespace n ON n.oid = c.relnamespace\nWHERE c.relkind IN ('r','p','v','m','S','f','')\n AND n.nspname <> 'pg_catalog'\n AND n.nspname <> 'information_schema'\n AND n.nspname !~ '^pg_toast'\n AND pg_catalog.pg_table_is_visible(c.oid)\nORDER BY 1,2\n) as t\n<auxiliary sql>;\n\nThat would allow things like\n\n\\d+ (order by size)\n\nwhich would sort by the numeric size, while displaying the pretty size\nin the output.\n\nThis would also allow more complex orderings that would be hard to\nachieve any other way, such as\n\n\\d+ (order by type, schema, size desc)\n\\dt (order by pg_total_relation_size(name::regclass))\n\n(note the size reported by \\d+ is not the total relation size, because\nit excludes indexes)\n\nAlso, it would allow user-defined WHERE clauses to filter the results\nshown, for example:\n\n\\d+ (where size > pg_size_bytes('1GB'))\n\\dv (where pg_relation_is_updatable(name::regclass, true) != 0)\n\nand many more things are possible, without needing to learn any new\nsyntax, and without needing to keep adding more and more options to\nthe psql syntax.\n\nRegards,\nDean\n\n","msg_date":"Mon, 13 Nov 2017 14:17:22 +0000","msg_from":"Dean Rasheed <dean.a.rasheed@gmail.com>","msg_from_op":false,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com>\nwrote:\n\n> On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com>\n> wrote:\n> > I though about Alexander proposal, and I am thinking so it can be\n> probably\n> > best if we respect psql design. I implemented two command suffixes\n> > (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n> >\n> > so list of tables can be sorted with commands:\n> >\n> > \\dt+sd (in this case, the order is not strict), so command\n> > \\dtsd+ is working too (same \\disd+ or \\di+sd)\n> >\n> > These two chars are acceptable. Same principle is used for \\l command\n> >\n> > \\lsd+ or \\l+sd\n> >\n> > What do you think about it?\n> >\n>\n> I really hate that syntax. This is going to turn into an\n> incomprehensible mess, and isn't easily extended to support other\n> options.\n>\n\n+1. While useful in itself, I think it's definitely a dangerous pattern to\ngo down, as it'll only get worse.\n\n\nI agree with people who have said they would prefer this to be\n> available as a per-command option rather than as a variable that you\n> have to set, but it needs a clearer syntax. I actually like Stephen's\n> idea of using a user-defined SQL snippet, because that's a familiar\n> syntax to people, and it avoids adding an ever-increasing number of\n> options to these commands. Instead, the syntax could simply be:\n>\n\n+1 here as well. And anybody who is actually going to need this level of\ncontrol definitely will know SQL...\n\nAnd if one wants to save some \"standard patterns\", it should be doable to\nsave the pattern itself in a variable and then use it with something like\n\"\\dt :mysort\" and have it expand the normal way there.\n\n-- \n Magnus Hagander\n Me: https://www.hagander.net/ <http://www.hagander.net/>\n Work: https://www.redpill-linpro.com/ <http://www.redpill-linpro.com/>\n\nOn Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com> wrote:On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com> wrote:\n> I though about Alexander proposal, and I am thinking so it can be probably\n> best if we respect psql design. I implemented two command suffixes\n> (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n>\n> so list of tables can be sorted with commands:\n>\n> \\dt+sd (in this case, the order is not strict), so command\n> \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>\n> These two chars are acceptable. Same principle is used for \\l command\n>\n> \\lsd+ or \\l+sd\n>\n> What do you think about it?\n>\n\nI really hate that syntax. This is going to turn into an\nincomprehensible mess, and isn't easily extended to support other\noptions.+1. While useful in itself, I think it's definitely a dangerous pattern to go down, as it'll only get worse.\nI agree with people who have said they would prefer this to be\navailable as a per-command option rather than as a variable that you\nhave to set, but it needs a clearer syntax. I actually like Stephen's\nidea of using a user-defined SQL snippet, because that's a familiar\nsyntax to people, and it avoids adding an ever-increasing number of\noptions to these commands. Instead, the syntax could simply be:+1 here as well. And anybody who is actually going to need this level of control definitely will know SQL...And if one wants to save some \"standard patterns\", it should be doable to save the pattern itself in a variable and then use it with something like \"\\dt :mysort\" and have it expand the normal way there.-- Magnus Hagander Me: https://www.hagander.net/ Work: https://www.redpill-linpro.com/","msg_date":"Tue, 14 Nov 2017 13:26:30 +0100","msg_from":"Magnus Hagander <magnus@hagander.net>","msg_from_op":false,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Tue, Nov 14, 2017 at 3:26 PM, Magnus Hagander <magnus@hagander.net>\nwrote:\n\n> On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com>\n> wrote:\n>\n>> On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com>\n>> wrote:\n>> > I though about Alexander proposal, and I am thinking so it can be\n>> probably\n>> > best if we respect psql design. I implemented two command suffixes\n>> > (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n>> >\n>> > so list of tables can be sorted with commands:\n>> >\n>> > \\dt+sd (in this case, the order is not strict), so command\n>> > \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>> >\n>> > These two chars are acceptable. Same principle is used for \\l command\n>> >\n>> > \\lsd+ or \\l+sd\n>> >\n>> > What do you think about it?\n>> >\n>>\n>> I really hate that syntax. This is going to turn into an\n>> incomprehensible mess, and isn't easily extended to support other\n>> options.\n>>\n>\n> +1. While useful in itself, I think it's definitely a dangerous pattern to\n> go down, as it'll only get worse.\n>\n>\n> I agree with people who have said they would prefer this to be\n>> available as a per-command option rather than as a variable that you\n>> have to set, but it needs a clearer syntax. I actually like Stephen's\n>> idea of using a user-defined SQL snippet, because that's a familiar\n>> syntax to people, and it avoids adding an ever-increasing number of\n>> options to these commands. Instead, the syntax could simply be:\n>>\n>\n> +1 here as well. And anybody who is actually going to need this level of\n> control definitely will know SQL...\n>\n> And if one wants to save some \"standard patterns\", it should be doable to\n> save the pattern itself in a variable and then use it with something like\n> \"\\dt :mysort\" and have it expand the normal way there.\n>\n\n+1\nI agree, that would look better, especially with \"standard patterns\" which\ncould help with too long to type each time SQL snippets.\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nOn Tue, Nov 14, 2017 at 3:26 PM, Magnus Hagander <magnus@hagander.net> wrote:On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com> wrote:On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com> wrote:\n> I though about Alexander proposal, and I am thinking so it can be probably\n> best if we respect psql design. I implemented two command suffixes\n> (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n>\n> so list of tables can be sorted with commands:\n>\n> \\dt+sd (in this case, the order is not strict), so command\n> \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>\n> These two chars are acceptable. Same principle is used for \\l command\n>\n> \\lsd+ or \\l+sd\n>\n> What do you think about it?\n>\n\nI really hate that syntax. This is going to turn into an\nincomprehensible mess, and isn't easily extended to support other\noptions.+1. While useful in itself, I think it's definitely a dangerous pattern to go down, as it'll only get worse.\nI agree with people who have said they would prefer this to be\navailable as a per-command option rather than as a variable that you\nhave to set, but it needs a clearer syntax. I actually like Stephen's\nidea of using a user-defined SQL snippet, because that's a familiar\nsyntax to people, and it avoids adding an ever-increasing number of\noptions to these commands. Instead, the syntax could simply be:+1 here as well. And anybody who is actually going to need this level of control definitely will know SQL...And if one wants to save some \"standard patterns\", it should be doable to save the pattern itself in a variable and then use it with something like \"\\dt :mysort\" and have it expand the normal way there.+1I agree, that would look better, especially with \"standard patterns\" which could help with too long to type each time SQL snippets.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Mon, 27 Nov 2017 10:58:59 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-11-27 8:58 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n\n> On Tue, Nov 14, 2017 at 3:26 PM, Magnus Hagander <magnus@hagander.net>\n> wrote:\n>\n>> On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com>\n>> wrote:\n>>\n>>> On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com>\n>>> wrote:\n>>> > I though about Alexander proposal, and I am thinking so it can be\n>>> probably\n>>> > best if we respect psql design. I implemented two command suffixes\n>>> > (supported only when it has sense) \"s\" sorted by size and \"d\" as\n>>> descent\n>>> >\n>>> > so list of tables can be sorted with commands:\n>>> >\n>>> > \\dt+sd (in this case, the order is not strict), so command\n>>> > \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>>> >\n>>> > These two chars are acceptable. Same principle is used for \\l command\n>>> >\n>>> > \\lsd+ or \\l+sd\n>>> >\n>>> > What do you think about it?\n>>> >\n>>>\n>>> I really hate that syntax. This is going to turn into an\n>>> incomprehensible mess, and isn't easily extended to support other\n>>> options.\n>>>\n>>\n>> +1. While useful in itself, I think it's definitely a dangerous pattern\n>> to go down, as it'll only get worse.\n>>\n>>\n>> I agree with people who have said they would prefer this to be\n>>> available as a per-command option rather than as a variable that you\n>>> have to set, but it needs a clearer syntax. I actually like Stephen's\n>>> idea of using a user-defined SQL snippet, because that's a familiar\n>>> syntax to people, and it avoids adding an ever-increasing number of\n>>> options to these commands. Instead, the syntax could simply be:\n>>>\n>>\n>> +1 here as well. And anybody who is actually going to need this level of\n>> control definitely will know SQL...\n>>\n>> And if one wants to save some \"standard patterns\", it should be doable to\n>> save the pattern itself in a variable and then use it with something like\n>> \"\\dt :mysort\" and have it expand the normal way there.\n>>\n>\n> +1\n> I agree, that would look better, especially with \"standard patterns\" which\n> could help with too long to type each time SQL snippets.\n>\n\nI though about this design more time. I see following disadvantages\n\n1. we are not able to check all possible variants of extended query. If\nthere will be some custom error, then we will raise pretty ugly error\nmessages,\n\n2. I don't thing so using \"Size\" as table size in human readable format and\n\"size\" as table size in raw format is intuitive, but any change of \"Size\"\ncan introduce some (less probability compatibility issues),\n\n3. What queries will contains size calculations? It is not cheap - requires\nAccessShareLock\n\nThis proposal is first time, when we cannot to detect full semantic from\n\\xxx command. When user extend query correctly, then it is better than\nbefore, when not it is worse than before.\n\nRegards\n\nPavel\n\n\n\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n\n2017-11-27 8:58 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Tue, Nov 14, 2017 at 3:26 PM, Magnus Hagander <magnus@hagander.net> wrote:On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com> wrote:On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com> wrote:\n> I though about Alexander proposal, and I am thinking so it can be probably\n> best if we respect psql design. I implemented two command suffixes\n> (supported only when it has sense) \"s\" sorted by size and \"d\" as descent\n>\n> so list of tables can be sorted with commands:\n>\n> \\dt+sd (in this case, the order is not strict), so command\n> \\dtsd+ is working too (same \\disd+ or \\di+sd)\n>\n> These two chars are acceptable. Same principle is used for \\l command\n>\n> \\lsd+ or \\l+sd\n>\n> What do you think about it?\n>\n\nI really hate that syntax. This is going to turn into an\nincomprehensible mess, and isn't easily extended to support other\noptions.+1. While useful in itself, I think it's definitely a dangerous pattern to go down, as it'll only get worse.\nI agree with people who have said they would prefer this to be\navailable as a per-command option rather than as a variable that you\nhave to set, but it needs a clearer syntax. I actually like Stephen's\nidea of using a user-defined SQL snippet, because that's a familiar\nsyntax to people, and it avoids adding an ever-increasing number of\noptions to these commands. Instead, the syntax could simply be:+1 here as well. And anybody who is actually going to need this level of control definitely will know SQL...And if one wants to save some \"standard patterns\", it should be doable to save the pattern itself in a variable and then use it with something like \"\\dt :mysort\" and have it expand the normal way there.+1I agree, that would look better, especially with \"standard patterns\" which could help with too long to type each time SQL snippets.I though about this design more time. I see following disadvantages1. we are not able to check all possible variants of extended query. If there will be some custom error, then we will raise pretty ugly error messages,2. I don't thing so using \"Size\" as table size in human readable format and \"size\" as table size in raw format is intuitive, but any change of \"Size\" can introduce some (less probability compatibility issues),3. What queries will contains size calculations? It is not cheap - requires AccessShareLockThis proposal is first time, when we cannot to detect full semantic from \\xxx command. When user extend query correctly, then it is better than before, when not it is worse than before.RegardsPavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Mon, 27 Nov 2017 20:18:17 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Tue, Nov 28, 2017 at 4:18 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:\n> This proposal is first time, when we cannot to detect full semantic from\n> \\xxx command. When user extend query correctly, then it is better than\n> before, when not it is worse than before.\n\nAs the discussion is still going on, I am moving this patch to next CF.\n-- \nMichael\n\n","msg_date":"Wed, 29 Nov 2017 14:17:16 +0900","msg_from":"Michael Paquier <michael.paquier@gmail.com>","msg_from_op":false,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"Hi!\n\nOn Mon, Nov 27, 2017 at 10:18 PM, Pavel Stehule <pavel.stehule@gmail.com>\nwrote:\n\n> I though about this design more time. I see following disadvantages\n>\n> 1. we are not able to check all possible variants of extended query. If\n> there will be some custom error, then we will raise pretty ugly error\n> messages,\n>\n\nYes, that's an inevitable shortcoming. psql is not backend and can't\nperform all the required checks on its side...\n\n2. I don't thing so using \"Size\" as table size in human readable format and\n> \"size\" as table size in raw format is intuitive, but any change of \"Size\"\n> can introduce some (less probability compatibility issues),\n>\n\nOh, this is surprisingly hard problem which probably have only imperative\nsolution...\n\n3. What queries will contains size calculations? It is not cheap - requires\n> AccessShareLock\n>\n\nSorry, I didn't understand this point. Yes, size calculation requires\nlocking, but that is already true for \\dt+ and \\l+. Why this is\ndisadvantage of proposed approach?\n\n------\nAlexander Korotkov\nPostgres Professional: http://www.postgrespro.com\nThe Russian Postgres Company\n\nHi!On Mon, Nov 27, 2017 at 10:18 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:I though about this design more time. I see following disadvantages1. we are not able to check all possible variants of extended query. If there will be some custom error, then we will raise pretty ugly error messages, Yes, that's an inevitable shortcoming. psql is not backend and can't perform all the required checks on its side... 2. I don't thing so using \"Size\" as table size in human readable format and \"size\" as table size in raw format is intuitive, but any change of \"Size\" can introduce some (less probability compatibility issues),Oh, this is surprisingly hard problem which probably have only imperative solution...3. What queries will contains size calculations? It is not cheap - requires AccessShareLock Sorry, I didn't understand this point. Yes, size calculation requires locking, but that is already true for \\dt+ and \\l+. Why this is disadvantage of proposed approach?------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Sat, 9 Dec 2017 00:49:57 +0300","msg_from":"Alexander Korotkov <a.korotkov@postgrespro.ru>","msg_from_op":false,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"2017-12-08 22:49 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\n\n> Hi!\n>\n> On Mon, Nov 27, 2017 at 10:18 PM, Pavel Stehule <pavel.stehule@gmail.com>\n> wrote:\n>\n>> I though about this design more time. I see following disadvantages\n>>\n>> 1. we are not able to check all possible variants of extended query. If\n>> there will be some custom error, then we will raise pretty ugly error\n>> messages,\n>>\n>\n> Yes, that's an inevitable shortcoming. psql is not backend and can't\n> perform all the required checks on its side...\n>\n> 2. I don't thing so using \"Size\" as table size in human readable format\n>> and \"size\" as table size in raw format is intuitive, but any change of\n>> \"Size\" can introduce some (less probability compatibility issues),\n>>\n>\n> Oh, this is surprisingly hard problem which probably have only imperative\n> solution...\n>\n> 3. What queries will contains size calculations? It is not cheap -\n>> requires AccessShareLock\n>>\n>\n> Sorry, I didn't understand this point. Yes, size calculation requires\n> locking, but that is already true for \\dt+ and \\l+. Why this is\n> disadvantage of proposed approach?\n>\n\nBecause you don't know the filter and sort clause (it can be generic), you\ndon't know, if you should to calculate or not the size. Or there should be\nrule, so filter, order must be limited to displayed columns.\n\nRegards\n\nPavel\n\n\n\n\n> ------\n> Alexander Korotkov\n> Postgres Professional: http://www.postgrespro.com\n> The Russian Postgres Company\n>\n\n2017-12-08 22:49 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:Hi!On Mon, Nov 27, 2017 at 10:18 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:I though about this design more time. I see following disadvantages1. we are not able to check all possible variants of extended query. If there will be some custom error, then we will raise pretty ugly error messages, Yes, that's an inevitable shortcoming. psql is not backend and can't perform all the required checks on its side... 2. I don't thing so using \"Size\" as table size in human readable format and \"size\" as table size in raw format is intuitive, but any change of \"Size\" can introduce some (less probability compatibility issues),Oh, this is surprisingly hard problem which probably have only imperative solution...3. What queries will contains size calculations? It is not cheap - requires AccessShareLock Sorry, I didn't understand this point. Yes, size calculation requires locking, but that is already true for \\dt+ and \\l+. Why this is disadvantage of proposed approach?Because you don't know the filter and sort clause (it can be generic), you don't know, if you should to calculate or not the size. Or there should be rule, so filter, order must be limited to displayed columns.RegardsPavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company","msg_date":"Sat, 9 Dec 2017 06:11:33 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"On Thu, Nov 2, 2017 at 12:44 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:\n> I am sending updated patch with some basic doc\n\nHi Pavel,\n\nI am not sure what the status of this patch is, but FYI:\n\nstartup.c: In function ‘main’:\nstartup.c:284:3: error: too few arguments to function ‘listAllDbs’\n success = listAllDbs(NULL, false);\n ^\nIn file included from startup.c:21:0:\ndescribe.h:68:13: note: declared here\n extern bool listAllDbs(const char *pattern, bool verbose, sortby_type\nsortby, bool sort_desc);\n ^\n\n-- \nThomas Munro\nhttp://www.enterprisedb.com\n\n","msg_date":"Fri, 12 Jan 2018 14:35:37 +1300","msg_from":"Thomas Munro <thomas.munro@enterprisedb.com>","msg_from_op":false,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"},{"msg_contents":"Hi\n\n2018-01-12 2:35 GMT+01:00 Thomas Munro <thomas.munro@enterprisedb.com>:\n\n> On Thu, Nov 2, 2017 at 12:44 AM, Pavel Stehule <pavel.stehule@gmail.com>\n> wrote:\n> > I am sending updated patch with some basic doc\n>\n> Hi Pavel,\n>\n> I am not sure what the status of this patch is, but FYI:\n>\n> startup.c: In function ‘main’:\n> startup.c:284:3: error: too few arguments to function ‘listAllDbs’\n> success = listAllDbs(NULL, false);\n> ^\n> In file included from startup.c:21:0:\n> describe.h:68:13: note: declared here\n> extern bool listAllDbs(const char *pattern, bool verbose, sortby_type\n> sortby, bool sort_desc);\n> ^\n>\n\nThere are no agreement about optimal form of this feature - so I'll remove\nthis patch from commitfest.\n\nRegards\n\nPavel\n\n\n>\n> --\n> Thomas Munro\n> http://www.enterprisedb.com\n>\n\nHi2018-01-12 2:35 GMT+01:00 Thomas Munro <thomas.munro@enterprisedb.com>:On Thu, Nov 2, 2017 at 12:44 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:\n> I am sending updated patch with some basic doc\n\nHi Pavel,\n\nI am not sure what the status of this patch is, but FYI:\n\nstartup.c: In function ‘main’:\nstartup.c:284:3: error: too few arguments to function ‘listAllDbs’\n success = listAllDbs(NULL, false);\n ^\nIn file included from startup.c:21:0:\ndescribe.h:68:13: note: declared here\n extern bool listAllDbs(const char *pattern, bool verbose, sortby_type\nsortby, bool sort_desc);\n ^There are no agreement about optimal form of this feature - so I'll remove this patch from commitfest.RegardsPavel \n\n--\nThomas Munro\nhttp://www.enterprisedb.com","msg_date":"Fri, 12 Jan 2018 16:11:02 +0100","msg_from":"Pavel Stehule <pavel.stehule@gmail.com>","msg_from_op":true,"msg_subject":"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\n describe commands, when size is printed"}],"string":"[\n {\n \"msg_contents\": \"Hi\\n\\nCurrently is not possible to control sort columns for \\\\d* commands. Usually\\nschema and table name is used. Really often task is collect the most big\\nobjects in database. \\\"\\\\dt+, \\\\di+\\\" shows necessary information, but not in\\npractical order.\\n\\nInstead introduction some additional flags to backslash commands, I propose\\na special psql variable that can be used for specification of order used\\nwhen some plus command is used.\\n\\nsome like\\n\\nset EXTENDED_DESCRIBE_SORT size_desc\\n\\\\dt+\\n\\\\l+\\n\\\\di+\\n\\nPossible variants: schema_table, table_schema, size_desc, size_asc\\n\\nComments, notes?\\n\\nRegards\\n\\nPavel\\n\\nHiCurrently is not possible to control sort columns for \\\\d* commands. Usually schema and table name is used. Really often task is collect the most big objects in database. \\\"\\\\dt+, \\\\di+\\\" shows necessary information, but not in practical order.Instead introduction some additional flags to backslash commands, I propose a special psql variable that can be used for specification of order used when some plus command is used.some likeset EXTENDED_DESCRIBE_SORT size_desc\\\\dt+\\\\l+\\\\di+Possible variants: schema_table, table_schema, size_desc, size_ascComments, notes?RegardsPavel\",\n \"msg_date\": \"Thu, 23 Feb 2017 12:17:31 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"proposal - psql: possibility to specify sort for describe commands,\\n when size is printed\"\n },\n {\n \"msg_contents\": \"Hi\\n\\n2017-02-23 12:17 GMT+01:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n\\n> Hi\\n>\\n> Currently is not possible to control sort columns for \\\\d* commands.\\n> Usually schema and table name is used. Really often task is collect the\\n> most big objects in database. \\\"\\\\dt+, \\\\di+\\\" shows necessary information, but\\n> not in practical order.\\n>\\n> Instead introduction some additional flags to backslash commands, I\\n> propose a special psql variable that can be used for specification of order\\n> used when some plus command is used.\\n>\\n> some like\\n>\\n> set EXTENDED_DESCRIBE_SORT size_desc\\n> \\\\dt+\\n> \\\\l+\\n> \\\\di+\\n>\\n> Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> Comments, notes?\\n>\\n\\nhere is a patch\\n\\nRegards\\n\\nPavel\\n\\n\\n>\\n> Regards\\n>\\n> Pavel\\n>\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\",\n \"msg_date\": \"Fri, 24 Feb 2017 22:32:47 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: proposal - psql: possibility to specify sort for describe\\n commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 2/24/17 16:32, Pavel Stehule wrote:\\n> set EXTENDED_DESCRIBE_SORT size_desc\\n> \\\\dt+\\n> \\\\l+\\n> \\\\di+\\n> \\n> Possible variants: schema_table, table_schema, size_desc, size_asc\\n\\nI can see this being useful, but I think it needs to be organized a\\nlittle better.\\n\\nSort key and sort direction should be separate settings.\\n\\nI'm not sure why we need to have separate settings to sort by schema\\nname and table name. But if we do, then we should support that for all\\nobject types. I think maybe that's something we shouldn't get into\\nright now.\\n\\nSo I would have one setting for sort key = {name|size} and on for sort\\ndirection = {asc|desc}.\\n\\n-- \\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Fri, 10 Mar 2017 09:10:08 -0500\",\n \"msg_from\": \"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> > \\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n> \\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n> \\n> Sort key and sort direction should be separate settings.\\n> \\n> I'm not sure why we need to have separate settings to sort by schema\\n> name and table name. But if we do, then we should support that for all\\n> object types. I think maybe that's something we shouldn't get into\\n> right now.\\n> \\n> So I would have one setting for sort key = {name|size} and on for sort\\n> direction = {asc|desc}.\\n\\nPerhaps I'm trying to be overly cute here, but why not let the user\\nsimply provide a bit of SQL to be put at the end of the query?\\n\\nThat is, something like:\\n\\n\\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\\n\\nThanks!\\n\\nStephen\",\n \"msg_date\": \"Fri, 10 Mar 2017 09:16:45 -0500\",\n \"msg_from\": \"Stephen Frost <sfrost@snowman.net>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Fri, Mar 10, 2017 at 5:10 PM, Peter Eisentraut <\\npeter.eisentraut@2ndquadrant.com> wrote:\\n\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> >\\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n>\\n> Sort key and sort direction should be separate settings.\\n>\\n\\nI agree.\\n\\nI'm not sure why we need to have separate settings to sort by schema\\n> name and table name.\\n\\n\\nI think sorting by schema name, object name makes sense for people, who\\nhave objects of same name in different schemas.\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Fri, Mar 10, 2017 at 5:10 PM, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:On 2/24/17 16:32, Pavel Stehule wrote:\\n> set EXTENDED_DESCRIBE_SORT size_desc\\n> \\\\dt+\\n> \\\\l+\\n> \\\\di+\\n>\\n> Possible variants: schema_table, table_schema, size_desc, size_asc\\n\\nI can see this being useful, but I think it needs to be organized a\\nlittle better.\\n\\nSort key and sort direction should be separate settings. I agree.\\nI'm not sure why we need to have separate settings to sort by schema\\nname and table name.I think sorting by schema name, object name makes sense for people, who have objects of same name in different schemas.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Fri, 10 Mar 2017 17:45:43 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-03-10 15:10 GMT+01:00 Peter Eisentraut <\\npeter.eisentraut@2ndquadrant.com>:\\n\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> >\\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n>\\n> Sort key and sort direction should be separate settings.\\n>\\n\\nok\\n\\nmaybe\\n\\nPREFERRED_SORT_COLUMNS\\nand PREFERRED_SORT_DIRECTION ?\\n\\n>\\n> I'm not sure why we need to have separate settings to sort by schema\\n> name and table name. But if we do, then we should support that for all\\n> object types. I think maybe that's something we shouldn't get into\\n> right now.\\n>\\n\\nIt can be useful, when you repeat one table name in more schema - usually,\\nwhere schema is related per one customer, project, ...\\n\\nRegards\\n\\nPavel\\n\\n\\n>\\n> So I would have one setting for sort key = {name|size} and on for sort\\n> direction = {asc|desc}.\\n>\\n> --\\n> Peter Eisentraut http://www.2ndQuadrant.com/\\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>\\n\\n2017-03-10 15:10 GMT+01:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 2/24/17 16:32, Pavel Stehule wrote:\\n> set EXTENDED_DESCRIBE_SORT size_desc\\n> \\\\dt+\\n> \\\\l+\\n> \\\\di+\\n>\\n> Possible variants: schema_table, table_schema, size_desc, size_asc\\n\\nI can see this being useful, but I think it needs to be organized a\\nlittle better.\\n\\nSort key and sort direction should be separate settings.okmaybePREFERRED_SORT_COLUMNSand PREFERRED_SORT_DIRECTION ? \\n\\nI'm not sure why we need to have separate settings to sort by schema\\nname and table name. But if we do, then we should support that for all\\nobject types. I think maybe that's something we shouldn't get into\\nright now.It can be useful, when you repeat one table name in more schema - usually, where schema is related per one customer, project, ...RegardsPavel \\n\\nSo I would have one setting for sort key = {name|size} and on for sort\\ndirection = {asc|desc}.\\n\\n--\\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\",\n \"msg_date\": \"Fri, 10 Mar 2017 15:57:38 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:\\n\\n> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> > On 2/24/17 16:32, Pavel Stehule wrote:\\n> > > set EXTENDED_DESCRIBE_SORT size_desc\\n> > > \\\\dt+\\n> > > \\\\l+\\n> > > \\\\di+\\n> > >\\n> > > Possible variants: schema_table, table_schema, size_desc, size_asc\\n> >\\n> > I can see this being useful, but I think it needs to be organized a\\n> > little better.\\n> >\\n> > Sort key and sort direction should be separate settings.\\n> >\\n> > I'm not sure why we need to have separate settings to sort by schema\\n> > name and table name. But if we do, then we should support that for all\\n> > object types. I think maybe that's something we shouldn't get into\\n> > right now.\\n> >\\n> > So I would have one setting for sort key = {name|size} and on for sort\\n> > direction = {asc|desc}.\\n>\\n> Perhaps I'm trying to be overly cute here, but why not let the user\\n> simply provide a bit of SQL to be put at the end of the query?\\n>\\n> That is, something like:\\n>\\n> \\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\\n>\\n\\nI think that's the question of usability. After all, one can manually type\\ncorresponding SQL instead of \\\\d* commands. However, it's quite cumbersome\\nto do this every time.\\nI found quite useful to being able to switch between different sortings\\nquickly. For instance, after seeing tables sorted by name, user can\\nrequire them sorted by size descending, then sorted by size ascending,\\netc...\\nTherefore, I find user-defined SQL clause to be cumbersome. Even psql\\nvariable itself seems to be cumbersome for me.\\nI would propose to add sorting as second optional argument to \\\\d*\\ncommands. Any thoughts?\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> >\\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n>\\n> Sort key and sort direction should be separate settings.\\n>\\n> I'm not sure why we need to have separate settings to sort by schema\\n> name and table name. But if we do, then we should support that for all\\n> object types. I think maybe that's something we shouldn't get into\\n> right now.\\n>\\n> So I would have one setting for sort key = {name|size} and on for sort\\n> direction = {asc|desc}.\\n\\nPerhaps I'm trying to be overly cute here, but why not let the user\\nsimply provide a bit of SQL to be put at the end of the query?\\n\\nThat is, something like:\\n\\n\\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\\\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\\\d* commands. Any thoughts?------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Fri, 10 Mar 2017 18:00:14 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-03-10 15:16 GMT+01:00 Stephen Frost <sfrost@snowman.net>:\\n\\n> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> > On 2/24/17 16:32, Pavel Stehule wrote:\\n> > > set EXTENDED_DESCRIBE_SORT size_desc\\n> > > \\\\dt+\\n> > > \\\\l+\\n> > > \\\\di+\\n> > >\\n> > > Possible variants: schema_table, table_schema, size_desc, size_asc\\n> >\\n> > I can see this being useful, but I think it needs to be organized a\\n> > little better.\\n> >\\n> > Sort key and sort direction should be separate settings.\\n> >\\n> > I'm not sure why we need to have separate settings to sort by schema\\n> > name and table name. But if we do, then we should support that for all\\n> > object types. I think maybe that's something we shouldn't get into\\n> > right now.\\n> >\\n> > So I would have one setting for sort key = {name|size} and on for sort\\n> > direction = {asc|desc}.\\n>\\n> Perhaps I'm trying to be overly cute here, but why not let the user\\n> simply provide a bit of SQL to be put at the end of the query?\\n>\\n> That is, something like:\\n>\\n> \\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\\n>\\n\\nFor example - the size is displayed in pretty form - raw form is not\\ndisplayed - so simple ORDER BY clause is not possible.\\n\\nBut setting LIMIT is not bad idea - although it is probably much more\\ncomplex for implementation.\\n\\n\\\\pset DESCRIBE_LIMIT 100\\n\\\\pset EXTENDED_DESCRIBE_LIMIT 100\\n\\ncan be implemented as next step\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n>\\n> Thanks!\\n>\\n> Stephen\\n>\\n\\n2017-03-10 15:16 GMT+01:00 Stephen Frost <sfrost@snowman.net>:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> >\\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n>\\n> Sort key and sort direction should be separate settings.\\n>\\n> I'm not sure why we need to have separate settings to sort by schema\\n> name and table name. But if we do, then we should support that for all\\n> object types. I think maybe that's something we shouldn't get into\\n> right now.\\n>\\n> So I would have one setting for sort key = {name|size} and on for sort\\n> direction = {asc|desc}.\\n\\nPerhaps I'm trying to be overly cute here, but why not let the user\\nsimply provide a bit of SQL to be put at the end of the query?\\n\\nThat is, something like:\\n\\n\\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'For example - the size is displayed in pretty form - raw form is not displayed - so simple ORDER BY clause is not possible.But setting LIMIT is not bad idea - although it is probably much more complex for implementation.\\\\pset DESCRIBE_LIMIT 100\\\\pset EXTENDED_DESCRIBE_LIMIT 100can be implemented as next stepRegardsPavel \\n\\nThanks!\\n\\nStephen\",\n \"msg_date\": \"Fri, 10 Mar 2017 16:01:13 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 3/10/17 09:57, Pavel Stehule wrote:\\n> PREFERRED_SORT_COLUMNS\\n> and PREFERRED_SORT_DIRECTION ? \\n\\nI think the name \\\"preferred\\\" implies that it will be ignored if it's not\\nfound or something like that, but I don't think that's what you are\\nimplementing.\\n\\n-- \\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Fri, 10 Mar 2017 10:05:04 -0500\",\n \"msg_from\": \"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n\\n> On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:\\n>\\n>> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n>> > On 2/24/17 16:32, Pavel Stehule wrote:\\n>> > > set EXTENDED_DESCRIBE_SORT size_desc\\n>> > > \\\\dt+\\n>> > > \\\\l+\\n>> > > \\\\di+\\n>> > >\\n>> > > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>> >\\n>> > I can see this being useful, but I think it needs to be organized a\\n>> > little better.\\n>> >\\n>> > Sort key and sort direction should be separate settings.\\n>> >\\n>> > I'm not sure why we need to have separate settings to sort by schema\\n>> > name and table name. But if we do, then we should support that for all\\n>> > object types. I think maybe that's something we shouldn't get into\\n>> > right now.\\n>> >\\n>> > So I would have one setting for sort key = {name|size} and on for sort\\n>> > direction = {asc|desc}.\\n>>\\n>> Perhaps I'm trying to be overly cute here, but why not let the user\\n>> simply provide a bit of SQL to be put at the end of the query?\\n>>\\n>> That is, something like:\\n>>\\n>> \\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\\n>>\\n>\\n> I think that's the question of usability. After all, one can manually\\n> type corresponding SQL instead of \\\\d* commands. However, it's quite\\n> cumbersome to do this every time.\\n> I found quite useful to being able to switch between different sortings\\n> quickly. For instance, after seeing tables sorted by name, user can\\n> require them sorted by size descending, then sorted by size ascending,\\n> etc...\\n> Therefore, I find user-defined SQL clause to be cumbersome. Even psql\\n> variable itself seems to be cumbersome for me.\\n> I would propose to add sorting as second optional argument to \\\\d*\\n> commands. Any thoughts?\\n>\\n\\nThis proposal was here already - maybe two years ago. The psql command\\nparser doesn't allow any complex syntax - more - the more parameters in one\\npsql commands is hard to remember, hard to read.\\n\\nWith my proposal, and patch I would to cover following use case. It is real\\ncase. Anytime when we used \\\\dt+ in psql we needed sort by size desc. When\\nwe should to see a size, then the top is interesting. This case is not\\nabsolute, but very often, so I would to create some simple way, how to do\\nsome parametrization (really simple).\\n\\nPavel\\n\\n\\n\\n>\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n\\n2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> >\\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n>\\n> Sort key and sort direction should be separate settings.\\n>\\n> I'm not sure why we need to have separate settings to sort by schema\\n> name and table name. But if we do, then we should support that for all\\n> object types. I think maybe that's something we shouldn't get into\\n> right now.\\n>\\n> So I would have one setting for sort key = {name|size} and on for sort\\n> direction = {asc|desc}.\\n\\nPerhaps I'm trying to be overly cute here, but why not let the user\\nsimply provide a bit of SQL to be put at the end of the query?\\n\\nThat is, something like:\\n\\n\\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\\\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\\\d* commands. Any thoughts?This proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.With my proposal, and patch I would to cover following use case. It is real case. Anytime when we used \\\\dt+ in psql we needed sort by size desc. When we should to see a size, then the top is interesting. This case is not absolute, but very often, so I would to create some simple way, how to do some parametrization (really simple).Pavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Fri, 10 Mar 2017 16:06:52 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-03-10 16:05 GMT+01:00 Peter Eisentraut <\\npeter.eisentraut@2ndquadrant.com>:\\n\\n> On 3/10/17 09:57, Pavel Stehule wrote:\\n> > PREFERRED_SORT_COLUMNS\\n> > and PREFERRED_SORT_DIRECTION ?\\n>\\n> I think the name \\\"preferred\\\" implies that it will be ignored if it's not\\n> found or something like that, but I don't think that's what you are\\n> implementing.\\n>\\n\\nok if it will be used only for verbose describe commands , then the name\\nEXTENDED_DESCRIBE_SORT_COLUMNS, and EXTENDED_DESCRIBE_SORT_DIRECTION.\\n\\nPavel\\n\\n\\n>\\n> --\\n> Peter Eisentraut http://www.2ndQuadrant.com/\\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>\\n\\n2017-03-10 16:05 GMT+01:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 3/10/17 09:57, Pavel Stehule wrote:\\n> PREFERRED_SORT_COLUMNS\\n> and PREFERRED_SORT_DIRECTION ?\\n\\nI think the name \\\"preferred\\\" implies that it will be ignored if it's not\\nfound or something like that, but I don't think that's what you are\\nimplementing.ok if it will be used only for verbose describe commands , then the name EXTENDED_DESCRIBE_SORT_COLUMNS, and EXTENDED_DESCRIBE_SORT_DIRECTION.Pavel \\n\\n--\\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\",\n \"msg_date\": \"Fri, 10 Mar 2017 16:08:58 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-03-10 15:45 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n\\n> On Fri, Mar 10, 2017 at 5:10 PM, Peter Eisentraut <\\n> peter.eisentraut@2ndquadrant.com> wrote:\\n>\\n>> On 2/24/17 16:32, Pavel Stehule wrote:\\n>> > set EXTENDED_DESCRIBE_SORT size_desc\\n>> > \\\\dt+\\n>> > \\\\l+\\n>> > \\\\di+\\n>> >\\n>> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>>\\n>> I can see this being useful, but I think it needs to be organized a\\n>> little better.\\n>>\\n>> Sort key and sort direction should be separate settings.\\n>>\\n>\\n> I agree.\\n>\\n> I'm not sure why we need to have separate settings to sort by schema\\n>> name and table name.\\n>\\n>\\n> I think sorting by schema name, object name makes sense for people, who\\n> have objects of same name in different schemas.\\n>\\n\\nI am sending a updated version with separated sort direction in special\\nvariable\\n\\nThere is a question. Has desc direction sense for columns like schema or\\ntable name?\\n\\nUsing desc, asc for size is natural. But for tablename?\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n>\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n>\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\",\n \"msg_date\": \"Sat, 11 Mar 2017 13:06:13 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"Hi Alexander,\\n\\nOn 3/11/17 7:06 AM, Pavel Stehule wrote:\\n\\n> I am sending a updated version with separated sort direction in special\\n> variable\\n>\\n> There is a question. Has desc direction sense for columns like schema or\\n> table name?\\n>\\n> Using desc, asc for size is natural. But for tablename?\\n\\nDo you know when you'll have a chance to review the updated patch?\\n\\nThanks,\\n-- \\n-David\\ndavid@pgmasters.net\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Tue, 21 Mar 2017 18:26:59 -0400\",\n \"msg_from\": \"David Steele <david@pgmasters.net>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: proposal - psql: possibility to specify sort for describe\\n commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com>\\nwrote:\\n\\n> 2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n>\\n>> On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net>\\n>> wrote:\\n>>\\n>>> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n>>> > On 2/24/17 16:32, Pavel Stehule wrote:\\n>>> > > set EXTENDED_DESCRIBE_SORT size_desc\\n>>> > > \\\\dt+\\n>>> > > \\\\l+\\n>>> > > \\\\di+\\n>>> > >\\n>>> > > Possible variants: schema_table, table_schema, size_desc,\\n>>> size_asc\\n>>> >\\n>>> > I can see this being useful, but I think it needs to be organized a\\n>>> > little better.\\n>>> >\\n>>> > Sort key and sort direction should be separate settings.\\n>>> >\\n>>> > I'm not sure why we need to have separate settings to sort by schema\\n>>> > name and table name. But if we do, then we should support that for all\\n>>> > object types. I think maybe that's something we shouldn't get into\\n>>> > right now.\\n>>> >\\n>>> > So I would have one setting for sort key = {name|size} and on for sort\\n>>> > direction = {asc|desc}.\\n>>>\\n>>> Perhaps I'm trying to be overly cute here, but why not let the user\\n>>> simply provide a bit of SQL to be put at the end of the query?\\n>>>\\n>>> That is, something like:\\n>>>\\n>>> \\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\\n>>>\\n>>\\n>> I think that's the question of usability. After all, one can manually\\n>> type corresponding SQL instead of \\\\d* commands. However, it's quite\\n>> cumbersome to do this every time.\\n>> I found quite useful to being able to switch between different sortings\\n>> quickly. For instance, after seeing tables sorted by name, user can\\n>> require them sorted by size descending, then sorted by size ascending,\\n>> etc...\\n>> Therefore, I find user-defined SQL clause to be cumbersome. Even psql\\n>> variable itself seems to be cumbersome for me.\\n>> I would propose to add sorting as second optional argument to \\\\d*\\n>> commands. Any thoughts?\\n>>\\n>\\n> This proposal was here already - maybe two years ago. The psql command\\n> parser doesn't allow any complex syntax - more - the more parameters in one\\n> psql commands is hard to remember, hard to read.\\n>\\n\\nCould you please provide a link to this discussion. Probably working with\\nmultiple parameters in psql commands require some rework, but that's\\ndefinitely doable.\\n\\n\\n> With my proposal, and patch I would to cover following use case. It is\\n> real case. Anytime when we used \\\\dt+ in psql we needed sort by size desc.\\n> When we should to see a size, then the top is interesting. This case is not\\n> absolute, but very often, so I would to create some simple way, how to do\\n> some parametrization (really simple).\\n>\\n\\nWe could combine both approaches: add parameters to psql commands and add\\npsql DEFAULT_(SORT_COLUMNS|DIRECTION|LIMIT) parameters.\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> >\\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n>\\n> Sort key and sort direction should be separate settings.\\n>\\n> I'm not sure why we need to have separate settings to sort by schema\\n> name and table name. But if we do, then we should support that for all\\n> object types. I think maybe that's something we shouldn't get into\\n> right now.\\n>\\n> So I would have one setting for sort key = {name|size} and on for sort\\n> direction = {asc|desc}.\\n\\nPerhaps I'm trying to be overly cute here, but why not let the user\\nsimply provide a bit of SQL to be put at the end of the query?\\n\\nThat is, something like:\\n\\n\\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\\\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\\\d* commands. Any thoughts?This proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.Could you please provide a link to this discussion. Probably working with multiple parameters in psql commands require some rework, but that's definitely doable. With my proposal, and patch I would to cover following use case. It is real case. Anytime when we used \\\\dt+ in psql we needed sort by size desc. When we should to see a size, then the top is interesting. This case is not absolute, but very often, so I would to create some simple way, how to do some parametrization (really simple).We could combine both approaches: add parameters to psql commands and add psql DEFAULT_(SORT_COLUMNS|DIRECTION|LIMIT) parameters.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Mon, 27 Mar 2017 14:59:32 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-03-27 13:59 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n\\n> On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com>\\n> wrote:\\n>\\n>> 2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>\\n>> :\\n>>\\n>>> On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net>\\n>>> wrote:\\n>>>\\n>>>> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n>>>> > On 2/24/17 16:32, Pavel Stehule wrote:\\n>>>> > > set EXTENDED_DESCRIBE_SORT size_desc\\n>>>> > > \\\\dt+\\n>>>> > > \\\\l+\\n>>>> > > \\\\di+\\n>>>> > >\\n>>>> > > Possible variants: schema_table, table_schema, size_desc,\\n>>>> size_asc\\n>>>> >\\n>>>> > I can see this being useful, but I think it needs to be organized a\\n>>>> > little better.\\n>>>> >\\n>>>> > Sort key and sort direction should be separate settings.\\n>>>> >\\n>>>> > I'm not sure why we need to have separate settings to sort by schema\\n>>>> > name and table name. But if we do, then we should support that for\\n>>>> all\\n>>>> > object types. I think maybe that's something we shouldn't get into\\n>>>> > right now.\\n>>>> >\\n>>>> > So I would have one setting for sort key = {name|size} and on for sort\\n>>>> > direction = {asc|desc}.\\n>>>>\\n>>>> Perhaps I'm trying to be overly cute here, but why not let the user\\n>>>> simply provide a bit of SQL to be put at the end of the query?\\n>>>>\\n>>>> That is, something like:\\n>>>>\\n>>>> \\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\\n>>>>\\n>>>\\n>>> I think that's the question of usability. After all, one can manually\\n>>> type corresponding SQL instead of \\\\d* commands. However, it's quite\\n>>> cumbersome to do this every time.\\n>>> I found quite useful to being able to switch between different sortings\\n>>> quickly. For instance, after seeing tables sorted by name, user can\\n>>> require them sorted by size descending, then sorted by size ascending,\\n>>> etc...\\n>>> Therefore, I find user-defined SQL clause to be cumbersome. Even psql\\n>>> variable itself seems to be cumbersome for me.\\n>>> I would propose to add sorting as second optional argument to \\\\d*\\n>>> commands. Any thoughts?\\n>>>\\n>>\\n>> This proposal was here already - maybe two years ago. The psql command\\n>> parser doesn't allow any complex syntax - more - the more parameters in one\\n>> psql commands is hard to remember, hard to read.\\n>>\\n>\\n> Could you please provide a link to this discussion. Probably working with\\n> multiple parameters in psql commands require some rework, but that's\\n> definitely doable.\\n>\\n\\nhttp://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/proposal-psql-show-longest-tables/oldest\\nhttps://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.com\\n\\n\\n>\\n>> With my proposal, and patch I would to cover following use case. It is\\n>> real case. Anytime when we used \\\\dt+ in psql we needed sort by size desc.\\n>> When we should to see a size, then the top is interesting. This case is not\\n>> absolute, but very often, so I would to create some simple way, how to do\\n>> some parametrization (really simple).\\n>>\\n>\\n> We could combine both approaches: add parameters to psql commands and add\\n> psql DEFAULT_(SORT_COLUMNS|DIRECTION|LIMIT) parameters.\\n>\\n\\nIt is possible - This moment is my interest concentrated to psql settings -\\nthe unpractical order in \\\\dt+ irritate me :). I understand so it depends on\\nuse-case. I worked in OLAP and still I have lot of customers with\\nperformance incidents - the first task - show most big tables, most big\\nindexes.\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n\\n\\n>\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n>\\n\\n2017-03-27 13:59 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> >\\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n>\\n> Sort key and sort direction should be separate settings.\\n>\\n> I'm not sure why we need to have separate settings to sort by schema\\n> name and table name. But if we do, then we should support that for all\\n> object types. I think maybe that's something we shouldn't get into\\n> right now.\\n>\\n> So I would have one setting for sort key = {name|size} and on for sort\\n> direction = {asc|desc}.\\n\\nPerhaps I'm trying to be overly cute here, but why not let the user\\nsimply provide a bit of SQL to be put at the end of the query?\\n\\nThat is, something like:\\n\\n\\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\\\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\\\d* commands. Any thoughts?This proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.Could you please provide a link to this discussion. Probably working with multiple parameters in psql commands require some rework, but that's definitely doable.http://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/proposal-psql-show-longest-tables/oldesthttps://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.com With my proposal, and patch I would to cover following use case. It is real case. Anytime when we used \\\\dt+ in psql we needed sort by size desc. When we should to see a size, then the top is interesting. This case is not absolute, but very often, so I would to create some simple way, how to do some parametrization (really simple).We could combine both approaches: add parameters to psql commands and add psql DEFAULT_(SORT_COLUMNS|DIRECTION|LIMIT) parameters.It is possible - This moment is my interest concentrated to psql settings - the unpractical order in \\\\dt+ irritate me :). I understand so it depends on use-case. I worked in OLAP and still I have lot of customers with performance incidents - the first task - show most big tables, most big indexes.RegardsPavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Tue, 28 Mar 2017 09:12:26 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Tue, Mar 28, 2017 at 10:12 AM, Pavel Stehule <pavel.stehule@gmail.com>\\nwrote:\\n\\n> 2017-03-27 13:59 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n>\\n>> On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com>\\n>> wrote:\\n>>\\n>>> 2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru\\n>>> >:\\n>>>\\n>>>> On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net>\\n>>>> wrote:\\n>>>>\\n>>>>> * Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n>>>>> > On 2/24/17 16:32, Pavel Stehule wrote:\\n>>>>> > > set EXTENDED_DESCRIBE_SORT size_desc\\n>>>>> > > \\\\dt+\\n>>>>> > > \\\\l+\\n>>>>> > > \\\\di+\\n>>>>> > >\\n>>>>> > > Possible variants: schema_table, table_schema, size_desc,\\n>>>>> size_asc\\n>>>>> >\\n>>>>> > I can see this being useful, but I think it needs to be organized a\\n>>>>> > little better.\\n>>>>> >\\n>>>>> > Sort key and sort direction should be separate settings.\\n>>>>> >\\n>>>>> > I'm not sure why we need to have separate settings to sort by schema\\n>>>>> > name and table name. But if we do, then we should support that for\\n>>>>> all\\n>>>>> > object types. I think maybe that's something we shouldn't get into\\n>>>>> > right now.\\n>>>>> >\\n>>>>> > So I would have one setting for sort key = {name|size} and on for\\n>>>>> sort\\n>>>>> > direction = {asc|desc}.\\n>>>>>\\n>>>>> Perhaps I'm trying to be overly cute here, but why not let the user\\n>>>>> simply provide a bit of SQL to be put at the end of the query?\\n>>>>>\\n>>>>> That is, something like:\\n>>>>>\\n>>>>> \\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'\\n>>>>>\\n>>>>\\n>>>> I think that's the question of usability. After all, one can manually\\n>>>> type corresponding SQL instead of \\\\d* commands. However, it's quite\\n>>>> cumbersome to do this every time.\\n>>>> I found quite useful to being able to switch between different sortings\\n>>>> quickly. For instance, after seeing tables sorted by name, user can\\n>>>> require them sorted by size descending, then sorted by size ascending,\\n>>>> etc...\\n>>>> Therefore, I find user-defined SQL clause to be cumbersome. Even psql\\n>>>> variable itself seems to be cumbersome for me.\\n>>>> I would propose to add sorting as second optional argument to \\\\d*\\n>>>> commands. Any thoughts?\\n>>>>\\n>>>\\n>>> This proposal was here already - maybe two years ago. The psql command\\n>>> parser doesn't allow any complex syntax - more - the more parameters in one\\n>>> psql commands is hard to remember, hard to read.\\n>>>\\n>>\\n>> Could you please provide a link to this discussion. Probably working\\n>> with multiple parameters in psql commands require some rework, but that's\\n>> definitely doable.\\n>>\\n>\\n> http://grokbase.com/t/postgresql/pgsql-hackers/\\n> 137nt5p6s0/proposal-psql-show-longest-tables/oldest\\n> https://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRT\\n> iUQJQHwNJ8ecN2W@mail.gmail.com\\n>\\n\\nI took a look to these threads, but I didn't find place where difficulties\\nof adding extra arguments to psql commands are pointed.\\nCould you, please, point particular messages about it?\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Tue, Mar 28, 2017 at 10:12 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-03-27 13:59 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 6:06 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-03-10 16:00 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Mar 10, 2017 at 5:16 PM, Stephen Frost <sfrost@snowman.net> wrote:* Peter Eisentraut (peter.eisentraut@2ndquadrant.com) wrote:\\n> On 2/24/17 16:32, Pavel Stehule wrote:\\n> > set EXTENDED_DESCRIBE_SORT size_desc\\n> > \\\\dt+\\n> > \\\\l+\\n> > \\\\di+\\n> >\\n> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n>\\n> I can see this being useful, but I think it needs to be organized a\\n> little better.\\n>\\n> Sort key and sort direction should be separate settings.\\n>\\n> I'm not sure why we need to have separate settings to sort by schema\\n> name and table name. But if we do, then we should support that for all\\n> object types. I think maybe that's something we shouldn't get into\\n> right now.\\n>\\n> So I would have one setting for sort key = {name|size} and on for sort\\n> direction = {asc|desc}.\\n\\nPerhaps I'm trying to be overly cute here, but why not let the user\\nsimply provide a bit of SQL to be put at the end of the query?\\n\\nThat is, something like:\\n\\n\\\\pset EXTENDED_DESCRIBE_ORDER_LIMIT 'ORDER BY 5 DESC LIMIT 10'I think that's the question of usability. After all, one can manually type corresponding SQL instead of \\\\d* commands. However, it's quite cumbersome to do this every time.I found quite useful to being able to switch between different sortings quickly. For instance, after seeing tables sorted by name, user can require them sorted by size descending, then sorted by size ascending, etc...Therefore, I find user-defined SQL clause to be cumbersome. Even psql variable itself seems to be cumbersome for me.I would propose to add sorting as second optional argument to \\\\d* commands. Any thoughts?This proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.Could you please provide a link to this discussion. Probably working with multiple parameters in psql commands require some rework, but that's definitely doable.http://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/proposal-psql-show-longest-tables/oldesthttps://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.comI took a look to these threads, but I didn't find place where difficulties of adding extra arguments to psql commands are pointed.Could you, please, point particular messages about it?------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Thu, 30 Mar 2017 18:09:51 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \">\\n>\\n>>>>\\n>>>> This proposal was here already - maybe two years ago. The psql command\\n>>>> parser doesn't allow any complex syntax - more - the more parameters in one\\n>>>> psql commands is hard to remember, hard to read.\\n>>>>\\n>>>\\n>>> Could you please provide a link to this discussion. Probably working\\n>>> with multiple parameters in psql commands require some rework, but that's\\n>>> definitely doable.\\n>>>\\n>>\\n>> http://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/\\n>> proposal-psql-show-longest-tables/oldest\\n>> https://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvq\\n>> PE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.com\\n>>\\n>\\n> I took a look to these threads, but I didn't find place where difficulties\\n> of adding extra arguments to psql commands are pointed.\\n> Could you, please, point particular messages about it?\\n>\\n\\nI am sorry - maybe my memory doesn't serve well\\n\\nPavel\\n\\n\\n>\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n>\\n\\nThis proposal was here already - maybe two years ago. The psql command parser doesn't allow any complex syntax - more - the more parameters in one psql commands is hard to remember, hard to read.Could you please provide a link to this discussion. Probably working with multiple parameters in psql commands require some rework, but that's definitely doable.http://grokbase.com/t/postgresql/pgsql-hackers/137nt5p6s0/proposal-psql-show-longest-tables/oldesthttps://www.postgresql.org/message-id/AANLkTikyaeJ0XdKDzxSvqPE8kaRRTiUQJQHwNJ8ecN2W@mail.gmail.comI took a look to these threads, but I didn't find place where difficulties of adding extra arguments to psql commands are pointed.Could you, please, point particular messages about it?I am sorry - maybe my memory doesn't serve well Pavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Thu, 30 Mar 2017 17:49:46 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"Hi,\\n\\nOn 2017-03-11 13:06:13 +0100, Pavel Stehule wrote:\\n> 2017-03-10 15:45 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n> \\n> > On Fri, Mar 10, 2017 at 5:10 PM, Peter Eisentraut <\\n> > peter.eisentraut@2ndquadrant.com> wrote:\\n> >\\n> >> On 2/24/17 16:32, Pavel Stehule wrote:\\n> >> > set EXTENDED_DESCRIBE_SORT size_desc\\n> >> > \\\\dt+\\n> >> > \\\\l+\\n> >> > \\\\di+\\n> >> >\\n> >> > Possible variants: schema_table, table_schema, size_desc, size_asc\\n> >>\\n> >> I can see this being useful, but I think it needs to be organized a\\n> >> little better.\\n> >>\\n> >> Sort key and sort direction should be separate settings.\\n> >>\\n> >\\n> > I agree.\\n> >\\n> > I'm not sure why we need to have separate settings to sort by schema\\n> >> name and table name.\\n> >\\n> >\\n> > I think sorting by schema name, object name makes sense for people, who\\n> > have objects of same name in different schemas.\\n> >\\n> \\n> I am sending a updated version with separated sort direction in special\\n> variable\\n> \\n> There is a question. Has desc direction sense for columns like schema or\\n> table name?\\n> \\n> Using desc, asc for size is natural. But for tablename?\\n\\nI think it's pretty clear that we don't have sufficient agreement on the\\ndesign, not to speak of an implementation for an agreed upon design, to\\nget this into v10. The patch also has been submitted late in the v10\\ncycle, and has received attention. I'm therefore moving it to the next\\ncommitfest.\\n\\nRegards,\\n\\nAndres\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Wed, 5 Apr 2017 16:10:20 -0700\",\n \"msg_from\": \"Andres Freund <andres@anarazel.de>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 3/11/17 07:06, Pavel Stehule wrote:\\n> I am sending a updated version with separated sort direction in special\\n> variable\\n\\nThis patch also needs a rebase.\\n\\n-- \\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Mon, 14 Aug 2017 22:37:35 -0400\",\n \"msg_from\": \"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"Hi\\n\\n2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com\\n>:\\n\\n> On 3/11/17 07:06, Pavel Stehule wrote:\\n> > I am sending a updated version with separated sort direction in special\\n> > variable\\n>\\n> This patch also needs a rebase.\\n>\\n\\nI am sending rebased patch\\n\\nRegards\\n\\nPavel\\n\\n\\n>\\n> --\\n> Peter Eisentraut http://www.2ndQuadrant.com/\\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\",\n \"msg_date\": \"Wed, 16 Aug 2017 14:06:44 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n\\n> Hi\\n>\\n> 2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.\\n> com>:\\n>\\n>> On 3/11/17 07:06, Pavel Stehule wrote:\\n>> > I am sending a updated version with separated sort direction in special\\n>> > variable\\n>>\\n>> This patch also needs a rebase.\\n>>\\n>\\n> I am sending rebased patch\\n>\\n\\nrebased again + fix obsolete help\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n>\\n> Regards\\n>\\n> Pavel\\n>\\n>\\n>>\\n>> --\\n>> Peter Eisentraut http://www.2ndQuadrant.com/\\n>> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>>\\n>\\n>\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\",\n \"msg_date\": \"Fri, 8 Sep 2017 06:13:47 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com>\\nwrote:\\n\\n> 2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n>\\n>> Hi\\n>>\\n>> 2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.\\n>> com>:\\n>>\\n>>> On 3/11/17 07:06, Pavel Stehule wrote:\\n>>> > I am sending a updated version with separated sort direction in special\\n>>> > variable\\n>>>\\n>>> This patch also needs a rebase.\\n>>>\\n>>\\n>> I am sending rebased patch\\n>>\\n>\\n> rebased again + fix obsolete help\\n>\\n\\nFor me, patch applies cleanly, builds and passed regression tests.\\nHowever, patch misses regression tests covering added functionality.\\nPatch is definitely harmless, i.e. it doesn't affect anybody who doesn't\\nuse new functionality.\\nBut I still would prefer ordering to be options of \\\\d* commands while psql\\nvariables be defaults for those options...\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:Hi2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 3/11/17 07:06, Pavel Stehule wrote:\\n> I am sending a updated version with separated sort direction in special\\n> variable\\n\\nThis patch also needs a rebase.I am sending rebased patchrebased again + fix obsolete helpFor me, patch applies cleanly, builds and passed regression tests.However, patch misses regression tests covering added functionality.Patch is definitely harmless, i.e. it doesn't affect anybody who doesn't use new functionality.But I still would prefer ordering to be options of \\\\d* commands while psql variables be defaults for those options...------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Tue, 19 Sep 2017 17:14:20 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"Hi\\n\\n2017-09-19 16:14 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n\\n> On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com>\\n> wrote:\\n>\\n>> 2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n>>\\n>>> Hi\\n>>>\\n>>> 2017-08-15 4:37 GMT+02:00 Peter Eisentraut <\\n>>> peter.eisentraut@2ndquadrant.com>:\\n>>>\\n>>>> On 3/11/17 07:06, Pavel Stehule wrote:\\n>>>> > I am sending a updated version with separated sort direction in\\n>>>> special\\n>>>> > variable\\n>>>>\\n>>>> This patch also needs a rebase.\\n>>>>\\n>>>\\n>>> I am sending rebased patch\\n>>>\\n>>\\n>> rebased again + fix obsolete help\\n>>\\n>\\n> For me, patch applies cleanly, builds and passed regression tests.\\n> However, patch misses regression tests covering added functionality.\\n>\\n\\nI am not sure if there are any tests related to output of \\\\dt+ commands -\\nthere result is platform depend.\\n\\n\\n> Patch is definitely harmless, i.e. it doesn't affect anybody who doesn't\\n> use new functionality.\\n> But I still would prefer ordering to be options of \\\\d* commands while psql\\n> variables be defaults for those options...\\n>\\n\\nI understand\\n\\na) I don't think so commands like \\\\dt++ (or similar) is good idea - these\\ncommands should be simple how it is possible\\n\\nb) this patch doesn't block any other design - more it opens the door\\nbecause the executive part will be implemented and users can have a\\nexperience with with different output sorts - so if people will need more\\nquick change of result sort, then the work in this area will continue.\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n\\n\\n\\n\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n\\nHi2017-09-19 16:14 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:Hi2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 3/11/17 07:06, Pavel Stehule wrote:\\n> I am sending a updated version with separated sort direction in special\\n> variable\\n\\nThis patch also needs a rebase.I am sending rebased patchrebased again + fix obsolete helpFor me, patch applies cleanly, builds and passed regression tests.However, patch misses regression tests covering added functionality.I am not sure if there are any tests related to output of \\\\dt+ commands - there result is platform depend. Patch is definitely harmless, i.e. it doesn't affect anybody who doesn't use new functionality.But I still would prefer ordering to be options of \\\\d* commands while psql variables be defaults for those options...I understand a) I don't think so commands like \\\\dt++ (or similar) is good idea - these commands should be simple how it is possibleb) this patch doesn't block any other design - more it opens the door because the executive part will be implemented and users can have a experience with with different output sorts - so if people will need more quick change of result sort, then the work in this area will continue.RegardsPavel------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Tue, 19 Sep 2017 18:54:03 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Tue, Sep 19, 2017 at 7:54 PM, Pavel Stehule <pavel.stehule@gmail.com>\\nwrote:\\n\\n> 2017-09-19 16:14 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n>\\n>> On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com>\\n>> wrote:\\n>>\\n>>> 2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n>>>\\n>>>> Hi\\n>>>>\\n>>>> 2017-08-15 4:37 GMT+02:00 Peter Eisentraut <\\n>>>> peter.eisentraut@2ndquadrant.com>:\\n>>>>\\n>>>>> On 3/11/17 07:06, Pavel Stehule wrote:\\n>>>>> > I am sending a updated version with separated sort direction in\\n>>>>> special\\n>>>>> > variable\\n>>>>>\\n>>>>> This patch also needs a rebase.\\n>>>>>\\n>>>>\\n>>>> I am sending rebased patch\\n>>>>\\n>>>\\n>>> rebased again + fix obsolete help\\n>>>\\n>>\\n>> For me, patch applies cleanly, builds and passed regression tests.\\n>> However, patch misses regression tests covering added functionality.\\n>>\\n>\\n> I am not sure if there are any tests related to output of \\\\dt+ commands -\\n> there result is platform depend.\\n>\\n\\nBTW, why isn't order by name_schema available for \\\\dt? If it's available\\nwe could at least cover this case by plain regression tests.\\n\\\\dt+ could be covered by TAP tests, but it isn't yet. I think one day we\\nshould add them. However, I don't think we should force you to write them\\nin order to push this simple patch.\\n\\nPatch is definitely harmless, i.e. it doesn't affect anybody who doesn't\\n>> use new functionality.\\n>> But I still would prefer ordering to be options of \\\\d* commands while\\n>> psql variables be defaults for those options...\\n>>\\n>\\n> I understand\\n>\\n> a) I don't think so commands like \\\\dt++ (or similar) is good idea - these\\n> commands should be simple how it is possible\\n>\\n\\nI don't particularly like \\\\dt++, but second argument is probably an option.\\n\\n\\n> b) this patch doesn't block any other design - more it opens the door\\n> because the executive part will be implemented and users can have a\\n> experience with with different output sorts - so if people will need more\\n> quick change of result sort, then the work in this area will continue.\\n>\\n\\nOK. As reviewer, I'm not going to block this patch if you see its\\nfunctionality limited by just psql variables.\\nI think you should add support of name_schema \\\\dt and some regression tests\\nfor this case, before I can mark this as \\\"ready for committer\\\".\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Tue, Sep 19, 2017 at 7:54 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-09-19 16:14 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Fri, Sep 8, 2017 at 7:13 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-08-16 14:06 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:Hi2017-08-15 4:37 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 3/11/17 07:06, Pavel Stehule wrote:\\n> I am sending a updated version with separated sort direction in special\\n> variable\\n\\nThis patch also needs a rebase.I am sending rebased patchrebased again + fix obsolete helpFor me, patch applies cleanly, builds and passed regression tests.However, patch misses regression tests covering added functionality.I am not sure if there are any tests related to output of \\\\dt+ commands - there result is platform depend. BTW, why isn't order by name_schema available for \\\\dt? If it's available we could at least cover this case by plain regression tests.\\\\dt+ could be covered by TAP tests, but it isn't yet. I think one day we should add them. However, I don't think we should force you to write them in order to push this simple patch.Patch is definitely harmless, i.e. it doesn't affect anybody who doesn't use new functionality.But I still would prefer ordering to be options of \\\\d* commands while psql variables be defaults for those options...I understand a) I don't think so commands like \\\\dt++ (or similar) is good idea - these commands should be simple how it is possibleI don't particularly like \\\\dt++, but second argument is probably an option. b) this patch doesn't block any other design - more it opens the door because the executive part will be implemented and users can have a experience with with different output sorts - so if people will need more quick change of result sort, then the work in this area will continue.OK. As reviewer, I'm not going to block this patch if you see its functionality limited by just psql variables.I think you should add support of name_schema \\\\dt and some regression tests for this case, before I can mark this as \\\"ready for committer\\\".------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Tue, 19 Sep 2017 23:03:04 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 9/19/17 12:54, Pavel Stehule wrote:\\n> However, patch misses regression tests covering added functionality.\\n> \\n> I am not sure if there are any tests related to output of \\\\dt+ commands\\n> - there result is platform depend.\\n\\nHow so?\\n\\n-- \\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Wed, 20 Sep 2017 18:52:40 -0400\",\n \"msg_from\": \"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 9/8/17 00:13, Pavel Stehule wrote:\\n> I am sending rebased patch\\n> \\n> rebased again + fix obsolete help\\n\\nWhy are the variables called VERBOSE_SORT_* ? What is verbose about them?\\n\\n-- \\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Wed, 20 Sep 2017 18:53:31 -0400\",\n \"msg_from\": \"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Thu, Sep 21, 2017 at 1:52 AM, Peter Eisentraut <\\npeter.eisentraut@2ndquadrant.com> wrote:\\n\\n> On 9/19/17 12:54, Pavel Stehule wrote:\\n> > However, patch misses regression tests covering added functionality.\\n> >\\n> > I am not sure if there are any tests related to output of \\\\dt+ commands\\n> > - there result is platform depend.\\n>\\n> How so?\\n\\n\\n\\\\dt+ reports relation sizes whose are platform depended.\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Thu, Sep 21, 2017 at 1:52 AM, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:On 9/19/17 12:54, Pavel Stehule wrote:\\n> However, patch misses regression tests covering added functionality.\\n>\\n> I am not sure if there are any tests related to output of \\\\dt+ commands\\n> - there result is platform depend.\\n\\nHow so?\\\\dt+ reports relation sizes whose are platform depended.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Thu, 21 Sep 2017 11:16:52 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Thu, Sep 21, 2017 at 1:53 AM, Peter Eisentraut <\\npeter.eisentraut@2ndquadrant.com> wrote:\\n\\n> On 9/8/17 00:13, Pavel Stehule wrote:\\n> > I am sending rebased patch\\n> >\\n> > rebased again + fix obsolete help\\n>\\n> Why are the variables called VERBOSE_SORT_* ? What is verbose about them?\\n\\n\\nI assume Pavel called them so, because they are working only for \\\"verbose\\\"\\nmode of command. I.e. they are working for \\\\dt+ not \\\\dt.\\nHowever, in \\\\dt 2 of 3 sorting modes might work: schema_name and\\nname_schema. Thus, I think it worths enabling these variables for \\\"non\\nverbose\\\" mode of commands too.\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Thu, Sep 21, 2017 at 1:53 AM, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:On 9/8/17 00:13, Pavel Stehule wrote:\\n> I am sending rebased patch\\n>\\n> rebased again + fix obsolete help\\n\\nWhy are the variables called VERBOSE_SORT_* ? What is verbose about them?I assume Pavel called them so, because they are working only for \\\"verbose\\\" mode of command. I.e. they are working for \\\\dt+ not \\\\dt.However, in \\\\dt 2 of 3 sorting modes might work: schema_name and name_schema. Thus, I think it worths enabling these variables for \\\"non verbose\\\" mode of commands too.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Thu, 21 Sep 2017 11:19:30 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-09-21 10:19 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n\\n> On Thu, Sep 21, 2017 at 1:53 AM, Peter Eisentraut <\\n> peter.eisentraut@2ndquadrant.com> wrote:\\n>\\n>> On 9/8/17 00:13, Pavel Stehule wrote:\\n>> > I am sending rebased patch\\n>> >\\n>> > rebased again + fix obsolete help\\n>>\\n>> Why are the variables called VERBOSE_SORT_* ? What is verbose about them?\\n>\\n>\\n> I assume Pavel called them so, because they are working only for \\\"verbose\\\"\\n> mode of command. I.e. they are working for \\\\dt+ not \\\\dt.\\n> However, in \\\\dt 2 of 3 sorting modes might work: schema_name and\\n> name_schema. Thus, I think it worths enabling these variables for \\\"non\\n> verbose\\\" mode of commands too.\\n>\\n\\nyes. It was designed for + commands only. Can be enhanced to all commands -\\nthen VERBOSE prefix should be removed - not sure if it is necessary. For me\\ninteresting different order than default is only in verbose mode.\\n\\n\\n\\n\\n\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n\\n2017-09-21 10:19 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Thu, Sep 21, 2017 at 1:53 AM, Peter Eisentraut <peter.eisentraut@2ndquadrant.com> wrote:On 9/8/17 00:13, Pavel Stehule wrote:\\n> I am sending rebased patch\\n>\\n> rebased again + fix obsolete help\\n\\nWhy are the variables called VERBOSE_SORT_* ? What is verbose about them?I assume Pavel called them so, because they are working only for \\\"verbose\\\" mode of command. I.e. they are working for \\\\dt+ not \\\\dt.However, in \\\\dt 2 of 3 sorting modes might work: schema_name and name_schema. Thus, I think it worths enabling these variables for \\\"non verbose\\\" mode of commands too.yes. It was designed for + commands only. Can be enhanced to all commands - then VERBOSE prefix should be removed - not sure if it is necessary. For me interesting different order than default is only in verbose mode. ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Thu, 21 Sep 2017 10:27:32 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 9/21/17 04:27, Pavel Stehule wrote:\\n> yes. It was designed for + commands only. Can be enhanced to all\\n> commands - then VERBOSE prefix should be removed - not sure if it is\\n> necessary. For me interesting different order than default is only in\\n> verbose mode.\\n\\nI see where you are coming from, but there is no association in the\\nexisting UI that equates \\\"+\\\" to the word \\\"verbose\\\". I think just\\nremoving the verbose prefix and applying the sorting behavior in all\\ncases should be easier to explain and implement.\\n\\n-- \\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Thu, 21 Sep 2017 09:30:13 -0400\",\n \"msg_from\": \"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-09-21 15:30 GMT+02:00 Peter Eisentraut <\\npeter.eisentraut@2ndquadrant.com>:\\n\\n> On 9/21/17 04:27, Pavel Stehule wrote:\\n> > yes. It was designed for + commands only. Can be enhanced to all\\n> > commands - then VERBOSE prefix should be removed - not sure if it is\\n> > necessary. For me interesting different order than default is only in\\n> > verbose mode.\\n>\\n> I see where you are coming from, but there is no association in the\\n> existing UI that equates \\\"+\\\" to the word \\\"verbose\\\". I think just\\n> removing the verbose prefix and applying the sorting behavior in all\\n> cases should be easier to explain and implement.\\n>\\n\\nI though about it - but I am not sure if one kind of these variables is\\npractical.\\n\\nif I don't need a size, then sort by schema, name is ok (I didn't need any\\nelse ever). With only one kind of these variables, this setting is common -\\nwhat is not practical.\\n\\nI need sort by size in verbose mode (where size is visible) in 100% - so it\\nwill be saved to psqlrc. And when size will be invisible, then sort by size\\nis not practical, and can be messy (because size is not visible).\\n\\nSo I don't think so removing VERBOSE prefix is a good idea - or we should\\nto do different design (have not a idea how)\\n\\nRegards\\n\\nPavel\\n\\n\\n> --\\n> Peter Eisentraut http://www.2ndQuadrant.com/\\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>\\n\\n2017-09-21 15:30 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/21/17 04:27, Pavel Stehule wrote:\\n> yes. It was designed for + commands only. Can be enhanced to all\\n> commands - then VERBOSE prefix should be removed - not sure if it is\\n> necessary. For me interesting different order than default is only in\\n> verbose mode.\\n\\nI see where you are coming from, but there is no association in the\\nexisting UI that equates \\\"+\\\" to the word \\\"verbose\\\". I think just\\nremoving the verbose prefix and applying the sorting behavior in all\\ncases should be easier to explain and implement.I though about it - but I am not sure if one kind of these variables is practical.if I don't need a size, then sort by schema, name is ok (I didn't need any else ever). With only one kind of these variables, this setting is common - what is not practical.I need sort by size in verbose mode (where size is visible) in 100% - so it will be saved to psqlrc. And when size will be invisible, then sort by size is not practical, and can be messy (because size is not visible). So I don't think so removing VERBOSE prefix is a good idea - or we should to do different design (have not a idea how)RegardsPavel\\n\\n--\\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\",\n \"msg_date\": \"Thu, 21 Sep 2017 19:54:01 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 9/21/17 13:54, Pavel Stehule wrote:\\n> I see where you are coming from, but there is no association in the\\n> existing UI that equates \\\"+\\\" to the word \\\"verbose\\\". I think just\\n> removing the verbose prefix and applying the sorting behavior in all\\n> cases should be easier to explain and implement.\\n> \\n> I though about it - but I am not sure if one kind of these variables is\\n> practical.\\n> \\n> if I don't need a size, then sort by schema, name is ok (I didn't need\\n> any else ever). With only one kind of these variables, this setting is\\n> common - what is not practical.\\n\\nBut you are proposing also to add a variable configuring the sort\\ndirection. It would be weird that \\\\dX+ observed the sort direction but\\n\\\\dX did not.\\n\\n-- \\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Thu, 21 Sep 2017 14:20:56 -0400\",\n \"msg_from\": \"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-09-21 20:20 GMT+02:00 Peter Eisentraut <\\npeter.eisentraut@2ndquadrant.com>:\\n\\n> On 9/21/17 13:54, Pavel Stehule wrote:\\n> > I see where you are coming from, but there is no association in the\\n> > existing UI that equates \\\"+\\\" to the word \\\"verbose\\\". I think just\\n> > removing the verbose prefix and applying the sorting behavior in all\\n> > cases should be easier to explain and implement.\\n> >\\n> > I though about it - but I am not sure if one kind of these variables is\\n> > practical.\\n> >\\n> > if I don't need a size, then sort by schema, name is ok (I didn't need\\n> > any else ever). With only one kind of these variables, this setting is\\n> > common - what is not practical.\\n>\\n> But you are proposing also to add a variable configuring the sort\\n> direction. It would be weird that \\\\dX+ observed the sort direction but\\n> \\\\dX did not.\\n>\\n\\nyes and no.\\n\\nschema_name, name_schema or SORT_DIRECTION has sense for both type of\\ncommands.\\n\\nsize sort has sense only for \\\\dX+ command.\\n\\nI am thinking about solution and the most clean I see two distinct\\nvariables:\\n\\nSORT_COLUMNS and VERBOSE_SORT_COLUMNS\\n\\nwhen VERBOSE_SORT_COLUMNS will be undefined, then SORT_COLUMNS is used for\\n\\\\dX+ command too.\\n\\nIs it acceptable?\\n\\n\\n\\n\\n> --\\n> Peter Eisentraut http://www.2ndQuadrant.com/\\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>\\n\\n2017-09-21 20:20 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/21/17 13:54, Pavel Stehule wrote:\\n> I see where you are coming from, but there is no association in the\\n> existing UI that equates \\\"+\\\" to the word \\\"verbose\\\". I think just\\n> removing the verbose prefix and applying the sorting behavior in all\\n> cases should be easier to explain and implement.\\n>\\n> I though about it - but I am not sure if one kind of these variables is\\n> practical.\\n>\\n> if I don't need a size, then sort by schema, name is ok (I didn't need\\n> any else ever). With only one kind of these variables, this setting is\\n> common - what is not practical.\\n\\nBut you are proposing also to add a variable configuring the sort\\ndirection. It would be weird that \\\\dX+ observed the sort direction but\\n\\\\dX did not.yes and no. schema_name, name_schema or SORT_DIRECTION has sense for both type of commands.size sort has sense only for \\\\dX+ command.I am thinking about solution and the most clean I see two distinct variables:SORT_COLUMNS and VERBOSE_SORT_COLUMNS when VERBOSE_SORT_COLUMNS will be undefined, then SORT_COLUMNS is used for \\\\dX+ command too.Is it acceptable?\\n\\n--\\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\",\n \"msg_date\": \"Thu, 21 Sep 2017 20:30:19 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-09-21 20:30 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n\\n>\\n>\\n> 2017-09-21 20:20 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.\\n> com>:\\n>\\n>> On 9/21/17 13:54, Pavel Stehule wrote:\\n>> > I see where you are coming from, but there is no association in the\\n>> > existing UI that equates \\\"+\\\" to the word \\\"verbose\\\". I think just\\n>> > removing the verbose prefix and applying the sorting behavior in all\\n>> > cases should be easier to explain and implement.\\n>> >\\n>> > I though about it - but I am not sure if one kind of these variables is\\n>> > practical.\\n>> >\\n>> > if I don't need a size, then sort by schema, name is ok (I didn't need\\n>> > any else ever). With only one kind of these variables, this setting is\\n>> > common - what is not practical.\\n>>\\n>> But you are proposing also to add a variable configuring the sort\\n>> direction. It would be weird that \\\\dX+ observed the sort direction but\\n>> \\\\dX did not.\\n>>\\n>\\n> yes and no.\\n>\\n> schema_name, name_schema or SORT_DIRECTION has sense for both type of\\n> commands.\\n>\\n> size sort has sense only for \\\\dX+ command.\\n>\\n> I am thinking about solution and the most clean I see two distinct\\n> variables:\\n>\\n> SORT_COLUMNS and VERBOSE_SORT_COLUMNS\\n>\\n> when VERBOSE_SORT_COLUMNS will be undefined, then SORT_COLUMNS is used for\\n> \\\\dX+ command too.\\n>\\n> Is it acceptable?\\n>\\n\\nI though more about it, and I am thinking so this direction is not good.\\n\\nExample: somebody set SORT_COLUMNS to schema_name value. This is nonsense\\nfor \\\\l command\\n\\nNow, I am thinking so more correct and practical design is based on special\\nmode, activated by variable\\n\\nPREFER_SIZE_SORT .. (off, asc, desc)\\n\\nThis has sense for wide group of commands that can show size. And when size\\nis not visible, then this option is not active.\\n\\nWhat do you think about this proposal?\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n>\\n>\\n>\\n>> --\\n>> Peter Eisentraut http://www.2ndQuadrant.com/\\n>> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>>\\n>\\n>\\n\\n2017-09-21 20:30 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:2017-09-21 20:20 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/21/17 13:54, Pavel Stehule wrote:\\n> I see where you are coming from, but there is no association in the\\n> existing UI that equates \\\"+\\\" to the word \\\"verbose\\\". I think just\\n> removing the verbose prefix and applying the sorting behavior in all\\n> cases should be easier to explain and implement.\\n>\\n> I though about it - but I am not sure if one kind of these variables is\\n> practical.\\n>\\n> if I don't need a size, then sort by schema, name is ok (I didn't need\\n> any else ever). With only one kind of these variables, this setting is\\n> common - what is not practical.\\n\\nBut you are proposing also to add a variable configuring the sort\\ndirection. It would be weird that \\\\dX+ observed the sort direction but\\n\\\\dX did not.yes and no. schema_name, name_schema or SORT_DIRECTION has sense for both type of commands.size sort has sense only for \\\\dX+ command.I am thinking about solution and the most clean I see two distinct variables:SORT_COLUMNS and VERBOSE_SORT_COLUMNS when VERBOSE_SORT_COLUMNS will be undefined, then SORT_COLUMNS is used for \\\\dX+ command too.Is it acceptable?I though more about it, and I am thinking so this direction is not good. Example: somebody set SORT_COLUMNS to schema_name value. This is nonsense for \\\\l commandNow, I am thinking so more correct and practical design is based on special mode, activated by variablePREFER_SIZE_SORT .. (off, asc, desc)This has sense for wide group of commands that can show size. And when size is not visible, then this option is not active.What do you think about this proposal?RegardsPavel \\n\\n--\\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\",\n \"msg_date\": \"Fri, 22 Sep 2017 15:16:35 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 9/22/17 09:16, Pavel Stehule wrote:\\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\\n> nonsense for \\\\l command\\n> \\n> Now, I am thinking so more correct and practical design is based on\\n> special mode, activated by variable\\n> \\n> PREFER_SIZE_SORT .. (off, asc, desc)\\n> \\n> This has sense for wide group of commands that can show size. And when\\n> size is not visible, then this option is not active.\\n\\nMaybe this shouldn't be a variable at all. It's not like you'll set\\nthis as a global preference. You probably want it for one command only.\\n So a per-command option might make more sense.\\n\\n-- \\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Fri, 22 Sep 2017 15:12:29 -0400\",\n \"msg_from\": \"Peter Eisentraut <peter.eisentraut@2ndquadrant.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-09-22 21:12 GMT+02:00 Peter Eisentraut <\\npeter.eisentraut@2ndquadrant.com>:\\n\\n> On 9/22/17 09:16, Pavel Stehule wrote:\\n> > Example: somebody set SORT_COLUMNS to schema_name value. This is\\n> > nonsense for \\\\l command\\n> >\\n> > Now, I am thinking so more correct and practical design is based on\\n> > special mode, activated by variable\\n> >\\n> > PREFER_SIZE_SORT .. (off, asc, desc)\\n> >\\n> > This has sense for wide group of commands that can show size. And when\\n> > size is not visible, then this option is not active.\\n>\\n> Maybe this shouldn't be a variable at all. It's not like you'll set\\n> this as a global preference. You probably want it for one command only.\\n> So a per-command option might make more sense.\\n>\\n\\nSure, I cannot to know, what users will do. But, when I need to see a size\\nof objects, then I prefer the sort by size desc every time. If I need to\\nfind some object, then I can to use a searching in pager. So in my case,\\nthis settings will be in psqlrc. In GoodData we used years own\\ncustomization - the order by size was hardcoded and nobody reported me any\\nissue.\\n\\nAlexander proposed some per command option, but current syntax of psql\\ncommands don't allows some simple parametrization. If it can be user\\nfriendly, then it should be short. From implementation perspective, it\\nshould be simply parsed. It should be intuitive too - too much symbols\\ntogether is not good idea.\\n\\nMaybe some prefix design - but it is not design for common people (although\\nthese people don't use psql usually)\\n\\n'\\\\sort size \\\\dt ?\\n\\n\\\\dt:sort_by_size\\n\\\\dt+:sort_by_size ?\\n\\nI don't see any good design in this direction\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n\\n\\n\\n\\n>\\n> --\\n> Peter Eisentraut http://www.2ndQuadrant.com/\\n> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>\\n\\n2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/22/17 09:16, Pavel Stehule wrote:\\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\\n> nonsense for \\\\l command\\n>\\n> Now, I am thinking so more correct and practical design is based on\\n> special mode, activated by variable\\n>\\n> PREFER_SIZE_SORT .. (off, asc, desc)\\n>\\n> This has sense for wide group of commands that can show size. And when\\n> size is not visible, then this option is not active.\\n\\nMaybe this shouldn't be a variable at all. It's not like you'll set\\nthis as a global preference. You probably want it for one command only.\\n So a per-command option might make more sense.Sure, I cannot to know, what users will do. But, when I need to see a size of objects, then I prefer the sort by size desc every time. If I need to find some object, then I can to use a searching in pager. So in my case, this settings will be in psqlrc. In GoodData we used years own customization - the order by size was hardcoded and nobody reported me any issue. Alexander proposed some per command option, but current syntax of psql commands don't allows some simple parametrization. If it can be user friendly, then it should be short. From implementation perspective, it should be simply parsed. It should be intuitive too - too much symbols together is not good idea.Maybe some prefix design - but it is not design for common people (although these people don't use psql usually)'\\\\sort size \\\\dt ?\\\\dt:sort_by_size \\\\dt+:sort_by_size ?I don't see any good design in this direction RegardsPavel \\n\\n--\\nPeter Eisentraut http://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\",\n \"msg_date\": \"Fri, 22 Sep 2017 21:31:12 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"Hi\\n\\n2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n\\n>\\n>\\n> 2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.\\n> com>:\\n>\\n>> On 9/22/17 09:16, Pavel Stehule wrote:\\n>> > Example: somebody set SORT_COLUMNS to schema_name value. This is\\n>> > nonsense for \\\\l command\\n>> >\\n>> > Now, I am thinking so more correct and practical design is based on\\n>> > special mode, activated by variable\\n>> >\\n>> > PREFER_SIZE_SORT .. (off, asc, desc)\\n>> >\\n>> > This has sense for wide group of commands that can show size. And when\\n>> > size is not visible, then this option is not active.\\n>>\\n>> Maybe this shouldn't be a variable at all. It's not like you'll set\\n>> this as a global preference. You probably want it for one command only.\\n>> So a per-command option might make more sense.\\n>>\\n>\\n> Sure, I cannot to know, what users will do. But, when I need to see a size\\n> of objects, then I prefer the sort by size desc every time. If I need to\\n> find some object, then I can to use a searching in pager. So in my case,\\n> this settings will be in psqlrc. In GoodData we used years own\\n> customization - the order by size was hardcoded and nobody reported me any\\n> issue.\\n>\\n> Alexander proposed some per command option, but current syntax of psql\\n> commands don't allows some simple parametrization. If it can be user\\n> friendly, then it should be short. From implementation perspective, it\\n> should be simply parsed. It should be intuitive too - too much symbols\\n> together is not good idea.\\n>\\n> Maybe some prefix design - but it is not design for common people\\n> (although these people don't use psql usually)\\n>\\n> '\\\\sort size \\\\dt ?\\n>\\n> \\\\dt:sort_by_size\\n> \\\\dt+:sort_by_size ?\\n>\\n> I don't see any good design in this direction\\n>\\n>\\nI though about Alexander proposal, and I am thinking so it can be probably\\nbest if we respect psql design. I implemented two command suffixes\\n(supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n\\nso list of tables can be sorted with commands:\\n\\n\\\\dt+sd (in this case, the order is not strict), so command\\n\\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n\\nThese two chars are acceptable. Same principle is used for \\\\l command\\n\\n\\\\lsd+ or \\\\l+sd\\n\\nWhat do you think about it?\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n> Regards\\n>\\n> Pavel\\n>\\n>\\n>\\n>\\n>\\n>\\n>\\n>>\\n>> --\\n>> Peter Eisentraut http://www.2ndQuadrant.com/\\n>> PostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n>>\\n>\\n>\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\",\n \"msg_date\": \"Sat, 28 Oct 2017 14:46:46 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com>\\nwrote:\\n\\n> 2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n>\\n>>\\n>>\\n>> 2017-09-22 21:12 GMT+02:00 Peter Eisentraut <\\n>> peter.eisentraut@2ndquadrant.com>:\\n>>\\n>>> On 9/22/17 09:16, Pavel Stehule wrote:\\n>>> > Example: somebody set SORT_COLUMNS to schema_name value. This is\\n>>> > nonsense for \\\\l command\\n>>> >\\n>>> > Now, I am thinking so more correct and practical design is based on\\n>>> > special mode, activated by variable\\n>>> >\\n>>> > PREFER_SIZE_SORT .. (off, asc, desc)\\n>>> >\\n>>> > This has sense for wide group of commands that can show size. And when\\n>>> > size is not visible, then this option is not active.\\n>>>\\n>>> Maybe this shouldn't be a variable at all. It's not like you'll set\\n>>> this as a global preference. You probably want it for one command only.\\n>>> So a per-command option might make more sense.\\n>>>\\n>>\\n>> Sure, I cannot to know, what users will do. But, when I need to see a\\n>> size of objects, then I prefer the sort by size desc every time. If I need\\n>> to find some object, then I can to use a searching in pager. So in my case,\\n>> this settings will be in psqlrc. In GoodData we used years own\\n>> customization - the order by size was hardcoded and nobody reported me any\\n>> issue.\\n>>\\n>> Alexander proposed some per command option, but current syntax of psql\\n>> commands don't allows some simple parametrization. If it can be user\\n>> friendly, then it should be short. From implementation perspective, it\\n>> should be simply parsed. It should be intuitive too - too much symbols\\n>> together is not good idea.\\n>>\\n>> Maybe some prefix design - but it is not design for common people\\n>> (although these people don't use psql usually)\\n>>\\n>> '\\\\sort size \\\\dt ?\\n>>\\n>> \\\\dt:sort_by_size\\n>> \\\\dt+:sort_by_size ?\\n>>\\n>> I don't see any good design in this direction\\n>>\\n>>\\n> I though about Alexander proposal, and I am thinking so it can be probably\\n> best if we respect psql design. I implemented two command suffixes\\n> (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n>\\n> so list of tables can be sorted with commands:\\n>\\n> \\\\dt+sd (in this case, the order is not strict), so command\\n> \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>\\n> These two chars are acceptable. Same principle is used for \\\\l command\\n>\\n> \\\\lsd+ or \\\\l+sd\\n>\\n> What do you think about it?\\n>\\n\\nI think \\\\lsd+ command would be another postgres meme :)\\nBTW, are you going to provide an ability to sort by name, schema?\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/22/17 09:16, Pavel Stehule wrote:\\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\\n> nonsense for \\\\l command\\n>\\n> Now, I am thinking so more correct and practical design is based on\\n> special mode, activated by variable\\n>\\n> PREFER_SIZE_SORT .. (off, asc, desc)\\n>\\n> This has sense for wide group of commands that can show size. And when\\n> size is not visible, then this option is not active.\\n\\nMaybe this shouldn't be a variable at all. It's not like you'll set\\nthis as a global preference. You probably want it for one command only.\\n So a per-command option might make more sense.Sure, I cannot to know, what users will do. But, when I need to see a size of objects, then I prefer the sort by size desc every time. If I need to find some object, then I can to use a searching in pager. So in my case, this settings will be in psqlrc. In GoodData we used years own customization - the order by size was hardcoded and nobody reported me any issue. Alexander proposed some per command option, but current syntax of psql commands don't allows some simple parametrization. If it can be user friendly, then it should be short. From implementation perspective, it should be simply parsed. It should be intuitive too - too much symbols together is not good idea.Maybe some prefix design - but it is not design for common people (although these people don't use psql usually)'\\\\sort size \\\\dt ?\\\\dt:sort_by_size \\\\dt+:sort_by_size ?I don't see any good design in this direction I though about Alexander proposal, and I am thinking so it can be probably best if we respect psql design. I implemented two command suffixes (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descentso list of tables can be sorted with commands:\\\\dt+sd (in this case, the order is not strict), so command\\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)These two chars are acceptable. Same principle is used for \\\\l command\\\\lsd+ or \\\\l+sdWhat do you think about it?I think \\\\lsd+ command would be another postgres meme :)BTW, are you going to provide an ability to sort by name, schema?------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Sun, 29 Oct 2017 00:35:56 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-10-28 23:35 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n\\n> On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com>\\n> wrote:\\n>\\n>> 2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n>>\\n>>>\\n>>>\\n>>> 2017-09-22 21:12 GMT+02:00 Peter Eisentraut <\\n>>> peter.eisentraut@2ndquadrant.com>:\\n>>>\\n>>>> On 9/22/17 09:16, Pavel Stehule wrote:\\n>>>> > Example: somebody set SORT_COLUMNS to schema_name value. This is\\n>>>> > nonsense for \\\\l command\\n>>>> >\\n>>>> > Now, I am thinking so more correct and practical design is based on\\n>>>> > special mode, activated by variable\\n>>>> >\\n>>>> > PREFER_SIZE_SORT .. (off, asc, desc)\\n>>>> >\\n>>>> > This has sense for wide group of commands that can show size. And when\\n>>>> > size is not visible, then this option is not active.\\n>>>>\\n>>>> Maybe this shouldn't be a variable at all. It's not like you'll set\\n>>>> this as a global preference. You probably want it for one command only.\\n>>>> So a per-command option might make more sense.\\n>>>>\\n>>>\\n>>> Sure, I cannot to know, what users will do. But, when I need to see a\\n>>> size of objects, then I prefer the sort by size desc every time. If I need\\n>>> to find some object, then I can to use a searching in pager. So in my case,\\n>>> this settings will be in psqlrc. In GoodData we used years own\\n>>> customization - the order by size was hardcoded and nobody reported me any\\n>>> issue.\\n>>>\\n>>> Alexander proposed some per command option, but current syntax of psql\\n>>> commands don't allows some simple parametrization. If it can be user\\n>>> friendly, then it should be short. From implementation perspective, it\\n>>> should be simply parsed. It should be intuitive too - too much symbols\\n>>> together is not good idea.\\n>>>\\n>>> Maybe some prefix design - but it is not design for common people\\n>>> (although these people don't use psql usually)\\n>>>\\n>>> '\\\\sort size \\\\dt ?\\n>>>\\n>>> \\\\dt:sort_by_size\\n>>> \\\\dt+:sort_by_size ?\\n>>>\\n>>> I don't see any good design in this direction\\n>>>\\n>>>\\n>> I though about Alexander proposal, and I am thinking so it can be\\n>> probably best if we respect psql design. I implemented two command suffixes\\n>> (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n>>\\n>> so list of tables can be sorted with commands:\\n>>\\n>> \\\\dt+sd (in this case, the order is not strict), so command\\n>> \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>>\\n>> These two chars are acceptable. Same principle is used for \\\\l command\\n>>\\n>> \\\\lsd+ or \\\\l+sd\\n>>\\n>> What do you think about it?\\n>>\\n>\\n> I think \\\\lsd+ command would be another postgres meme :)\\n> BTW, are you going to provide an ability to sort by name, schema?\\n>\\n\\nIt has sense only for tables - probably only \\\\dtn \\\"n\\\" like name\\n\\n\\n\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n\\n2017-10-28 23:35 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/22/17 09:16, Pavel Stehule wrote:\\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\\n> nonsense for \\\\l command\\n>\\n> Now, I am thinking so more correct and practical design is based on\\n> special mode, activated by variable\\n>\\n> PREFER_SIZE_SORT .. (off, asc, desc)\\n>\\n> This has sense for wide group of commands that can show size. And when\\n> size is not visible, then this option is not active.\\n\\nMaybe this shouldn't be a variable at all. It's not like you'll set\\nthis as a global preference. You probably want it for one command only.\\n So a per-command option might make more sense.Sure, I cannot to know, what users will do. But, when I need to see a size of objects, then I prefer the sort by size desc every time. If I need to find some object, then I can to use a searching in pager. So in my case, this settings will be in psqlrc. In GoodData we used years own customization - the order by size was hardcoded and nobody reported me any issue. Alexander proposed some per command option, but current syntax of psql commands don't allows some simple parametrization. If it can be user friendly, then it should be short. From implementation perspective, it should be simply parsed. It should be intuitive too - too much symbols together is not good idea.Maybe some prefix design - but it is not design for common people (although these people don't use psql usually)'\\\\sort size \\\\dt ?\\\\dt:sort_by_size \\\\dt+:sort_by_size ?I don't see any good design in this direction I though about Alexander proposal, and I am thinking so it can be probably best if we respect psql design. I implemented two command suffixes (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descentso list of tables can be sorted with commands:\\\\dt+sd (in this case, the order is not strict), so command\\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)These two chars are acceptable. Same principle is used for \\\\l command\\\\lsd+ or \\\\l+sdWhat do you think about it?I think \\\\lsd+ command would be another postgres meme :)BTW, are you going to provide an ability to sort by name, schema?It has sense only for tables - probably only \\\\dtn \\\"n\\\" like name ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Sat, 28 Oct 2017 23:47:27 +0200\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Sun, Oct 29, 2017 at 12:47 AM, Pavel Stehule <pavel.stehule@gmail.com>\\nwrote:\\n\\n> 2017-10-28 23:35 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n>\\n>> On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com>\\n>> wrote:\\n>>\\n>>> 2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:\\n>>>\\n>>>>\\n>>>>\\n>>>> 2017-09-22 21:12 GMT+02:00 Peter Eisentraut <\\n>>>> peter.eisentraut@2ndquadrant.com>:\\n>>>>\\n>>>>> On 9/22/17 09:16, Pavel Stehule wrote:\\n>>>>> > Example: somebody set SORT_COLUMNS to schema_name value. This is\\n>>>>> > nonsense for \\\\l command\\n>>>>> >\\n>>>>> > Now, I am thinking so more correct and practical design is based on\\n>>>>> > special mode, activated by variable\\n>>>>> >\\n>>>>> > PREFER_SIZE_SORT .. (off, asc, desc)\\n>>>>> >\\n>>>>> > This has sense for wide group of commands that can show size. And\\n>>>>> when\\n>>>>> > size is not visible, then this option is not active.\\n>>>>>\\n>>>>> Maybe this shouldn't be a variable at all. It's not like you'll set\\n>>>>> this as a global preference. You probably want it for one command\\n>>>>> only.\\n>>>>> So a per-command option might make more sense.\\n>>>>>\\n>>>>\\n>>>> Sure, I cannot to know, what users will do. But, when I need to see a\\n>>>> size of objects, then I prefer the sort by size desc every time. If I need\\n>>>> to find some object, then I can to use a searching in pager. So in my case,\\n>>>> this settings will be in psqlrc. In GoodData we used years own\\n>>>> customization - the order by size was hardcoded and nobody reported me any\\n>>>> issue.\\n>>>>\\n>>>> Alexander proposed some per command option, but current syntax of psql\\n>>>> commands don't allows some simple parametrization. If it can be user\\n>>>> friendly, then it should be short. From implementation perspective, it\\n>>>> should be simply parsed. It should be intuitive too - too much symbols\\n>>>> together is not good idea.\\n>>>>\\n>>>> Maybe some prefix design - but it is not design for common people\\n>>>> (although these people don't use psql usually)\\n>>>>\\n>>>> '\\\\sort size \\\\dt ?\\n>>>>\\n>>>> \\\\dt:sort_by_size\\n>>>> \\\\dt+:sort_by_size ?\\n>>>>\\n>>>> I don't see any good design in this direction\\n>>>>\\n>>>>\\n>>> I though about Alexander proposal, and I am thinking so it can be\\n>>> probably best if we respect psql design. I implemented two command suffixes\\n>>> (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n>>>\\n>>> so list of tables can be sorted with commands:\\n>>>\\n>>> \\\\dt+sd (in this case, the order is not strict), so command\\n>>> \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>>>\\n>>> These two chars are acceptable. Same principle is used for \\\\l command\\n>>>\\n>>> \\\\lsd+ or \\\\l+sd\\n>>>\\n>>> What do you think about it?\\n>>>\\n>>\\n>> I think \\\\lsd+ command would be another postgres meme :)\\n>> BTW, are you going to provide an ability to sort by name, schema?\\n>>\\n>\\n> It has sense only for tables - probably only \\\\dtn \\\"n\\\" like name\\n>\\n\\nIn general, this approach looks good for me.\\nRegarding current state of patch, I'd like to see new options documented.\\nAlso, it would be better to replace \\\"bool sort_size\\\" with enum assuming\\nthere could be other sorting orders in future.\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Sun, Oct 29, 2017 at 12:47 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-10-28 23:35 GMT+02:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Sat, Oct 28, 2017 at 3:46 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:2017-09-22 21:31 GMT+02:00 Pavel Stehule <pavel.stehule@gmail.com>:2017-09-22 21:12 GMT+02:00 Peter Eisentraut <peter.eisentraut@2ndquadrant.com>:On 9/22/17 09:16, Pavel Stehule wrote:\\n> Example: somebody set SORT_COLUMNS to schema_name value. This is\\n> nonsense for \\\\l command\\n>\\n> Now, I am thinking so more correct and practical design is based on\\n> special mode, activated by variable\\n>\\n> PREFER_SIZE_SORT .. (off, asc, desc)\\n>\\n> This has sense for wide group of commands that can show size. And when\\n> size is not visible, then this option is not active.\\n\\nMaybe this shouldn't be a variable at all. It's not like you'll set\\nthis as a global preference. You probably want it for one command only.\\n So a per-command option might make more sense.Sure, I cannot to know, what users will do. But, when I need to see a size of objects, then I prefer the sort by size desc every time. If I need to find some object, then I can to use a searching in pager. So in my case, this settings will be in psqlrc. In GoodData we used years own customization - the order by size was hardcoded and nobody reported me any issue. Alexander proposed some per command option, but current syntax of psql commands don't allows some simple parametrization. If it can be user friendly, then it should be short. From implementation perspective, it should be simply parsed. It should be intuitive too - too much symbols together is not good idea.Maybe some prefix design - but it is not design for common people (although these people don't use psql usually)'\\\\sort size \\\\dt ?\\\\dt:sort_by_size \\\\dt+:sort_by_size ?I don't see any good design in this direction I though about Alexander proposal, and I am thinking so it can be probably best if we respect psql design. I implemented two command suffixes (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descentso list of tables can be sorted with commands:\\\\dt+sd (in this case, the order is not strict), so command\\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)These two chars are acceptable. Same principle is used for \\\\l command\\\\lsd+ or \\\\l+sdWhat do you think about it?I think \\\\lsd+ command would be another postgres meme :)BTW, are you going to provide an ability to sort by name, schema?It has sense only for tables - probably only \\\\dtn \\\"n\\\" like nameIn general, this approach looks good for me.Regarding current state of patch, I'd like to see new options documented. Also, it would be better to replace \\\"bool sort_size\\\" with enum assuming there could be other sorting orders in future.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Tue, 31 Oct 2017 11:26:52 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"Hi\\n\\n\\nIn general, this approach looks good for me.\\n> Regarding current state of patch, I'd like to see new options documented.\\n> Also, it would be better to replace \\\"bool sort_size\\\" with enum assuming\\n> there could be other sorting orders in future.\\n>\\n\\nI am sending updated patch with some basic doc\\n\\nRegards\\n\\nPavel\\n\\n>\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\",\n \"msg_date\": \"Wed, 1 Nov 2017 12:44:31 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com> wrote:\\n> I though about Alexander proposal, and I am thinking so it can be probably\\n> best if we respect psql design. I implemented two command suffixes\\n> (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n>\\n> so list of tables can be sorted with commands:\\n>\\n> \\\\dt+sd (in this case, the order is not strict), so command\\n> \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>\\n> These two chars are acceptable. Same principle is used for \\\\l command\\n>\\n> \\\\lsd+ or \\\\l+sd\\n>\\n> What do you think about it?\\n>\\n\\nI really hate that syntax. This is going to turn into an\\nincomprehensible mess, and isn't easily extended to support other\\noptions.\\n\\nI agree with people who have said they would prefer this to be\\navailable as a per-command option rather than as a variable that you\\nhave to set, but it needs a clearer syntax. I actually like Stephen's\\nidea of using a user-defined SQL snippet, because that's a familiar\\nsyntax to people, and it avoids adding an ever-increasing number of\\noptions to these commands. Instead, the syntax could simply be:\\n\\n\\\\d[S+] [ pattern ] [ ( auxiliary sql ) ]\\n\\n(and similar for the other commands)\\n\\nThe auxiliary SQL could be pretty much anything to allow user-defined\\nordering and filtering.\\n\\nI think parsing the optional auxiliary SQL snippet in parentheses at\\nthe end should be quite straightforward, provided that psql makes no\\nattempt to actually parse the SQL contained in the parentheses -- it\\nshould just add it to the SQL it sends to the backend (like \\\\copy\\ndoes). For example, for \\\\d+, instead of generating\\n\\nSELECT n.nspname as \\\"Schema\\\",\\n c.relname as \\\"Name\\\",\\n CASE c.relkind WHEN 'r' THEN 'table' WHEN 'v' THEN 'view' WHEN 'm'\\nTHEN 'materialized view' WHEN 'i' THEN 'index' WHEN 'S' THEN\\n'sequence' WHEN 's' THEN 'special' WHEN 'f' THEN 'foreign table' WHEN\\n'p' THEN 'table' END as \\\"Type\\\",\\n pg_catalog.pg_get_userbyid(c.relowner) as \\\"Owner\\\",\\n pg_catalog.pg_size_pretty(pg_catalog.pg_table_size(c.oid)) as \\\"Size\\\",\\n pg_catalog.obj_description(c.oid, 'pg_class') as \\\"Description\\\"\\nFROM pg_catalog.pg_class c\\n LEFT JOIN pg_catalog.pg_namespace n ON n.oid = c.relnamespace\\nWHERE c.relkind IN ('r','p','v','m','S','f','')\\n AND n.nspname <> 'pg_catalog'\\n AND n.nspname <> 'information_schema'\\n AND n.nspname !~ '^pg_toast'\\n AND pg_catalog.pg_table_is_visible(c.oid)\\nORDER BY 1,2;\\n\\nwe could generate\\n\\nSELECT schema as \\\"Schema\\\",\\n name as \\\"Name\\\",\\n type as \\\"Type\\\",\\n owner as \\\"Owner\\\",\\n pg_catalog.pg_size_pretty(size) as \\\"Size\\\",\\n description as \\\"Description\\\"\\nFROM (\\nSELECT n.nspname as schema,\\n c.relname as name,\\n CASE c.relkind WHEN 'r' THEN 'table' WHEN 'v' THEN 'view' WHEN 'm'\\nTHEN 'materialized view' WHEN 'i' THEN 'index' WHEN 'S' THEN\\n'sequence' WHEN 's' THEN 'special' WHEN 'f' THEN 'foreign table' WHEN\\n'p' THEN 'table' END as type,\\n pg_catalog.pg_get_userbyid(c.relowner) as owner,\\n pg_catalog.pg_table_size(c.oid) as size,\\n pg_catalog.obj_description(c.oid, 'pg_class') as description\\nFROM pg_catalog.pg_class c\\n LEFT JOIN pg_catalog.pg_namespace n ON n.oid = c.relnamespace\\nWHERE c.relkind IN ('r','p','v','m','S','f','')\\n AND n.nspname <> 'pg_catalog'\\n AND n.nspname <> 'information_schema'\\n AND n.nspname !~ '^pg_toast'\\n AND pg_catalog.pg_table_is_visible(c.oid)\\nORDER BY 1,2\\n) as t\\n<auxiliary sql>;\\n\\nThat would allow things like\\n\\n\\\\d+ (order by size)\\n\\nwhich would sort by the numeric size, while displaying the pretty size\\nin the output.\\n\\nThis would also allow more complex orderings that would be hard to\\nachieve any other way, such as\\n\\n\\\\d+ (order by type, schema, size desc)\\n\\\\dt (order by pg_total_relation_size(name::regclass))\\n\\n(note the size reported by \\\\d+ is not the total relation size, because\\nit excludes indexes)\\n\\nAlso, it would allow user-defined WHERE clauses to filter the results\\nshown, for example:\\n\\n\\\\d+ (where size > pg_size_bytes('1GB'))\\n\\\\dv (where pg_relation_is_updatable(name::regclass, true) != 0)\\n\\nand many more things are possible, without needing to learn any new\\nsyntax, and without needing to keep adding more and more options to\\nthe psql syntax.\\n\\nRegards,\\nDean\\n\\n\",\n \"msg_date\": \"Mon, 13 Nov 2017 14:17:22 +0000\",\n \"msg_from\": \"Dean Rasheed <dean.a.rasheed@gmail.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com>\\nwrote:\\n\\n> On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com>\\n> wrote:\\n> > I though about Alexander proposal, and I am thinking so it can be\\n> probably\\n> > best if we respect psql design. I implemented two command suffixes\\n> > (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n> >\\n> > so list of tables can be sorted with commands:\\n> >\\n> > \\\\dt+sd (in this case, the order is not strict), so command\\n> > \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n> >\\n> > These two chars are acceptable. Same principle is used for \\\\l command\\n> >\\n> > \\\\lsd+ or \\\\l+sd\\n> >\\n> > What do you think about it?\\n> >\\n>\\n> I really hate that syntax. This is going to turn into an\\n> incomprehensible mess, and isn't easily extended to support other\\n> options.\\n>\\n\\n+1. While useful in itself, I think it's definitely a dangerous pattern to\\ngo down, as it'll only get worse.\\n\\n\\nI agree with people who have said they would prefer this to be\\n> available as a per-command option rather than as a variable that you\\n> have to set, but it needs a clearer syntax. I actually like Stephen's\\n> idea of using a user-defined SQL snippet, because that's a familiar\\n> syntax to people, and it avoids adding an ever-increasing number of\\n> options to these commands. Instead, the syntax could simply be:\\n>\\n\\n+1 here as well. And anybody who is actually going to need this level of\\ncontrol definitely will know SQL...\\n\\nAnd if one wants to save some \\\"standard patterns\\\", it should be doable to\\nsave the pattern itself in a variable and then use it with something like\\n\\\"\\\\dt :mysort\\\" and have it expand the normal way there.\\n\\n-- \\n Magnus Hagander\\n Me: https://www.hagander.net/ <http://www.hagander.net/>\\n Work: https://www.redpill-linpro.com/ <http://www.redpill-linpro.com/>\\n\\nOn Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com> wrote:On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com> wrote:\\n> I though about Alexander proposal, and I am thinking so it can be probably\\n> best if we respect psql design. I implemented two command suffixes\\n> (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n>\\n> so list of tables can be sorted with commands:\\n>\\n> \\\\dt+sd (in this case, the order is not strict), so command\\n> \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>\\n> These two chars are acceptable. Same principle is used for \\\\l command\\n>\\n> \\\\lsd+ or \\\\l+sd\\n>\\n> What do you think about it?\\n>\\n\\nI really hate that syntax. This is going to turn into an\\nincomprehensible mess, and isn't easily extended to support other\\noptions.+1. While useful in itself, I think it's definitely a dangerous pattern to go down, as it'll only get worse.\\nI agree with people who have said they would prefer this to be\\navailable as a per-command option rather than as a variable that you\\nhave to set, but it needs a clearer syntax. I actually like Stephen's\\nidea of using a user-defined SQL snippet, because that's a familiar\\nsyntax to people, and it avoids adding an ever-increasing number of\\noptions to these commands. Instead, the syntax could simply be:+1 here as well. And anybody who is actually going to need this level of control definitely will know SQL...And if one wants to save some \\\"standard patterns\\\", it should be doable to save the pattern itself in a variable and then use it with something like \\\"\\\\dt :mysort\\\" and have it expand the normal way there.-- Magnus Hagander Me: https://www.hagander.net/ Work: https://www.redpill-linpro.com/\",\n \"msg_date\": \"Tue, 14 Nov 2017 13:26:30 +0100\",\n \"msg_from\": \"Magnus Hagander <magnus@hagander.net>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Tue, Nov 14, 2017 at 3:26 PM, Magnus Hagander <magnus@hagander.net>\\nwrote:\\n\\n> On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com>\\n> wrote:\\n>\\n>> On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com>\\n>> wrote:\\n>> > I though about Alexander proposal, and I am thinking so it can be\\n>> probably\\n>> > best if we respect psql design. I implemented two command suffixes\\n>> > (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n>> >\\n>> > so list of tables can be sorted with commands:\\n>> >\\n>> > \\\\dt+sd (in this case, the order is not strict), so command\\n>> > \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>> >\\n>> > These two chars are acceptable. Same principle is used for \\\\l command\\n>> >\\n>> > \\\\lsd+ or \\\\l+sd\\n>> >\\n>> > What do you think about it?\\n>> >\\n>>\\n>> I really hate that syntax. This is going to turn into an\\n>> incomprehensible mess, and isn't easily extended to support other\\n>> options.\\n>>\\n>\\n> +1. While useful in itself, I think it's definitely a dangerous pattern to\\n> go down, as it'll only get worse.\\n>\\n>\\n> I agree with people who have said they would prefer this to be\\n>> available as a per-command option rather than as a variable that you\\n>> have to set, but it needs a clearer syntax. I actually like Stephen's\\n>> idea of using a user-defined SQL snippet, because that's a familiar\\n>> syntax to people, and it avoids adding an ever-increasing number of\\n>> options to these commands. Instead, the syntax could simply be:\\n>>\\n>\\n> +1 here as well. And anybody who is actually going to need this level of\\n> control definitely will know SQL...\\n>\\n> And if one wants to save some \\\"standard patterns\\\", it should be doable to\\n> save the pattern itself in a variable and then use it with something like\\n> \\\"\\\\dt :mysort\\\" and have it expand the normal way there.\\n>\\n\\n+1\\nI agree, that would look better, especially with \\\"standard patterns\\\" which\\ncould help with too long to type each time SQL snippets.\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nOn Tue, Nov 14, 2017 at 3:26 PM, Magnus Hagander <magnus@hagander.net> wrote:On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com> wrote:On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com> wrote:\\n> I though about Alexander proposal, and I am thinking so it can be probably\\n> best if we respect psql design. I implemented two command suffixes\\n> (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n>\\n> so list of tables can be sorted with commands:\\n>\\n> \\\\dt+sd (in this case, the order is not strict), so command\\n> \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>\\n> These two chars are acceptable. Same principle is used for \\\\l command\\n>\\n> \\\\lsd+ or \\\\l+sd\\n>\\n> What do you think about it?\\n>\\n\\nI really hate that syntax. This is going to turn into an\\nincomprehensible mess, and isn't easily extended to support other\\noptions.+1. While useful in itself, I think it's definitely a dangerous pattern to go down, as it'll only get worse.\\nI agree with people who have said they would prefer this to be\\navailable as a per-command option rather than as a variable that you\\nhave to set, but it needs a clearer syntax. I actually like Stephen's\\nidea of using a user-defined SQL snippet, because that's a familiar\\nsyntax to people, and it avoids adding an ever-increasing number of\\noptions to these commands. Instead, the syntax could simply be:+1 here as well. And anybody who is actually going to need this level of control definitely will know SQL...And if one wants to save some \\\"standard patterns\\\", it should be doable to save the pattern itself in a variable and then use it with something like \\\"\\\\dt :mysort\\\" and have it expand the normal way there.+1I agree, that would look better, especially with \\\"standard patterns\\\" which could help with too long to type each time SQL snippets.------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Mon, 27 Nov 2017 10:58:59 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-11-27 8:58 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n\\n> On Tue, Nov 14, 2017 at 3:26 PM, Magnus Hagander <magnus@hagander.net>\\n> wrote:\\n>\\n>> On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com>\\n>> wrote:\\n>>\\n>>> On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com>\\n>>> wrote:\\n>>> > I though about Alexander proposal, and I am thinking so it can be\\n>>> probably\\n>>> > best if we respect psql design. I implemented two command suffixes\\n>>> > (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as\\n>>> descent\\n>>> >\\n>>> > so list of tables can be sorted with commands:\\n>>> >\\n>>> > \\\\dt+sd (in this case, the order is not strict), so command\\n>>> > \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>>> >\\n>>> > These two chars are acceptable. Same principle is used for \\\\l command\\n>>> >\\n>>> > \\\\lsd+ or \\\\l+sd\\n>>> >\\n>>> > What do you think about it?\\n>>> >\\n>>>\\n>>> I really hate that syntax. This is going to turn into an\\n>>> incomprehensible mess, and isn't easily extended to support other\\n>>> options.\\n>>>\\n>>\\n>> +1. While useful in itself, I think it's definitely a dangerous pattern\\n>> to go down, as it'll only get worse.\\n>>\\n>>\\n>> I agree with people who have said they would prefer this to be\\n>>> available as a per-command option rather than as a variable that you\\n>>> have to set, but it needs a clearer syntax. I actually like Stephen's\\n>>> idea of using a user-defined SQL snippet, because that's a familiar\\n>>> syntax to people, and it avoids adding an ever-increasing number of\\n>>> options to these commands. Instead, the syntax could simply be:\\n>>>\\n>>\\n>> +1 here as well. And anybody who is actually going to need this level of\\n>> control definitely will know SQL...\\n>>\\n>> And if one wants to save some \\\"standard patterns\\\", it should be doable to\\n>> save the pattern itself in a variable and then use it with something like\\n>> \\\"\\\\dt :mysort\\\" and have it expand the normal way there.\\n>>\\n>\\n> +1\\n> I agree, that would look better, especially with \\\"standard patterns\\\" which\\n> could help with too long to type each time SQL snippets.\\n>\\n\\nI though about this design more time. I see following disadvantages\\n\\n1. we are not able to check all possible variants of extended query. If\\nthere will be some custom error, then we will raise pretty ugly error\\nmessages,\\n\\n2. I don't thing so using \\\"Size\\\" as table size in human readable format and\\n\\\"size\\\" as table size in raw format is intuitive, but any change of \\\"Size\\\"\\ncan introduce some (less probability compatibility issues),\\n\\n3. What queries will contains size calculations? It is not cheap - requires\\nAccessShareLock\\n\\nThis proposal is first time, when we cannot to detect full semantic from\\n\\\\xxx command. When user extend query correctly, then it is better than\\nbefore, when not it is worse than before.\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n\\n2017-11-27 8:58 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:On Tue, Nov 14, 2017 at 3:26 PM, Magnus Hagander <magnus@hagander.net> wrote:On Mon, Nov 13, 2017 at 3:17 PM, Dean Rasheed <dean.a.rasheed@gmail.com> wrote:On 28 October 2017 at 13:46, Pavel Stehule <pavel.stehule@gmail.com> wrote:\\n> I though about Alexander proposal, and I am thinking so it can be probably\\n> best if we respect psql design. I implemented two command suffixes\\n> (supported only when it has sense) \\\"s\\\" sorted by size and \\\"d\\\" as descent\\n>\\n> so list of tables can be sorted with commands:\\n>\\n> \\\\dt+sd (in this case, the order is not strict), so command\\n> \\\\dtsd+ is working too (same \\\\disd+ or \\\\di+sd)\\n>\\n> These two chars are acceptable. Same principle is used for \\\\l command\\n>\\n> \\\\lsd+ or \\\\l+sd\\n>\\n> What do you think about it?\\n>\\n\\nI really hate that syntax. This is going to turn into an\\nincomprehensible mess, and isn't easily extended to support other\\noptions.+1. While useful in itself, I think it's definitely a dangerous pattern to go down, as it'll only get worse.\\nI agree with people who have said they would prefer this to be\\navailable as a per-command option rather than as a variable that you\\nhave to set, but it needs a clearer syntax. I actually like Stephen's\\nidea of using a user-defined SQL snippet, because that's a familiar\\nsyntax to people, and it avoids adding an ever-increasing number of\\noptions to these commands. Instead, the syntax could simply be:+1 here as well. And anybody who is actually going to need this level of control definitely will know SQL...And if one wants to save some \\\"standard patterns\\\", it should be doable to save the pattern itself in a variable and then use it with something like \\\"\\\\dt :mysort\\\" and have it expand the normal way there.+1I agree, that would look better, especially with \\\"standard patterns\\\" which could help with too long to type each time SQL snippets.I though about this design more time. I see following disadvantages1. we are not able to check all possible variants of extended query. If there will be some custom error, then we will raise pretty ugly error messages,2. I don't thing so using \\\"Size\\\" as table size in human readable format and \\\"size\\\" as table size in raw format is intuitive, but any change of \\\"Size\\\" can introduce some (less probability compatibility issues),3. What queries will contains size calculations? It is not cheap - requires AccessShareLockThis proposal is first time, when we cannot to detect full semantic from \\\\xxx command. When user extend query correctly, then it is better than before, when not it is worse than before.RegardsPavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Mon, 27 Nov 2017 20:18:17 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Tue, Nov 28, 2017 at 4:18 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:\\n> This proposal is first time, when we cannot to detect full semantic from\\n> \\\\xxx command. When user extend query correctly, then it is better than\\n> before, when not it is worse than before.\\n\\nAs the discussion is still going on, I am moving this patch to next CF.\\n-- \\nMichael\\n\\n\",\n \"msg_date\": \"Wed, 29 Nov 2017 14:17:16 +0900\",\n \"msg_from\": \"Michael Paquier <michael.paquier@gmail.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"Hi!\\n\\nOn Mon, Nov 27, 2017 at 10:18 PM, Pavel Stehule <pavel.stehule@gmail.com>\\nwrote:\\n\\n> I though about this design more time. I see following disadvantages\\n>\\n> 1. we are not able to check all possible variants of extended query. If\\n> there will be some custom error, then we will raise pretty ugly error\\n> messages,\\n>\\n\\nYes, that's an inevitable shortcoming. psql is not backend and can't\\nperform all the required checks on its side...\\n\\n2. I don't thing so using \\\"Size\\\" as table size in human readable format and\\n> \\\"size\\\" as table size in raw format is intuitive, but any change of \\\"Size\\\"\\n> can introduce some (less probability compatibility issues),\\n>\\n\\nOh, this is surprisingly hard problem which probably have only imperative\\nsolution...\\n\\n3. What queries will contains size calculations? It is not cheap - requires\\n> AccessShareLock\\n>\\n\\nSorry, I didn't understand this point. Yes, size calculation requires\\nlocking, but that is already true for \\\\dt+ and \\\\l+. Why this is\\ndisadvantage of proposed approach?\\n\\n------\\nAlexander Korotkov\\nPostgres Professional: http://www.postgrespro.com\\nThe Russian Postgres Company\\n\\nHi!On Mon, Nov 27, 2017 at 10:18 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:I though about this design more time. I see following disadvantages1. we are not able to check all possible variants of extended query. If there will be some custom error, then we will raise pretty ugly error messages, Yes, that's an inevitable shortcoming. psql is not backend and can't perform all the required checks on its side... 2. I don't thing so using \\\"Size\\\" as table size in human readable format and \\\"size\\\" as table size in raw format is intuitive, but any change of \\\"Size\\\" can introduce some (less probability compatibility issues),Oh, this is surprisingly hard problem which probably have only imperative solution...3. What queries will contains size calculations? It is not cheap - requires AccessShareLock Sorry, I didn't understand this point. Yes, size calculation requires locking, but that is already true for \\\\dt+ and \\\\l+. Why this is disadvantage of proposed approach?------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Sat, 9 Dec 2017 00:49:57 +0300\",\n \"msg_from\": \"Alexander Korotkov <a.korotkov@postgrespro.ru>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"2017-12-08 22:49 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:\\n\\n> Hi!\\n>\\n> On Mon, Nov 27, 2017 at 10:18 PM, Pavel Stehule <pavel.stehule@gmail.com>\\n> wrote:\\n>\\n>> I though about this design more time. I see following disadvantages\\n>>\\n>> 1. we are not able to check all possible variants of extended query. If\\n>> there will be some custom error, then we will raise pretty ugly error\\n>> messages,\\n>>\\n>\\n> Yes, that's an inevitable shortcoming. psql is not backend and can't\\n> perform all the required checks on its side...\\n>\\n> 2. I don't thing so using \\\"Size\\\" as table size in human readable format\\n>> and \\\"size\\\" as table size in raw format is intuitive, but any change of\\n>> \\\"Size\\\" can introduce some (less probability compatibility issues),\\n>>\\n>\\n> Oh, this is surprisingly hard problem which probably have only imperative\\n> solution...\\n>\\n> 3. What queries will contains size calculations? It is not cheap -\\n>> requires AccessShareLock\\n>>\\n>\\n> Sorry, I didn't understand this point. Yes, size calculation requires\\n> locking, but that is already true for \\\\dt+ and \\\\l+. Why this is\\n> disadvantage of proposed approach?\\n>\\n\\nBecause you don't know the filter and sort clause (it can be generic), you\\ndon't know, if you should to calculate or not the size. Or there should be\\nrule, so filter, order must be limited to displayed columns.\\n\\nRegards\\n\\nPavel\\n\\n\\n\\n\\n> ------\\n> Alexander Korotkov\\n> Postgres Professional: http://www.postgrespro.com\\n> The Russian Postgres Company\\n>\\n\\n2017-12-08 22:49 GMT+01:00 Alexander Korotkov <a.korotkov@postgrespro.ru>:Hi!On Mon, Nov 27, 2017 at 10:18 PM, Pavel Stehule <pavel.stehule@gmail.com> wrote:I though about this design more time. I see following disadvantages1. we are not able to check all possible variants of extended query. If there will be some custom error, then we will raise pretty ugly error messages, Yes, that's an inevitable shortcoming. psql is not backend and can't perform all the required checks on its side... 2. I don't thing so using \\\"Size\\\" as table size in human readable format and \\\"size\\\" as table size in raw format is intuitive, but any change of \\\"Size\\\" can introduce some (less probability compatibility issues),Oh, this is surprisingly hard problem which probably have only imperative solution...3. What queries will contains size calculations? It is not cheap - requires AccessShareLock Sorry, I didn't understand this point. Yes, size calculation requires locking, but that is already true for \\\\dt+ and \\\\l+. Why this is disadvantage of proposed approach?Because you don't know the filter and sort clause (it can be generic), you don't know, if you should to calculate or not the size. Or there should be rule, so filter, order must be limited to displayed columns.RegardsPavel ------Alexander KorotkovPostgres Professional: http://www.postgrespro.comThe Russian Postgres Company\",\n \"msg_date\": \"Sat, 9 Dec 2017 06:11:33 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"On Thu, Nov 2, 2017 at 12:44 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:\\n> I am sending updated patch with some basic doc\\n\\nHi Pavel,\\n\\nI am not sure what the status of this patch is, but FYI:\\n\\nstartup.c: In function ‘main’:\\nstartup.c:284:3: error: too few arguments to function ‘listAllDbs’\\n success = listAllDbs(NULL, false);\\n ^\\nIn file included from startup.c:21:0:\\ndescribe.h:68:13: note: declared here\\n extern bool listAllDbs(const char *pattern, bool verbose, sortby_type\\nsortby, bool sort_desc);\\n ^\\n\\n-- \\nThomas Munro\\nhttp://www.enterprisedb.com\\n\\n\",\n \"msg_date\": \"Fri, 12 Jan 2018 14:35:37 +1300\",\n \"msg_from\": \"Thomas Munro <thomas.munro@enterprisedb.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n },\n {\n \"msg_contents\": \"Hi\\n\\n2018-01-12 2:35 GMT+01:00 Thomas Munro <thomas.munro@enterprisedb.com>:\\n\\n> On Thu, Nov 2, 2017 at 12:44 AM, Pavel Stehule <pavel.stehule@gmail.com>\\n> wrote:\\n> > I am sending updated patch with some basic doc\\n>\\n> Hi Pavel,\\n>\\n> I am not sure what the status of this patch is, but FYI:\\n>\\n> startup.c: In function ‘main’:\\n> startup.c:284:3: error: too few arguments to function ‘listAllDbs’\\n> success = listAllDbs(NULL, false);\\n> ^\\n> In file included from startup.c:21:0:\\n> describe.h:68:13: note: declared here\\n> extern bool listAllDbs(const char *pattern, bool verbose, sortby_type\\n> sortby, bool sort_desc);\\n> ^\\n>\\n\\nThere are no agreement about optimal form of this feature - so I'll remove\\nthis patch from commitfest.\\n\\nRegards\\n\\nPavel\\n\\n\\n>\\n> --\\n> Thomas Munro\\n> http://www.enterprisedb.com\\n>\\n\\nHi2018-01-12 2:35 GMT+01:00 Thomas Munro <thomas.munro@enterprisedb.com>:On Thu, Nov 2, 2017 at 12:44 AM, Pavel Stehule <pavel.stehule@gmail.com> wrote:\\n> I am sending updated patch with some basic doc\\n\\nHi Pavel,\\n\\nI am not sure what the status of this patch is, but FYI:\\n\\nstartup.c: In function ‘main’:\\nstartup.c:284:3: error: too few arguments to function ‘listAllDbs’\\n success = listAllDbs(NULL, false);\\n ^\\nIn file included from startup.c:21:0:\\ndescribe.h:68:13: note: declared here\\n extern bool listAllDbs(const char *pattern, bool verbose, sortby_type\\nsortby, bool sort_desc);\\n ^There are no agreement about optimal form of this feature - so I'll remove this patch from commitfest.RegardsPavel \\n\\n--\\nThomas Munro\\nhttp://www.enterprisedb.com\",\n \"msg_date\": \"Fri, 12 Jan 2018 16:11:02 +0100\",\n \"msg_from\": \"Pavel Stehule <pavel.stehule@gmail.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: [HACKERS] Re: proposal - psql: possibility to specify sort for\\n describe commands, when size is printed\"\n }\n]","isConversational":false}}},{"rowIdx":43998,"cells":{"threads":{"kind":"list like","value":[{"msg_contents":"Hello, while I investigated a bug of async-exec patch. I found a\n(maybe) typo in mcxt.c.\n\n| * It's not entirely clear whether 'tis better to do this before or after\n| * delinking the context; but an error in a callback will likely result in\n\nI'm not so confident, but the \"'tis\" seems to me to be a typo of\n\"it's\".\n\n\nregards\n\n-- \nKyotaro Horiguchi\nNTT Open Source Software Center\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers","msg_date":"Thu, 23 Feb 2017 21:10:37 +0900 (Tokyo Standard Time)","msg_from":"Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>","msg_from_op":true,"msg_subject":"A typo in mcxt.c"},{"msg_contents":"On Thu, Feb 23, 2017 at 5:40 PM, Kyotaro HORIGUCHI\n<horiguchi.kyotaro@lab.ntt.co.jp> wrote:\n> Hello, while I investigated a bug of async-exec patch. I found a\n> (maybe) typo in mcxt.c.\n>\n> | * It's not entirely clear whether 'tis better to do this before or after\n> | * delinking the context; but an error in a callback will likely result in\n>\n> I'm not so confident, but the \"'tis\" seems to me to be a typo of\n> \"it's\".\n\nThat is an archaic way of contracting the same words differently:\n\nhttps://en.wiktionary.org/wiki/'tis\n\nIt sounds amusing and a tiny bit like a famous line from Hamlet,\n\"Whether 'tis nobler in the mind to suffer...\".\n\nhttps://en.wikipedia.org/wiki/To_be,_or_not_to_be\n\n-- \nThomas Munro\nhttp://www.enterprisedb.com\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Thu, 23 Feb 2017 18:08:44 +0530","msg_from":"Thomas Munro <thomas.munro@enterprisedb.com>","msg_from_op":false,"msg_subject":"Re: A typo in mcxt.c"},{"msg_contents":"On 2/23/17 6:38 AM, Thomas Munro wrote:\n>> I'm not so confident, but the \"'tis\" seems to me to be a typo of\n>> \"it's\".\n> That is an archaic way of contracting the same words differently:\n\nGiven the number of non-native English speakers we have, it's probably \nworth changing it...\n-- \nJim Nasby, Data Architect, Blue Treble Consulting, Austin TX\nExperts in Analytics, Data Architecture and PostgreSQL\nData in Trouble? Get it in Treble! http://BlueTreble.com\n855-TREBLE2 (855-873-2532)\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Thu, 23 Feb 2017 14:26:07 -0600","msg_from":"Jim Nasby <Jim.Nasby@BlueTreble.com>","msg_from_op":false,"msg_subject":"Re: A typo in mcxt.c"},{"msg_contents":"On 2017-02-23 14:26:07 -0600, Jim Nasby wrote:\n> On 2/23/17 6:38 AM, Thomas Munro wrote:\n> > > I'm not so confident, but the \"'tis\" seems to me to be a typo of\n> > > \"it's\".\n> > That is an archaic way of contracting the same words differently:\n> \n> Given the number of non-native English speakers we have, it's probably worth\n> changing it...\n\nI'm a non-native speaker and I actually like discovering new language\n\"features\" every now and then. I think as long as it's not inhibiting\nunderstanding to much - which doesn't seem to be the case here - it's\nfine to keep things like this.\n\n- Andres\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Thu, 23 Feb 2017 19:45:14 -0800","msg_from":"Andres Freund <andres@anarazel.de>","msg_from_op":false,"msg_subject":"Re: A typo in mcxt.c"},{"msg_contents":"Andres Freund <andres@anarazel.de> writes:\n> On 2017-02-23 14:26:07 -0600, Jim Nasby wrote:\n>> On 2/23/17 6:38 AM, Thomas Munro wrote:\n>>> That is an archaic way of contracting the same words differently:\n\n>> Given the number of non-native English speakers we have, it's probably worth\n>> changing it...\n\n> I'm a non-native speaker and I actually like discovering new language\n> \"features\" every now and then. I think as long as it's not inhibiting\n> understanding to much - which doesn't seem to be the case here - it's\n> fine to keep things like this.\n\nWhile I don't recall it specifically, git blame says that comment is mine.\nI'm pretty sure it's not a typo, but that the allusion to Hamlet was\nintentional. I think it's good to have a bit of levity and external\nreferences in our comments; cuts down on the boredom of reading totally\ndry code.\n\n(But see commit d2783bee3 for one hazard of this sort of thing.)\n\n\t\t\tregards, tom lane\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Thu, 23 Feb 2017 22:59:51 -0500","msg_from":"Tom Lane <tgl@sss.pgh.pa.us>","msg_from_op":false,"msg_subject":"Re: A typo in mcxt.c"},{"msg_contents":"I'm happy to know such a thing. mcxt.c is very stable part of the\ncode so those who don't know such things like me rarely comes.\n\nAt Thu, 23 Feb 2017 22:59:51 -0500, Tom Lane <tgl@sss.pgh.pa.us> wrote in <31880.1487908791@sss.pgh.pa.us>\n> Andres Freund <andres@anarazel.de> writes:\n> > On 2017-02-23 14:26:07 -0600, Jim Nasby wrote:\n> >> On 2/23/17 6:38 AM, Thomas Munro wrote:\n> >>> That is an archaic way of contracting the same words differently:\n> \n> >> Given the number of non-native English speakers we have, it's probably worth\n> >> changing it...\n> \n> > I'm a non-native speaker and I actually like discovering new language\n> > \"features\" every now and then. I think as long as it's not inhibiting\n> > understanding to much - which doesn't seem to be the case here - it's\n> > fine to keep things like this.\n> \n> While I don't recall it specifically, git blame says that comment is mine.\n> I'm pretty sure it's not a typo, but that the allusion to Hamlet was\n> intentional. I think it's good to have a bit of levity and external\n> references in our comments; cuts down on the boredom of reading totally\n> dry code.\n> \n> (But see commit d2783bee3 for one hazard of this sort of thing.)\n> \n> \t\t\tregards, tom lane\n\n-- \nKyotaro Horiguchi\nNTT Open Source Software Center\n\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Mon, 27 Feb 2017 13:46:57 +0900 (Tokyo Standard Time)","msg_from":"Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>","msg_from_op":true,"msg_subject":"Re: A typo in mcxt.c"}],"string":"[\n {\n \"msg_contents\": \"Hello, while I investigated a bug of async-exec patch. I found a\\n(maybe) typo in mcxt.c.\\n\\n| * It's not entirely clear whether 'tis better to do this before or after\\n| * delinking the context; but an error in a callback will likely result in\\n\\nI'm not so confident, but the \\\"'tis\\\" seems to me to be a typo of\\n\\\"it's\\\".\\n\\n\\nregards\\n\\n-- \\nKyotaro Horiguchi\\nNTT Open Source Software Center\\n\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\",\n \"msg_date\": \"Thu, 23 Feb 2017 21:10:37 +0900 (Tokyo Standard Time)\",\n \"msg_from\": \"Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"A typo in mcxt.c\"\n },\n {\n \"msg_contents\": \"On Thu, Feb 23, 2017 at 5:40 PM, Kyotaro HORIGUCHI\\n<horiguchi.kyotaro@lab.ntt.co.jp> wrote:\\n> Hello, while I investigated a bug of async-exec patch. I found a\\n> (maybe) typo in mcxt.c.\\n>\\n> | * It's not entirely clear whether 'tis better to do this before or after\\n> | * delinking the context; but an error in a callback will likely result in\\n>\\n> I'm not so confident, but the \\\"'tis\\\" seems to me to be a typo of\\n> \\\"it's\\\".\\n\\nThat is an archaic way of contracting the same words differently:\\n\\nhttps://en.wiktionary.org/wiki/'tis\\n\\nIt sounds amusing and a tiny bit like a famous line from Hamlet,\\n\\\"Whether 'tis nobler in the mind to suffer...\\\".\\n\\nhttps://en.wikipedia.org/wiki/To_be,_or_not_to_be\\n\\n-- \\nThomas Munro\\nhttp://www.enterprisedb.com\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Thu, 23 Feb 2017 18:08:44 +0530\",\n \"msg_from\": \"Thomas Munro <thomas.munro@enterprisedb.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: A typo in mcxt.c\"\n },\n {\n \"msg_contents\": \"On 2/23/17 6:38 AM, Thomas Munro wrote:\\n>> I'm not so confident, but the \\\"'tis\\\" seems to me to be a typo of\\n>> \\\"it's\\\".\\n> That is an archaic way of contracting the same words differently:\\n\\nGiven the number of non-native English speakers we have, it's probably \\nworth changing it...\\n-- \\nJim Nasby, Data Architect, Blue Treble Consulting, Austin TX\\nExperts in Analytics, Data Architecture and PostgreSQL\\nData in Trouble? Get it in Treble! http://BlueTreble.com\\n855-TREBLE2 (855-873-2532)\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Thu, 23 Feb 2017 14:26:07 -0600\",\n \"msg_from\": \"Jim Nasby <Jim.Nasby@BlueTreble.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: A typo in mcxt.c\"\n },\n {\n \"msg_contents\": \"On 2017-02-23 14:26:07 -0600, Jim Nasby wrote:\\n> On 2/23/17 6:38 AM, Thomas Munro wrote:\\n> > > I'm not so confident, but the \\\"'tis\\\" seems to me to be a typo of\\n> > > \\\"it's\\\".\\n> > That is an archaic way of contracting the same words differently:\\n> \\n> Given the number of non-native English speakers we have, it's probably worth\\n> changing it...\\n\\nI'm a non-native speaker and I actually like discovering new language\\n\\\"features\\\" every now and then. I think as long as it's not inhibiting\\nunderstanding to much - which doesn't seem to be the case here - it's\\nfine to keep things like this.\\n\\n- Andres\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Thu, 23 Feb 2017 19:45:14 -0800\",\n \"msg_from\": \"Andres Freund <andres@anarazel.de>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: A typo in mcxt.c\"\n },\n {\n \"msg_contents\": \"Andres Freund <andres@anarazel.de> writes:\\n> On 2017-02-23 14:26:07 -0600, Jim Nasby wrote:\\n>> On 2/23/17 6:38 AM, Thomas Munro wrote:\\n>>> That is an archaic way of contracting the same words differently:\\n\\n>> Given the number of non-native English speakers we have, it's probably worth\\n>> changing it...\\n\\n> I'm a non-native speaker and I actually like discovering new language\\n> \\\"features\\\" every now and then. I think as long as it's not inhibiting\\n> understanding to much - which doesn't seem to be the case here - it's\\n> fine to keep things like this.\\n\\nWhile I don't recall it specifically, git blame says that comment is mine.\\nI'm pretty sure it's not a typo, but that the allusion to Hamlet was\\nintentional. I think it's good to have a bit of levity and external\\nreferences in our comments; cuts down on the boredom of reading totally\\ndry code.\\n\\n(But see commit d2783bee3 for one hazard of this sort of thing.)\\n\\n\\t\\t\\tregards, tom lane\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Thu, 23 Feb 2017 22:59:51 -0500\",\n \"msg_from\": \"Tom Lane <tgl@sss.pgh.pa.us>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: A typo in mcxt.c\"\n },\n {\n \"msg_contents\": \"I'm happy to know such a thing. mcxt.c is very stable part of the\\ncode so those who don't know such things like me rarely comes.\\n\\nAt Thu, 23 Feb 2017 22:59:51 -0500, Tom Lane <tgl@sss.pgh.pa.us> wrote in <31880.1487908791@sss.pgh.pa.us>\\n> Andres Freund <andres@anarazel.de> writes:\\n> > On 2017-02-23 14:26:07 -0600, Jim Nasby wrote:\\n> >> On 2/23/17 6:38 AM, Thomas Munro wrote:\\n> >>> That is an archaic way of contracting the same words differently:\\n> \\n> >> Given the number of non-native English speakers we have, it's probably worth\\n> >> changing it...\\n> \\n> > I'm a non-native speaker and I actually like discovering new language\\n> > \\\"features\\\" every now and then. I think as long as it's not inhibiting\\n> > understanding to much - which doesn't seem to be the case here - it's\\n> > fine to keep things like this.\\n> \\n> While I don't recall it specifically, git blame says that comment is mine.\\n> I'm pretty sure it's not a typo, but that the allusion to Hamlet was\\n> intentional. I think it's good to have a bit of levity and external\\n> references in our comments; cuts down on the boredom of reading totally\\n> dry code.\\n> \\n> (But see commit d2783bee3 for one hazard of this sort of thing.)\\n> \\n> \\t\\t\\tregards, tom lane\\n\\n-- \\nKyotaro Horiguchi\\nNTT Open Source Software Center\\n\\n\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Mon, 27 Feb 2017 13:46:57 +0900 (Tokyo Standard Time)\",\n \"msg_from\": \"Kyotaro HORIGUCHI <horiguchi.kyotaro@lab.ntt.co.jp>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: A typo in mcxt.c\"\n }\n]","isConversational":false}}},{"rowIdx":43999,"cells":{"threads":{"kind":"list like","value":[{"msg_contents":"Andrew Dunstan wrote:\n> \n> While reviewing the IF NOT EXISTS patch for CREATE USER MAPPING I\n> noticed that in several places we treat the user name as the name of the\n> user mapping. Strictly ISTM that user mappings are really anonymous\n> objects, so instead of something like user \"mapping \\\"%s\\\" does not\n> exist for the server\" we should possibly have \"user mapping for user\n> \\\"%s\\\" does not exist for the server\". I was about to make that change\n> in the patch when I saw that it was consistent with current usage. Do we\n> want to stick with the current usage where we treat the user name as the\n> mapping name, or change it?\n\nHmm, I vaguely recall that due to some previous discussion I changed\nsome of uses of the former wording to your proposed one, which I agree\nis an improvement.\n\n-- \n�lvaro Herrera https://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Thu, 23 Feb 2017 11:52:44 -0300","msg_from":"Alvaro Herrera <alvherre@2ndquadrant.com>","msg_from_op":true,"msg_subject":"Re: user mapping messages"},{"msg_contents":"On Thu, Feb 23, 2017 at 8:06 PM, Andrew Dunstan\n<andrew.dunstan@2ndquadrant.com> wrote:\n>\n> While reviewing the IF NOT EXISTS patch for CREATE USER MAPPING I\n> noticed that in several places we treat the user name as the name of the\n> user mapping. Strictly ISTM that user mappings are really anonymous\n> objects, so instead of something like user \"mapping \\\"%s\\\" does not\n> exist for the server\" we should possibly have \"user mapping for user\n> \\\"%s\\\" does not exist for the server\".\n\nYour proposed usage is better than the existing one.\n\n> I was about to make that change\n> in the patch when I saw that it was consistent with current usage. Do we\n> want to stick with the current usage where we treat the user name as the\n> mapping name, or change it?\n>\n\nWe should change existing usage and then commit the patch with new\nusage. The new message being added should be consistent with other\nplaces.\n\n-- \nBest Wishes,\nAshutosh Bapat\nEnterpriseDB Corporation\nThe Postgres Database Company\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers\n","msg_date":"Thu, 2 Mar 2017 12:35:19 +0530","msg_from":"Ashutosh Bapat <ashutosh.bapat@enterprisedb.com>","msg_from_op":false,"msg_subject":"Re: user mapping messages"}],"string":"[\n {\n \"msg_contents\": \"Andrew Dunstan wrote:\\n> \\n> While reviewing the IF NOT EXISTS patch for CREATE USER MAPPING I\\n> noticed that in several places we treat the user name as the name of the\\n> user mapping. Strictly ISTM that user mappings are really anonymous\\n> objects, so instead of something like user \\\"mapping \\\\\\\"%s\\\\\\\" does not\\n> exist for the server\\\" we should possibly have \\\"user mapping for user\\n> \\\\\\\"%s\\\\\\\" does not exist for the server\\\". I was about to make that change\\n> in the patch when I saw that it was consistent with current usage. Do we\\n> want to stick with the current usage where we treat the user name as the\\n> mapping name, or change it?\\n\\nHmm, I vaguely recall that due to some previous discussion I changed\\nsome of uses of the former wording to your proposed one, which I agree\\nis an improvement.\\n\\n-- \\n�lvaro Herrera https://www.2ndQuadrant.com/\\nPostgreSQL Development, 24x7 Support, Remote DBA, Training & Services\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Thu, 23 Feb 2017 11:52:44 -0300\",\n \"msg_from\": \"Alvaro Herrera <alvherre@2ndquadrant.com>\",\n \"msg_from_op\": true,\n \"msg_subject\": \"Re: user mapping messages\"\n },\n {\n \"msg_contents\": \"On Thu, Feb 23, 2017 at 8:06 PM, Andrew Dunstan\\n<andrew.dunstan@2ndquadrant.com> wrote:\\n>\\n> While reviewing the IF NOT EXISTS patch for CREATE USER MAPPING I\\n> noticed that in several places we treat the user name as the name of the\\n> user mapping. Strictly ISTM that user mappings are really anonymous\\n> objects, so instead of something like user \\\"mapping \\\\\\\"%s\\\\\\\" does not\\n> exist for the server\\\" we should possibly have \\\"user mapping for user\\n> \\\\\\\"%s\\\\\\\" does not exist for the server\\\".\\n\\nYour proposed usage is better than the existing one.\\n\\n> I was about to make that change\\n> in the patch when I saw that it was consistent with current usage. Do we\\n> want to stick with the current usage where we treat the user name as the\\n> mapping name, or change it?\\n>\\n\\nWe should change existing usage and then commit the patch with new\\nusage. The new message being added should be consistent with other\\nplaces.\\n\\n-- \\nBest Wishes,\\nAshutosh Bapat\\nEnterpriseDB Corporation\\nThe Postgres Database Company\\n\\n\\n-- \\nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\\nTo make changes to your subscription:\\nhttp://www.postgresql.org/mailpref/pgsql-hackers\\n\",\n \"msg_date\": \"Thu, 2 Mar 2017 12:35:19 +0530\",\n \"msg_from\": \"Ashutosh Bapat <ashutosh.bapat@enterprisedb.com>\",\n \"msg_from_op\": false,\n \"msg_subject\": \"Re: user mapping messages\"\n }\n]","isConversational":false}}}],"truncated":false,"partial":false},"paginationData":{"pageIndex":439,"numItemsPerPage":100,"numTotalItems":58371,"offset":43900,"length":100}},"jwt":"eyJhbGciOiJFZERTQSIsImtpZCI6IjVHZDBvd0g5MTM2eDZjc1FvbE1zcktNWUZoRVFoUm5PVVVybEpjOUhaUEEifQ.eyJyZWFkIjp0cnVlLCJwZXJtaXNzaW9ucyI6eyJyZXBvLmNvbnRlbnQucmVhZCI6dHJ1ZX0sImlhdCI6MTc4NTQyMDM3OSwianRpIjoiNDNmM2U2YjctNDMwYi00YzAxLWI2YWQtMzVkNDY1M2Q3YzRlIiwic3ViIjoiL2RhdGFzZXRzL3dhbnNoZW5sL3Bnc3FsLWhhY2tlcnMiLCJleHAiOjE3ODU0MjM5NzksImlzcyI6Imh0dHBzOi8vaHVnZ2luZ2ZhY2UuY28ifQ.eeYsCrx4c4E6IwvIWly5B0pDw_necQyhuv9kjs14mTGEFeW4BqAzzLzrGpjvvBeCn0N0kjBAqXkm-ntWH1UXCg","displayUrls":true,"splitSizeSummaries":[{"config":"default","split":"train","numRows":58371,"numBytesParquet":368254123}]},"discussionsStats":{"closed":0,"open":0,"total":0},"fullWidth":true,"hasGatedAccess":true,"hasFullAccess":true,"isEmbedded":false,"savedQueries":{"community":[],"user":[]}}">
threads listlengths 1 2.99k |
|---|
[
{
"msg_contents": "Hi,\n\nI'd like to make it easier to analyze slow queries in ECPG using \nlog_min_duration_statement. Especially when DBA and C application developer \nis different, DBA can feedback slow embedded query to the developer without \ndifficulty and mistakes.\n\nWhen our customers log and look int... |
[
{
"msg_contents": "I noticed that running ALTER TABLE table_name DETACH PARTITION crashes, if\ntable_name is not a partitioned table. That's because of an Assert in\nATExecDetachPartition(). We really should error out much sooner in this\ncase, IOW during transformAlterTableStmt(), as is done in the case of\... |
[
{
"msg_contents": "Hi\nBulletted list in Devel branch are all showing up bold as seen in the\nscreenshot of [1] attached. But the same page for 9.6 branch shows\nnormal bulletted lists. If I try to bulid docs from the latest\nresources (at commit ae0e550ce1703621efb3b75f4558df253cef5bca), the\nhtml renders bull... |
[
{
"msg_contents": "Hi all guys,\nI have a database named edb. In this db , I have lots of schemas. Here I use\ntwo schemas: a and b. The two schemas belong to different role.\nNow , I want to export all objects of schema a . And then using these export\nfile to recover schema b. The user of the schema b can use... |
[
{
"msg_contents": "Hi hackers,\n\nI wonder why SUM aggregate is calculated for real (float4) type using \nfloating point accumulator?\nIt cause very confusing and unexpected behavior:\n\n-- postgres=# select sum(l_quantity) from lineitem where l_shipdate <= '1998-12-01';\n sum\n-------------\n 1.52688e+0... |
[
{
"msg_contents": "Hello.\n\nWhile playing with logical publications/subscriptions I’ve noted that walsender crashes\nwhen i’m trying to CREATE SUBSCRIPTION to a current server.\n\nSo having one running postgres instance on a standard port:\n\nstas=# CREATE SUBSCRIPTION mysub CONNECTION 'dbname=stas host=127.0.... |
[
{
"msg_contents": "Currently, an update of a partition key of a partition is not allowed,\nsince it requires to move the row(s) into the applicable partition.\n\nAttached is a WIP patch (update-partition-key.patch) that removes this\nrestriction. When an UPDATE causes the row of a partition to violate\nits part... |
[
{
"msg_contents": "Hello hackers,\n\nThere were many discussions concerning possible ways of speeding-up \nPostgres. Different approaches were suggested:\n\n- JIT (now we have three different prototype implementations based on LLVM)\n- Chunked (vectorized) executor\n- Replacing pull with push\n- Columnar store... |
[
{
"msg_contents": "Is there a reason not to allow $SUBJECT? Specifically, it'd be nice to \nbe able to do something like WHEN tag LIKE 'ALTER%'.\n-- \nJim Nasby, Data Architect, Blue Treble Consulting, Austin TX\nExperts in Analytics, Data Architecture and PostgreSQL\nData in Trouble? Get it in Treble! http://B... |
[
{
"msg_contents": "Upgrading from 9.6 to dev, I now get:\n\n$ rm bisectdata -r ; bisect/bin/pg_ctl initdb -D bisectdata;\nbisect/bin/pg_upgrade -b /usr/local/pgsql9_6/bin/ -B bisect/bin/ -d 96 -D\nbisectdata/\n\n\ncheck for \"/usr/local/pgsql9_6/bin/pg_resetwal\" failed: No such file or\ndirectory\n\nThis looks... |
[
{
"msg_contents": "Hi all,\n\nBeginning a new thread to raise awareness... As already reported here,\nI had a look at what has been committed in a507b869:\nhttps://www.postgresql.org/message-id/CAB7nPqRzCQb=vdfHvMtP0HMLBHU6z1aGdo4GJsUP-HP8jx+Pkw@mail.gmail.com\n\nHere are a couple of things I have noticed while... |
[
{
"msg_contents": "Dilip complained on another thread about the apparent difficulty of\ngetting the system to compile after adding a reference to dsa_pointer\nto tidbitmap.c:\n\nhttps://www.postgresql.org/message-id/CAFiTN-u%3DBH_b0QE9FG%2B%2B_Ht6Q%3DF84am%3DJ-rt7fNbQkq3%2BqX3VQ%40mail.gmail.com\n\nI dug into t... |
[
{
"msg_contents": "Hi All,\n\nCurrently, Hash Index scan works tuple-at-a-time, i.e. for every\nqualifying tuple in a page, it acquires and releases the lock which\neventually increases the lock/unlock traffic. For example, if an index\npage contains 100 qualified tuples, the current hash index scan has to\nacq... |
[
{
"msg_contents": "Hi all, \n\nI propose the patch that adds a function GetOldestXminExtended that is like GetOldestXmin but can ignore arbitrary vacuum flags. And then, rewrite GetOldestXmin to use it. Note that this is done so as not to change the behavior of GetOldestXmin.\n\nThis change will be useful for f... |
[
{
"msg_contents": "Sync message causes backend to return a \"Ready for query\" response\neven if there's no query previously sent to the backend. I don't think\nthis is a bug but I'd think it would be better to write this behavior\nin the doc, because it might help someone who want to write an API\nwhich needs ... |
[
{
"msg_contents": "The revised draft of the proposed Code of Conduct for the PostgreSQL\ncommunity is at https://wiki.postgresql.org/wiki/Code_of_Conduct.\n\nThis updated draft incorporates comments and suggestions from the\ncommunity received at PgCon Ottawa and subsequent discussion.\n\nWe will not be monitor... |
[
{
"msg_contents": "After seeing Yet Another Missing Psql Tab Completion it occurred to \nme... why not add a checklist item to the commitfest review page? I \nrealize most regular contributors don't use the form, but a fair number \nof people do. I don't see how it could hurt.\n\nAnother possible idea is a git ... |
[
{
"msg_contents": "make installcheck currently fails against a server running\nwith bytea_output = escape.\n\nMaking it succeed is fairly easy, and I think it is worth doing.\n\nAttached are two options for doing that. One overrides bytea_output\nlocally where-ever needed, and the other overrides it for the en... |
[
{
"msg_contents": "Hi all,\n\nI propose a new feature for high availability. \n\nThis configuration is effective for following configuration: \n1. Primary and synchronous standby are in the same center; called main center. \n2. Asynchronous standby is in the another center; called backup center. \n (The backu... |
[
{
"msg_contents": "I ran into a case where a hash join took a really long time to respond\nto a cancel request --- long enough that I gave up and kill -9'd it,\nbecause its memory usage was also growing to the point where the kernel\nwould likely soon choose to do that for me. The culprit seems to be\nthat the... |
[
{
"msg_contents": "make installcheck fails against a server running with\noperator_precedence_warning\n= on.\n\nThe difference is in update.out, and consists of an error-locating carat\ngetting moved over by one position. I've attached the regression diff.\n\nI don't know why the setting of this GUC causes the... |
[
{
"msg_contents": "Hello,\n\n(I'm posting to hackers since I got no response on the general list.)\n\nI use Postgres + PostGIS quite heavily, and recently have been taking full\nadvantage of the new parallelism in 9.6. I'm now running queries in a few\nhours that would otherwise take more than a day.\n\nHowever... |
[
{
"msg_contents": "Throws a build error if we encounter a different number of fields in a\nDATA() line than we expect for the catalog in question.\n\nPreviously, it was possible to silently ignore any mismatches at build\ntime which could result in symbol undefined errors at link time. Now\nwe stop and identif... |
[
{
"msg_contents": "Fixes some DESCR() grammar mistakes introduced by the xlog -> wal changes.\n\n---\n src/include/catalog/pg_proc.h | 4 ++--\n 1 file changed, 2 insertions(+), 2 deletions(-)\n\ndiff --git a/src/include/catalog/pg_proc.h b/src/include/catalog/pg_proc.h\nindex 41c12af..bb7053a 100644\n--- a/src/... |
[
{
"msg_contents": "Hi all,\n\nI was looking through some of the implementation details of the\nheap/tuples, specifically src/include/access/htup_details.h - and I came\nacross the big macro fastgetattr, and had a question about it. I've\nincluded the code here for clarity and convenience:\n\n #define fastge... |
[
{
"msg_contents": "If RegisterBackgroundWorker() (the non-dynamic kind that is only\nloadable from shared_preload_libraries) fails to register the worker, it\nwrites a log message and proceeds, ignoring the registration request. I\nthink that is a mistake, it should be a hard error. The only way in\npractice ... |
[
{
"msg_contents": "The current help text for pg_basebackup -R is \"write recovery.conf after\nbackup\".\n\nThis says nothing about what it actually does. I've had a number of people\nask me now why that's not default \"because you need a recovery.conf to\nrestore from backup\". The point being that it doesn't s... |
[
{
"msg_contents": "Hi hackers,\n\n\"A system that is not case-preserving is necessarily case-insensitive,\nbut it is possible and common for a system to be case-insensitive, yet\ncase-preserving\" [1]\n\nImagine if you could turn on a GUC that would turn PostgreSQL into\nsuch a system,\nwhere the case would be ... |
[
{
"msg_contents": ">\n> It would help to know the data types of the columns involved in this\n> query; but just eyeballing it, it doesn't look like it involves any\n> array operations, so it's pretty hard to believe that the expanded-object\n> code could have gotten invoked intentionally. (The mere presence of... |
[
{
"msg_contents": "Hi,\n\nPlease find attached a patch to fix 2 typos.\n\n1) s/mypubclication/mypublication/\n\n2) Removed trailing comma from last column definition in example.\n\nThanks\n\nThom\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription... |
[
{
"msg_contents": "Hi,\n\nI've noticed that when creating a subscription, it can't be\ninterrupted. One must wait until it times out, which takes just over\n2 minutes. I'm guessing ALTER SUBSCRIPTION would have the same\nproblem.\n\nShouldn't we have an interrupt for this?\n\nThom\n\n\n-- \nSent via pgsql-hac... |
[
{
"msg_contents": "Hi,\n\nAt the moment, partitioned tables have a restriction that prevents\nthem allowing INSERT ... ON CONFLICT ... statements:\n\npostgres=# INSERT INTO cities SELECT 1, 'Crawley',105000 ON CONFLICT\n(city_id) DO NOTHING;\nERROR: ON CONFLICT clause is not supported with partitioned tables\n... |
[
{
"msg_contents": "When trying to build the documentation there is a minor warning:\n\n collateindex.pl: duplicated index entry found: MEDIAN\n\nIndeed, the \"median\" index term is specified twice in \"syntax.sgml\". The \nattached patch removes the warning.\n\n-- \nFabien.\n\n\n-- \nSent via pgsql-hackers m... |
[
{
"msg_contents": "Add new function dsa_allocate0.\n\nThis does the same thing as dsa_allocate, except that the memory\nis guaranteed to be zero-filled on return.\n\nDilip Kumar, adjusted by me.\n\nBranch\n------\nmaster\n\nDetails\n-------\nhttp://git.postgresql.org/pg/commitdiff/9acb85597f1223ac26a5b19a934584... |
[
{
"msg_contents": "\nWhen consulting the online \"devel\" documentation, for instance for \n\"pgbench\":\n\n \thttps://www.postgresql.org/docs/devel/static/pgbench.html\n\nUnordered lists <ul> are shown in bold, as can be seen in the -M option \nand in the description of random functions. This is because \"ul\"... |
[
{
"msg_contents": "The planner doesn't currently worry about work_mem restrictions when\nplanning a hash join, figuring that the executor should be able to\nsubdivide the data arbitrarily finely by splitting buckets at runtime.\nHowever there's a thread here:\nhttps://www.postgresql.org/message-id/flat/CACw4T0p... |
[
{
"msg_contents": "Extracted from a larger patch, this patch provides the basic infrastructure for turning data\nchecksums off in a cluster. This also sets up the necessary pg_control fields to support the\nnecessary multiple states for handling the transitions.\n\nWe do a few things:\n\n- Change \"data_checks... |
[
{
"msg_contents": "Hi\n\nI want to implement my custom join algorithm into postgres. Where should I\nstart? Where in the code base and how the current algorithms are currently\nimplemented?\n\nHiI want to implement my custom join algorithm into postgres. Where should I start? Where in the code base and how the ... |
[
{
"msg_contents": "Peter,\n\n* Peter Eisentraut (peter_e@gmx.net) wrote:\n> Logical replication\n> \n> - Add PUBLICATION catalogs and DDL\n> - Add SUBSCRIPTION catalog and DDL\n> - Define logical replication protocol and output plugin\n> - Add logical replication workers\n\nI'm not entirely sure about the reaso... |
[
{
"msg_contents": "Buildfarm members gaur, pademelon, and gharial have all recently shown\nfailures like this:\n\n*** /home/bfarm/bf-data/HEAD/pgsql.build/src/test/regress/expected/select_parallel.out\tThu Feb 16 20:35:14 2017\n--- /home/bfarm/bf-data/HEAD/pgsql.build/src/test/regress/results/select_parallel.ou... |
[
{
"msg_contents": "Hi,\n\nSince commit 7c030783a5bd07cadffc2a1018bc33119a4c7505 it seems that $SUBJECT.\n\npg_recvlogical.c:501:37: error: incompatible pointer types passing\n'int64 *' (aka 'long *') to parameter of type 'TimestampTz *' (aka\n'double *') [-Werror,-Wincompatible-pointer-types]\n ... |
[
{
"msg_contents": "In certain cases, pg_dump's dumpTableSchema() emits a separate ALTER TABLE\ncommand for those schema elements of a table that could not be included\ndirectly in the CREATE TABLE command for the table.\n\nFor example:\n\ncreate table p (a int, b int) partition by range (a);\ncreate table p1 pa... |
[
{
"msg_contents": "Hi all,\n\nAs of now, we expand the hash index by doubling the number of bucket\nblocks. But unfortunately, those blocks will not be used immediately.\nSo I thought if we can differ bucket block allocation by some factor,\nhash index size can grow much efficiently. I have written a POC patch\... |
[
{
"msg_contents": "Hi,\n\nWhen working on REINDEX CONCURRENTLY I noticed that the new memory \ncontext created in the ReindexMultipleTables() seems pointless.\n\nThe purpose claimed in the code for introducing the \nReindexMultipleTables context is to make sure the list we build with \nrelation IDs survive the ... |
[
{
"msg_contents": "Thomas Munro pointed out that commit 7c030783a broke things on\n--disable-integer-datetimes builds, because somebody cleverly used\nTimestampTz to declare timestamp variables, no doubt not having\nread the comment (which doesn't even appear in the same file :-()\nthat\n\n * ... The replicatio... |
[
{
"msg_contents": "Here is a patch set to refine various access control settings in logical\nreplication. Currently, you need to be replication or superuser for\nmost things, and the goal of these patches is to allow ordinary users\nequipped with explicit privileges to do most things. (Btw., current\ndocument... |
[
{
"msg_contents": "See below. ISTM that pg_get_object_address should support everything \npg_identify_object_as_address can output, no?\n\nI'm guessing the answer here is to have pg_identify_object_as_address \ncomplain if you ask it for something that's not mapable.\n\n> ~@decina.local/5621# CREATE TYPE comp A... |
[
{
"msg_contents": "I'm confused by this:\n\n\"pg_event_trigger_ddl_commands returns one row for each base command \nexecuted; some commands that are a single SQL sentence may return more \nthan one row.\"\n\nWhat is a \"SQL sentence\"?\n\n(FWIW, I'm wondering because I was just looking to see why there's no \nd... |
[
{
"msg_contents": "Hi\n\nI am checking new Fedora 26, where new gcc compiler is used.\n\nfloat.c: In function ‘float4out’:\nfloat.c:382:41: warning: ‘%.*g’ directive output may be truncated writing\nbetween 1 and 310 bytes into a region of size 65 [-Wformat-truncation=]\n snprintf(ascii, MAXFLOATWIDTH + 1, ... |
[
{
"msg_contents": "Hi all!\n\nI decided to start new thread for this patch for following two reasons.\n * It's renamed from \"Partial sort\" to \"Incremental sort\" per suggestion by\nRobert Haas [1]. New name much better characterizes the essence of\nalgorithm.\n * I think it's not PoC anymore. Patch receive... |
[
{
"msg_contents": "While experimenting with enabling -Werror in the buildfarm, I got\nannoyed about the fact that we have to apply -Wno-error while\nbuilding some of the flex scanners, because with flex versions\nbefore 2.5.36 you get\n\nscan.c: In function 'yy_try_NUL_trans':\nscan.c:10317: warning: unused var... |
[
{
"msg_contents": "Hi,\n\nI would like to work on a patch to accommodate restricted environments\n(such as AWS RDS Postgres) which don't allow pg_authid access since their\ndefinition of Superuser is just a regular user with extra permissions.\n\nWould you consider a patch to add a flag to work around this rest... |
[
{
"msg_contents": "Both the streaming replication and logical replication areas of the code\nare, approximately, utterly broken when !HAVE_INT64_TIMESTAMPS. (The fact\nthat \"make check-world\" passes anyway is an indictment of the quality of\nthe regression tests.)\n\nI started poking around in this area afte... |
[
{
"msg_contents": "Hi hackers:\r\n When using pg_stat_statements to collect running SQL of PG, we find it is hard for our program to get exact operation type of the SQL, such as SELECT, DELETE, UPDATE, INSERT, and so on.\r\n So we modify the the source code of pg_stat_statements and add another outpu... |
[
{
"msg_contents": "Hi hackers:\r\n When using pg_stat_statements to collect running SQL of PG, we find it is hard for our program to get exact operation type of the SQL, such as SELECT, DELETE, UPDATE, INSERT, and so on.\r\n So we modify the the source code of pg_stat_statements and add another outpu... |
[
{
"msg_contents": "Hi hackers:\r\n When using pg_stat_statements to collect running SQL of PG, we find it is hard for our program to get exact operation type of the SQL, such as SELECT, DELETE, UPDATE, INSERT, and so on.\r\n So we modify the the source code of pg_stat_statements and add another outpu... |
[
{
"msg_contents": "Hi PG hackers:\r\n When using pg_stat_statements to collect running SQL of PG, we find it is hard for our program to get exact operation type of the SQL, such as SELECT, DELETE, UPDATE, INSERT, and so on.\r\n So we modify the the source code of pg_stat_statements and add another ou... |
[
{
"msg_contents": "Hi PG hackers: When using pg_stat_statements to collect running SQL of PG, we find it is hard for our program to get exact operation type of the SQL, such as SELECT, DELETE, UPDATE, INSERT, and so on. So we modify the the source code of pg_stat_statements and add another output par... |
[
{
"msg_contents": "On Thu, Feb 16, 2017 at 7:15 AM, Amit Langote\n<Langote_Amit_f8@lab.ntt.co.jp> wrote:\n> Also attaching 0002 (unchanged) for tab-completion support for the new\n> partitioning syntax.\n\nAt one point you have this:\n\n+ /* Limited completion support for partition bound specification */\n+ ... |
[
{
"msg_contents": "Over in the \"Keeping pg_recvlogical's \"feTimestamp\" separate from \nTimestampTz\"...\n\nOn 2/17/17 12:15 PM, Tom Lane wrote:\n > I am not sure that it was really a good design to pretend that the\n > replication protocol is independent of --disable-integer-datetimes\n > when the underlying... |
[
{
"msg_contents": "Hi,\n\nAttached is a patch that fixes a comment typo in varlena.c\n\n\nThanks,\nNeha\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.org/mailpref/pgsql-hackers",
"msg_date": "Mon, 20 Feb 2017 09:34:... |
[
{
"msg_contents": "Re-posting the patch I posted in a nearby thread [0].\n\nOn 2017/02/16 2:08, Robert Haas wrote:\n> On Wed, Feb 15, 2017 at 11:34 AM, Alvaro Herrera\n> <alvherre@2ndquadrant.com> wrote:\n>> I think new-style partitioning is supposed to consider each partition as\n>> an implementation detail of... |
[
{
"msg_contents": "I noticed while researching bug #14555 that fd.c contains two separate\ncases like\n\n\tvfdP->seekPos = lseek(vfdP->fd, (off_t) 0, SEEK_CUR);\n\tAssert(vfdP->seekPos != (off_t) -1);\n\nThis seems, um, unwise. It might somehow fail to fail in production\nbuilds, because elsewhere it's assumed... |
[
{
"msg_contents": "Hello everyone,\n\nI am writing a custom logical replication decoder, and I took test decoder\nfrom Postgres sources as an example.\n\nCould anyone tell me how to read \"unchanged toast datum\" in case it is\nVARTT_IS_EXTERNAL_ONDISK.\n\nIn the test decoder it is ignored:\n\nhttps://github.co... |
[
{
"msg_contents": "Hi\n\nFollowing various conversations on list and in person, including the\ndeveloper meeting in Brussels earlier this month, here is a patch that\nimplements pg_ls_logdir() and pg_ls_waldir() functions.\n\nThe ultimate aim of this (and followup work I'll be doing) is to\nprovide functionalit... |
[
{
"msg_contents": "Further to the patch I just submitted\n(https://www.postgresql.org/message-id/CA%2BOCxow-X%3DD2fWdKy%2BHP%2BvQ1LtrgbsYQ%3DCshzZBqyFT5jOYrFw%40mail.gmail.com)\nI'd like to propose the addition of a default role, pg_monitor.\n\nThe intent is to make it easy for users to setup a role for fully\n... |
[
{
"msg_contents": "The point/polygon regression tests have started to fail on 32-bit\npowerpc on Debian Jessie. So far I could reproduce the problem with\nPostgreSQL 9.4.10+11 and 9.6.1, on several different machines. Debian\nunstable is unaffected.\n\nThe failure looks like this:\n\nhttps://buildd.debian.org/s... |
[
{
"msg_contents": "Hi\n\nWhen I did tests of own XML import functions I found a strange error.\n\nI successfully imported XML to PostgreSQL. This document is readable\nwithout any visual defects. But when I tested this document against any\nlibxml2 function I found a error -\n\nERROR: could not parse XML docum... |
[
{
"msg_contents": "Hello.\n\nI've just tried to build PostgreSQL with Clang 3.9.1 (default version\ncurrently available in Arch Linux) and noticed that it outputs lots of\nwarning messages. Most of them are result of a bug in Clang itself:\n\n```\npostinit.c:846:3: note: include the header <string.h> or explici... |
[
{
"msg_contents": "Hi,\n\nWhen building with a new-ish gcc (6.3.0 right now, but I've seen this\nfor a while) with optimization I get a number of warnings:\n\nIn file included from /home/andres/src/postgresql/src/include/postgres.h:48:0,\n from /home/andres/src/postgresql/src/backend/parser/pars... |
[
{
"msg_contents": "I've come across a number of times where the statistics on materialized\nviews become stale producing bad plans. It turns out that autovacuum only\ntouches a materialized view when it is first created and ignores it on a\nrefresh. When you have a materialized view like yesterdays_sales the da... |
[
{
"msg_contents": "Hi,\n\nwhile investigating some checksum-related issues, I needed to perform \nsome forensics on a copy of a btree page (taken from another instance \nusing 'dd').\n\nBut I've ran into two pageinspect limitations, hopefully addressed by \nthis patch:\n\n1) bt_page_items(bytea) not defined\n\n... |
[
{
"msg_contents": "Hi\n\nToday I played with xml_recv function and with xml processing functions.\n\nxml_recv function ensures correct encoding from document encoding to server\nencoding. But the decl section holds original encoding info - that should\nbe obsolete after encoding. Sometimes we solve this issue b... |
[
{
"msg_contents": "Hi.\nWhile verifying the logical replication of PostgreSQL 10-devel, I found the\nfollowing problem.\n\n* When you run the SUBSCRIPTION command against a table in the same\nPostgreSQL server, hang up.\n* Canceling the hung SUBSCRIPTION command with CTRL + C causes a server\nprocess occurs Seg... |
[
{
"msg_contents": "Simon pointed out in a nearby thread [0] that the detail part of\npartition-not-found error should show just the partition keys. I posted a\npatch on that thread [1], but to avoid confusion being caused by multitude\nof patches over there I'm re-posting it here.\n\n* What the patch does:\n\n... |
[
{
"msg_contents": "Hello,\n\nThis is just a correction from \"index\" to \"table\". I was a bit confused when I first read this part.\n\n\nRegards\nTakayuki Tsunakawa\n\n\n\n\n-- \nSent via pgsql-hackers mailing list (pgsql-hackers@postgresql.org)\nTo make changes to your subscription:\nhttp://www.postgresql.o... |
[
{
"msg_contents": "Hi,\n\ncurrently, the backup_label and (I think) the tablespace_map files are\n(by design) conveniently located at the beginning of the main tablespace\ntarball when making a basebackup. However, (by accident or also by\ndesign?) the main tablespace is the last tarball[1] to be streamed via\n... |
[
{
"msg_contents": "Greetings,\n\nWhen pg_upgrade calls pg_dump, it passes in \"--schema-only\", which is\ngenerally correct, except that this causes everything having to do with\nlarge objects to be excluded. That's still usually correct, because\npg_upgrade will simply copy the pg_largeobject and\npg_largeobj... |
[
{
"msg_contents": "I think $SUBJECT is a bug. While it would be unusual for an extension to \nhave a matview, it's still allowed, and as it stands right now that view \nwon't be refreshed at the end of a restore, unlike other matviews.\n\nI started looking into a patch for this, but I'm not sure of the best \nw... |
[
{
"msg_contents": "AFAICT in older versions only object types that absolutely had to wait \nfor DO_POST_DATA_BOUNDARY would do so. More recently though, objects are \nbeing added after that (presumably because it's easier than renumbering \neverything in dbObjectTypePriority).\n\nIs this change a good or bad id... |
[
{
"msg_contents": "The attached tiny patch lets the to_json(b) routines convert objects\nwith array pseudotypes to json{b}. The main impetus for this came from\nmy trying to convert the annyarray columns in pg_stats to json and\nfinding that they ended up as text instead of json arrays. This way\nto_json on the... |
[
{
"msg_contents": "Ashutosh Bapat pointed out [0] that regression tests are missing for the\nforeign partition DDL commands. Attached patch takes care of that.\n\nThanks,\nAmit\n\n[0]\nhttps://www.postgresql.org/message-id/CAFjFpRcrdzBRj0cZ%2BJAQmfSa2Tv8wSEcWAeYtDpV-YZnNna2sA%40mail.gmail.com\n\n\n\n-- \nSent ... |
[
{
"msg_contents": "Last year I have proposed an enhancement regarding behavior of the\nstatement timeout in extended queries.\n\nhttps://www.postgresql.org/message-id/20160528.220442.1489791680347556026.t-ishii%40sraoss.co.jp\n\nIMO the current behavior is counter intuitive and I would like to\nchange it toward... |
[
{
"msg_contents": "Hello everybody,\n\nIn the current version, queries in SQL or PL functions can not\nleverage parallelism. To relax this restriction I prepared a patch,\nthe approach used in the patch is explained next,\n\nApproach:\n\n1. Allow parallelism for queries in PL functions by passing\nCURSOR_OPT_PA... |
[
{
"msg_contents": "Hello everyone,\n\nWhile conducting the experiments for parallelism, Rafia came across a\nhang in Q18 when plan uses partial and finalize hash aggregate. This\ncould be seen on both scale factors - 20 and 300, on setting work_mem\nhigh enough so that the query uses hash aggregate. It seems th... |
[
{
"msg_contents": "Hi Hackers,\n\nCurrently, it's only possible to grant/revoke execute on functions to roles.\n\nI think it would be useful in many situations, both for documentation purposes,\nbut also for increased security, to in a precise way control what\nother function(s)\nare allowed to execute a specif... |
[
{
"msg_contents": "Attached patch fixes an oversight that tablesample cannot be used with\npartitioned tables:\n\ncreate table p (a int) partition by list (a);\nselect * from p tablesample bernoulli (50);\nERROR: TABLESAMPLE clause can only be applied to tables and materialized\nviews\n\nThanks,\nAmit\n\n\n\n-... |
[
{
"msg_contents": "Hello All,\n\nI would like to propose the attached patch which removes all direct\nreferences to ip_posid and ip_blkid members of ItemPointerData struct and\ninstead use ItemPointerGetOffsetNumber and ItemPointerGetBlockNumber macros\nrespectively to access these members.\n\nMy motivation to ... |
[
{
"msg_contents": "There is support for COMMENT ON RULE <name> without specifying a table\nname, for upgrading from pre-7.3 instances. I think it might be time\nfor that workaround to go.\n\nPatch attached.\n\n-- \nPeter Eisentraut http://www.2ndQuadrant.com/\nPostgreSQL Development, 24x7 Support,... |
[
{
"msg_contents": "From time to time, especially during migration projects from Oracle to\nPostgreSQL, i'm faced with people questioning why the alias in the FROM\nclause for subqueries in PostgreSQL is mandatory. The default answer\nhere is, the SQL standard requires it.\n\nThis also is exactly the comment in ... |
[
{
"msg_contents": "I looked into the report in bug #14563 about pg_trgm giving wrong\nanswers for regexp searches. There is a whole lot of complex mechanism\nin trgm_regexp.c for extracting trigram patterns out of regexps, and\nso I was afraid that it might be a very subtle bug, but actually the\nproblem seems... |
[
{
"msg_contents": "The topic has been previously discussed[0] on the -performance mailing list,\nabout four years ago.\n\nIn that thread, Tom suggested[0] the planner could be made to \"expand\n\"intcol <@\n'x,y'::int4range\" into \"intcol between x and y\", using something similar\nto the\nindex LIKE optimizat... |
[
{
"msg_contents": "pg_get_object_address() currently returns a field called subobjid, while \npg_depend calls that objsubid. I'm guessing that wasn't on purpose \n(especially because internally the function uses objsubid), and it'd be \nnice to fix it.\n\nAttached does that, as well as updating the input naming... |
[
{
"msg_contents": "Hi Hackers,\n\nI have noticed the following behaviour in range partitioning which i felt\nis not quite correct (i missed reporting this) -\n\nI have tested by creating a date ranged partition.\n\nI created the following table.\n\ndb03=# CREATE TABLE orders (\n o_orderkey INTEGER,\n ... |
[
{
"msg_contents": "Hi all,\n\nWhen storing WAL segments on a dedicated partition with\npg_receivexlog, for some deployments, the removal of past WAL segments\ndepends on the frequency of base backups happening on the server. In\nshort, once a new base backup is taken, it may not be necessary to\nkeep around tho... |
[
{
"msg_contents": "Hi\n\nI propose a schema private functions as analogy to Oracle package functions.\n\nMy target of this proposal is better isolation generally available top\nlevel callable functions from other auxiliary functions. A execution of\nfunctions can be little bit more robust due less dependency on... |
[
{
"msg_contents": "Have a nice day\nI`m using \\pset format html for generating tables for some documentation,\ntraining, reports, blogs... Mainly sample of data for SQL queries, or\nexplaining structure etc.\nBut in some software like sphinx, wiki on redmine, or markdown on github\nissues I`m not able to inser... |
[
{
"msg_contents": "Hi\n\nCurrently is not possible to control sort columns for \\d* commands. Usually\nschema and table name is used. Really often task is collect the most big\nobjects in database. \"\\dt+, \\di+\" shows necessary information, but not in\npractical order.\n\nInstead introduction some additional... |
[
{
"msg_contents": "Hello, while I investigated a bug of async-exec patch. I found a\n(maybe) typo in mcxt.c.\n\n| * It's not entirely clear whether 'tis better to do this before or after\n| * delinking the context; but an error in a callback will likely result in\n\nI'm not so confident, but the \"'tis\" seem... |
[
{
"msg_contents": "Andrew Dunstan wrote:\n> \n> While reviewing the IF NOT EXISTS patch for CREATE USER MAPPING I\n> noticed that in several places we treat the user name as the name of the\n> user mapping. Strictly ISTM that user mappings are really anonymous\n> objects, so instead of something like user \"map... |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.