issue_owner_repo listlengths 2 2 | issue_body stringlengths 0 261k ⌀ | issue_title stringlengths 1 925 | issue_comments_url stringlengths 56 81 | issue_comments_count int64 0 2.5k | issue_created_at stringlengths 20 20 | issue_updated_at stringlengths 20 20 | issue_html_url stringlengths 37 62 | issue_github_id int64 387k 2.46B | issue_number int64 1 127k |
|---|---|---|---|---|---|---|---|---|---|
[
"MonetDB",
"MonetDB"
] | Date: 2009-10-12 02:43:49 +0200
From: @swingbit
To: XQuery devs <<bugs-xquery>>
Version: -- development
Last updated: 2011-04-28 16:38:58 +0200
## Comment 13913
Date: 2009-10-12 14:43:49 +0200
From: @swingbit
Currently, runtime errors are reported as an error code + textual description.
Often this is not enough, especially because the error refers to algebra expressions that could have been generated by the compiler, so they are hard to find in the query text.
Example:
ERROR = !ERROR: err:FORG0003, fn:zero-or-one called with a sequence containing more than one item.
With enough experience, most of the times it is possible to find which expression in the original query generated this,
although there is no explicit zero-or-one() call. However, when the query is non-trivial, this becomes a real pain.
If it is too hard to trace back the exact line in the query text (which should be possible anyway), a printout of the algebra expression being evaluated, whith some context around, would certainly help to quickly spot the error.
## Comment 13914
Date: 2009-10-13 00:03:42 +0200
From: Jan Rittinger <<jan.rittinger>>
As far as I can oversee almost nobody would understand the algebra plan in a textual representation. In most cases it's not easy to digest.
The best solution would be to give an estimate position with respect to the input query. Currently we loose the position information before the translation to our XQuery Core representation -- probably because due to rewrites positions might be misleading. Necessary changes to make 'assumptions on the position' would include making the Core representation location-aware, keeping positions over rewrites, and incorporating location information in the core2algebra translation...
## Comment 13915
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2877036 at http://sourceforge.net/support/tracker.php?aid=2877036
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-xquery@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 15770
Date: 2011-04-28 16:38:58 +0200
From: @sjoerdmullender
Development on MonetDB4 and MonetDB/XQuery has ceased. It is unlikely that this bug will be fixed.
The last release of MonetDB/XQuery and MonetDB4 is Mar2011.
| PF: more useful error reports | https://api.github.com/repos/MonetDB/MonetDB/issues/2510/comments | 0 | 2020-11-30T11:00:38Z | 2024-06-28T07:23:08Z | https://github.com/MonetDB/MonetDB/issues/2510 | 753,377,158 | 2,510 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-10-09 03:03:48 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2009-12-04 11:21:14 +0100
## Comment 13908
Date: 2009-10-09 03:03:48 +0200
From: @skinkie
The sum of an crossproduct is equal to the sum of the table multiplicated by the amount of joins. Hence we can calculate the sum without doing the actual cross product and summing over the result.
## Comment 13909
Date: 2009-12-01 09:51:17 +0100
From: @grobian
wasn't this optimisation applied?
## Comment 13910
Date: 2009-12-01 09:56:11 +0100
From: @skinkie
What is written in the summary is. But since there was one rewrite option left I didn't close it.
## Comment 13911
Date: 2009-12-04 11:21:13 +0100
From: @mlkersten
A too specific optimization, which would call for a math reasoner
within the SQL front-end.
It won't be added in the foreseeable future.
## Comment 13912
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2875295 at http://sourceforge.net/support/tracker.php?aid=2875295
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| SQL: rel_optimizer: apply sum calculation using n-fold mult | https://api.github.com/repos/MonetDB/MonetDB/issues/2509/comments | 0 | 2020-11-30T11:00:35Z | 2024-06-28T07:23:08Z | https://github.com/MonetDB/MonetDB/issues/2509 | 753,377,121 | 2,509 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-10-09 02:58:31 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
CC: holger.pirk, @yzchang
Last updated: 2016-04-11 11:45:06 +0200
## Comment 13906
Date: 2009-10-09 02:58:31 +0200
From: @skinkie
=project (
=| group by (
=| | crossproduct (
=| | | table(sys.tralala) [ tralala.test as a.test, tralala.%TID% NOT NULL as a.%TID% ],
=| | | table(sys.tralala) [ tralala.%TID% NOT NULL ]
=| | )
=| ) [ ] [ max no nil (a.test) as L1 ]
=) [ L1 ]
=>
=project (
=| group by (
=| | table(sys.tralala) [ tralala.test as a.test, tralala.%TID% NOT NULL as a.%TID% ]
=| ) [ ] [ max no nil (a.test) as L1 ]
=) [ L1 ]
The general solution should push down the aggregates min/max down to the tables where they operate upon and purge the remaining tables.
## Comment 13907
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2875293 at http://sourceforge.net/support/tracker.php?aid=2875293
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 18027
Date: 2012-11-27 11:07:45 +0100
From: Holger <<holger.pirk>>
This is a very specific case that doesn't seem very important. This can be fixed by simply rewriting the query.
## Comment 18144
Date: 2012-11-27 16:23:16 +0100
From: @yzchang
No test needed for feature request
| SQL:rel_optimizer apply min/max only on source table in join | https://api.github.com/repos/MonetDB/MonetDB/issues/2508/comments | 0 | 2020-11-30T11:00:32Z | 2024-06-28T07:19:39Z | https://github.com/MonetDB/MonetDB/issues/2508 | 753,377,095 | 2,508 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-10-05 02:05:52 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
CC: dennis
Last updated: 2016-04-11 11:46:07 +0200
## Comment 13902
Date: 2009-10-05 14:05:52 +0200
From: @skinkie
PostgreSQL implements a nice number of extract features that can be used as DATEPART. We don't have these functions yet so it could be useful to implement them. (I'll volunteer)
http://www.postgresql.org/docs/7.4/static/functions-datetime.html (section 9.8.1)
## Comment 13903
Date: 2009-12-01 09:52:55 +0100
From: @grobian
This seems like a useful addition to the codebase to me
## Comment 13904
Date: 2009-12-01 11:13:59 +0100
From: @grobian
we should at least implement EPOCH (1752627)
## Comment 13905
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2872897 at http://sourceforge.net/support/tracker.php?aid=2872897
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 15155
Date: 2010-10-28 15:44:28 +0200
From: @grobian
I'm affraid we won't make the EPOCH implementation for Oct2010
## Comment 16034
Date: 2011-07-29 11:00:03 +0200
From: @sjoerdmullender
Apr2011-SP2 has been released.
## Comment 16272
Date: 2011-09-16 15:10:37 +0200
From: @sjoerdmullender
The Aug2011 version has been released.
## Comment 17676
Date: 2012-08-24 14:55:55 +0200
From: @sjoerdmullender
Jul2012-SP1 has been released.
## Comment 19905
Date: 2014-07-16 11:09:05 +0200
From: Dennis Pallett <<dennis>>
Created attachment 290
UNIX_TIMESTAMP SQL function to convert a timestamp into an epoch
> Attached file: [unix_timestamp_ddl.sql](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_2507_unix_timestamp_ddl.sql_290) (text/plain, 151 bytes)
> Description: UNIX_TIMESTAMP SQL function to convert a timestamp into an epoch
## Comment 19906
Date: 2014-07-16 11:10:36 +0200
From: Dennis Pallett <<dennis>>
I would also like to have this included into a future release of MonetDB. At the moment I'm using a simple SQL function that converts a timestamp into an EPOCH timestamp but it'd be better if it's properly implemented as part of the EXTRACT function.
| SQL: Implement complementary EXTRACT functionality | https://api.github.com/repos/MonetDB/MonetDB/issues/2507/comments | 0 | 2020-11-30T11:00:29Z | 2024-06-28T07:19:39Z | https://github.com/MonetDB/MonetDB/issues/2507 | 753,377,068 | 2,507 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-10-02 12:53:44 +0200
From: @skinkie
To: @njnes
Version: 1.40.1 (Oct2010) [obsolete]
Last updated: 2011-01-25 15:32:51 +0100
## Comment 13886
Date: 2009-10-02 00:53:44 +0200
From: @skinkie
While loading: http://mirror.openstreetmap.nl/knmi/KNMI.sql
I see:
MAPI = monetdb@localhost:50001
QUERY = DEZE GEGEVENS MOGEN VRIJ WORDEN GEBRUIKT MITS DE VOLGENDE BRONVERMELDING WORDT GEGEVEN:
KONINKLIJK NEDERLANDS METEOROLOGISCH INSTITUUT (KNMI)
THESE DATA CAN BE USED FREELY PROVIDED THAT THE FOLLOWING SOURCE IS ACKNOWLEDGED:
ROYAL NETHERLANDS METEOROLOGICAL INSTITUTE
DROP TABLE stations;
CREATE TABLE stations (stationid integer, long decimal(5,2), lat decimal(5,2), altitude smallint, name varchar(128), primary key(stationid));
COPY 36 RECORDS INTO stations from stdin USING DELIMITERS '\t', '\n'
ERROR = !DROP TABLE: no such table 'stations'
Obviously the right query is not selected.
## Comment 13887
Date: 2009-10-02 01:25:10 +0200
From: @mlkersten
The conclusion that there is here a MonetDB bug can not be deciphered from this report. Of course it works when you have auto-commit enabled. Moreover, mapi client throws the complete sql script to the server when you pass it as an argument to mclient. If you need line-based errors, then use either <inputfile or the -i option.
## Comment 13888
Date: 2009-10-02 02:21:02 +0200
From: @mlkersten
mclient -lsql -i KNMI.sql
Welcome to mclient, the MonetDB/SQL interactive terminal
Database: MonetDB v5.15.0, 'demo'
Type \q to quit, \? for a list of available commands
auto commit mode: on
DROP TABLE: no such table 'stations'
DROP TABLE: no such table 'stations'
36 affected rows
DROP TABLE: no such table 'knmi_tmp'
544054 affected rows
DROP TABLE: no such table 'knmi'
544054 affected rows
sql>
## Comment 13889
Date: 2009-10-02 08:23:28 +0200
From: @grobian
You are probably using the candidate release (Nov2009) or current, and are piping the output into mclient, or making mclient read from file. Our server doesn't report its state, so we can't do more in that case than we do right now. If you want mclient to process line by line, use the -i flag.
It is a feature request to have (any) client to report exactly what went wrong where, based on a state inside the server.
## Comment 13890
Date: 2009-10-02 10:15:41 +0200
From: @drstmane
Like others, I could not get this from the original posting of this bug report --- nor can I tell, whether this was the problem that the original poster intended to report ---, but there is indeed an error/bug involved:
The correct/expected first error message (DROP TABLE: no such table 'stations') is issued a second time, as soon as the first query result is returned/printed. Looks like some error/output buffer is not initialized or cleared correctly.
Might be a bug in SQL, but might also be a bug in MAPI or mclient.
## Comment 13891
Date: 2009-10-02 11:10:46 +0200
From: @drstmane
The double (first, only) error message also occurs with the Aug2009 release branch.
As said, I'm not sure, whether it is a SQL, MAPI, or mclient problem ...
## Comment 13892
Date: 2009-10-02 11:25:07 +0200
From: @grobian
stmane, how do you reproduce that error? It works fine for me:
sql>drop table test;
DROP TABLE: no such table 'test'
sql>select 1;
+--------------+
| single_value |
+==============+
| 1 |
+--------------+
1 tuple
sql>
## Comment 13893
Date: 2009-10-02 11:30:41 +0200
From: @mlkersten
Download the KNMI.sql file and issue the command
mclient -lsql -i KNMI.sql
## Comment 13894
Date: 2009-10-02 11:32:13 +0200
From: @grobian
without -i I see this:
% $INSTALL_DIR/bin/mclient -dslave ~/tmp/KNMI.sql
MAPI = monetdb@localhost:50000
QUERY = DEZE GEGEVENS MOGEN VRIJ WORDEN GEBRUIKT MITS DE VOLGENDE BRONVERMELDING WORDT GEGEVEN:
KONINKLIJK NEDERLANDS METEOROLOGISCH INSTITUUT (KNMI)
THESE DATA CAN BE USED FREELY PROVIDED THAT THE FOLLOWING SOURCE IS ACKNOWLEDGED:
ROYAL NETHERLANDS METEOROLOGICAL INSTITUTE
DROP TABLE stations;
CREATE TABLE stations (stationid integer, long decimal(5,2), lat decimal(5,2), altitude smallint, name varchar(128), primary key(stationid));
COPY 36 RECORDS INTO stations from stdin USING DELIMITERS '\t', '\n'
ERROR = !DROP TABLE: no such table 'stations'
MAPI = monetdb@localhost:50000
QUERY = DEZE GEGEVENS MOGEN VRIJ WORDEN GEBRUIKT MITS DE VOLGENDE BRONVERMELDING WORDT GEGEVEN:
KONINKLIJK NEDERLANDS METEOROLOGISCH INSTITUUT (KNMI)
THESE DATA CAN BE USED FREELY PROVIDED THAT THE FOLLOWING SOURCE IS ACKNOWLEDGED:
ROYAL NETHERLANDS METEOROLOGICAL INSTITUTE
DROP TABLE stations;
CREATE TABLE stations (stationid integer, long decimal(5,2), lat decimal(5,2), altitude smallint, name varchar(128), primary key(stationid));
COPY 36 RECORDS INTO stations from stdin USING DELIMITERS '\t', '\n'
ERROR = !DROP TABLE: no such table 'stations'
!MALException:!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!ERROR: madvise on (&b->T->heap) tb failed.
!OS: Invalid argument
!COMMIT: failed
!ERROR: COMMIT: transaction is aborted, will ROLLBACK instead
!DROP TABLE: no such table 'knmi'
[ 36 ]
MAPI = monetdb@localhost:50000
QUERY = DEZE GEGEVENS MOGEN VRIJ WORDEN GEBRUIKT MITS DE VOLGENDE BRONVERMELDING WORDT GEGEVEN:
KONINKLIJK NEDERLANDS METEOROLOGISCH INSTITUUT (KNMI)
THESE DATA CAN BE USED FREELY PROVIDED THAT THE FOLLOWING SOURCE IS ACKNOWLEDGED:
ROYAL NETHERLANDS METEOROLOGICAL INSTITUTE
DROP TABLE stations;
CREATE TABLE stations (stationid integer, long decimal(5,2), lat decimal(5,2), altitude smallint, name varchar(128), primary key(stationid));
COPY 36 RECORDS INTO stations from stdin USING DELIMITERS '\t', '\n'
ERROR = !DROP TABLE: no such table 'knmi_tmp'
MAPI = monetdb@localhost:50000
QUERY = DEZE GEGEVENS MOGEN VRIJ WORDEN GEBRUIKT MITS DE VOLGENDE BRONVERMELDING WORDT GEGEVEN:
KONINKLIJK NEDERLANDS METEOROLOGISCH INSTITUUT (KNMI)
THESE DATA CAN BE USED FREELY PROVIDED THAT THE FOLLOWING SOURCE IS ACKNOWLEDGED:
ROYAL NETHERLANDS METEOROLOGICAL INSTITUTE
DROP TABLE stations;
CREATE TABLE stations (stationid integer, long decimal(5,2), lat decimal(5,2), altitude smallint, name varchar(128), primary key(stationid));
COPY 36 RECORDS INTO stations from stdin USING DELIMITERS '\t', '\n'
ERROR = !SELECT: no such table 'knmi_tmp'
!current transaction is aborted (please ROLLBACK)
!COMMIT: failed
!ERROR: COMMIT: transaction is aborted, will ROLLBACK instead
sql>
## Comment 13895
Date: 2009-10-02 11:34:04 +0200
From: @grobian
with -i:
% $INSTALL_DIR/bin/mclient -i -dslave ~/tmp/KNMI.sql
Welcome to mclient, the MonetDB/SQL interactive terminal
Database: MonetDB v5.16.0, 'slave'
Type \q to quit, \? for a list of available commands
auto commit mode: on
DROP TABLE: no such table 'stations'
DROP TABLE: no such table 'stations'
MALException:!ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
ERROR: madvise on (&b->T->heap) tb failed.
OS: Invalid argument
COMMIT: failed
ERROR: COMMIT: transaction is aborted, will ROLLBACK instead
DROP TABLE: no such table 'knmi'
36 affected rows
DROP TABLE: no such table 'knmi_tmp'
SELECT: no such table 'knmi_tmp'
current transaction is aborted (please ROLLBACK)
COMMIT: failed
ERROR: COMMIT: transaction is aborted, will ROLLBACK instead
sql>
there is way to much crap here for me to debug/see the right problem.
## Comment 13896
Date: 2009-10-02 11:34:27 +0200
From: @grobian
the server actually crashes even
## Comment 13897
Date: 2009-10-02 13:00:44 +0200
From: @grobian
Turning this one into a feature request again. After talking to Stefan (dK) it seems that what he expects as output would be:
MAPI = monetdb@localhost:50001
QUERY = DROP TABLE stations;
ERROR = !DROP TABLE: no such table 'stations'
This is just the conclusion I made earlier in this bug.
We will have to make separate bugs for all issues encountered here that we noticed, as per the "one issue per bug is enough" rule.
## Comment 13898
Date: 2009-12-04 16:18:42 +0100
From: @mlkersten
This script works on the current version.
MonetDB/SQL module v2.35.0 loaded
!DROP TABLE: no such table 'stations'
&3
&2 36 -1
!DROP TABLE: no such table 'knmi_tmp'
&4 f
&3
&2 544054 -1
&4 t
!DROP TABLE: no such table 'knmi'
&4 f
&3
&2 544054 -1
&3
&4 t
## Comment 13899
Date: 2009-12-04 17:08:34 +0100
From: @grobian
this feature request is about the server sending back the query that failed, not about a loading problem.
## Comment 13900
Date: 2009-12-23 19:24:33 +0100
From: Mark Bucciarelli <<mkbucc>>
ref: Foreign key violation error message gives wrong record - ID: 2920022
At least change the error message to say something like "somewhere around" or "somewhere after" or better yet, just don't output any sql at all from the buffer. It is misdirection and confusing to someone trying to figure out which exact SQL statement caused the error.
## Comment 13901
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2871572 at http://sourceforge.net/support/tracker.php?aid=2871572
## Comment 15434
Date: 2011-01-25 15:32:51 +0100
From: @grobian
we did some improvements here, but it will never be perfect, since we don't want to pay the costs of maintaining this in the server side.
| mclient: QUERY does not report the actual query on error | https://api.github.com/repos/MonetDB/MonetDB/issues/2506/comments | 0 | 2020-11-30T11:00:25Z | 2024-06-28T07:23:08Z | https://github.com/MonetDB/MonetDB/issues/2506 | 753,377,018 | 2,506 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-09-29 05:19:03 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2009-12-04 11:27:24 +0100
## Comment 13880
Date: 2009-09-29 17:19:03 +0200
From: @skinkie
I think it must be possible for a query like:
select count(*) from bigdatabase1, bigdatabase2;
To rewrite it to:
select (select count(*) from bigdatabase1) * (select count(*) from bigdatabase2);
## Comment 13881
Date: 2009-10-09 02:30:14 +0200
From: @skinkie
Because I am seriously proud... I want to publish the patch first :D
select count(*) from kvk, kvk;
% . table_name
% L60 name
% wrd type
% 13 length
[ 5830679502400 ]
Timer 2.650 msec 1 rows
We have to work on some sort of overflow protection too... but that was already in the /real/ bugtracker :)
## Comment 13882
Date: 2009-12-01 09:55:46 +0100
From: @grobian
Sounds very similar to 2875295, is there a big difference?
## Comment 13883
Date: 2009-12-01 10:01:55 +0100
From: @skinkie
This one is on the count, and is partly implemented. The example above here works. The other one is that the sum of the cross product equals the sum of one multiplied by the amount of joins to come to the cross product itself.
So count(*)^N vs sum(something)*N
## Comment 13884
Date: 2009-12-04 11:27:24 +0100
From: @mlkersten
Too specific transformation. Calls for a math reasoner
in the optimizer.
Won't be implemented in the foreseable future.
## Comment 13885
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2869851 at http://sourceforge.net/support/tracker.php?aid=2869851
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| SQL: Cross product count, algebra | https://api.github.com/repos/MonetDB/MonetDB/issues/2505/comments | 0 | 2020-11-30T11:00:22Z | 2024-06-28T07:23:08Z | https://github.com/MonetDB/MonetDB/issues/2505 | 753,376,984 | 2,505 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-09-10 10:26:46 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2009-12-01 09:57:01 +0100
## Comment 13877
Date: 2009-09-10 10:26:46 +0200
From: @grobian
It's a big mess with memory management, currently three systems in use, GDK, merovingian's own, and plain malloc + free. We should get rid of all but the last, and also make Sabaoth have a multi-headed interface to have merovingian and monetdb really GDK free.
## Comment 13878
Date: 2009-12-01 09:57:00 +0100
From: @grobian
unfortunately this is not really possible, so we're stuck with GDK until that beast is detangled from the rest of the codebase as well.
## Comment 13879
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2855855 at http://sourceforge.net/support/tracker.php?aid=2855855
| merovingian: clean from GDK{free,malloc,strdup} usage | https://api.github.com/repos/MonetDB/MonetDB/issues/2504/comments | 0 | 2020-11-30T11:00:19Z | 2024-06-28T07:23:07Z | https://github.com/MonetDB/MonetDB/issues/2504 | 753,376,952 | 2,504 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-09-10 10:08:01 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2009-11-30 10:27:26 +0100
## Comment 13868
Date: 2009-09-10 10:08:01 +0200
From: @grobian
Currently, some monetdb commands give confirmative output like 'database X is now under maintenance', while others remain silent. For consistency it would be better if all commands return confirmative answers that something has taken place (and not e.g. the tool crashed or failed to recognise the command). Then to make it silent (better in script environments) a global option -q should be added to suppress all these messages.
## Comment 13869
Date: 2009-09-25 11:21:38 +0200
From: @grobian
implemented for next feature release
## Comment 13870
Date: 2009-10-10 04:22:05 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13871
Date: 2009-10-25 03:20:15 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13872
Date: 2009-11-09 03:20:53 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13873
Date: 2009-11-24 03:20:20 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13874
Date: 2009-11-24 08:44:06 +0100
From: @grobian
patience...
## Comment 13875
Date: 2009-11-30 10:27:26 +0100
From: @grobian
This feature is available starting from the Nov2009 release
## Comment 13876
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2855846 at http://sourceforge.net/support/tracker.php?aid=2855846
| monetdb: add -q option and make output consistent | https://api.github.com/repos/MonetDB/MonetDB/issues/2503/comments | 0 | 2020-11-30T11:00:16Z | 2024-06-28T07:23:07Z | https://github.com/MonetDB/MonetDB/issues/2503 | 753,376,918 | 2,503 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-08-01 11:33:09 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2009-11-30 10:26:32 +0100
## Comment 13857
Date: 2009-08-01 11:33:09 +0200
From: @grobian
Instead of just allowing local access (and using the trust relationship by the OS) allow monetdb to control remote merovingians. Needs a proper security model, but allows administrators to manage multiple hosts in a cluster form a single one in a script to e.g. do:
- create database on remote site
- share it in a certain way
to e.g. on the fly initialise new worker nodes, when e.g. the load of existing databases in a pool becomes too high, or when e.g. octopus/armada wants to expand its territory to newly discovered machines (new neighbours discovered by merovingian)
## Comment 13858
Date: 2009-08-01 11:35:18 +0200
From: @grobian
this would also allow other applications to easily retrieve data from merovingian, such as (Java/web) apps (meronitor) that could monitor an entire cluster
## Comment 13859
Date: 2009-09-23 18:23:18 +0200
From: @grobian
this is now implemented and will appear in the next feature release
## Comment 13860
Date: 2009-10-08 04:20:40 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13861
Date: 2009-10-08 09:39:14 +0200
From: @grobian
give me a way to flag this feature request as being implemented but not released, or stop bothering me, sf-robot!
## Comment 13862
Date: 2009-10-23 04:20:50 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13863
Date: 2009-11-07 03:20:46 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13864
Date: 2009-11-22 03:20:15 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13865
Date: 2009-11-22 10:16:30 +0100
From: @grobian
just a few more days until we release...
## Comment 13866
Date: 2009-11-30 10:26:32 +0100
From: @grobian
feature is available starting from the Nov2009 release
## Comment 13867
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2830754 at http://sourceforge.net/support/tracker.php?aid=2830754
| Merovingian: allow control from remote monetdb | https://api.github.com/repos/MonetDB/MonetDB/issues/2502/comments | 0 | 2020-11-30T11:00:13Z | 2024-06-28T07:23:07Z | https://github.com/MonetDB/MonetDB/issues/2502 | 753,376,882 | 2,502 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-07-31 12:49:31 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2009-11-30 11:06:36 +0100
## Comment 13847
Date: 2009-07-31 12:49:31 +0200
From: @grobian
it is very handy to be able to e.g. rename the monetdb user to something else as first line of defence.
## Comment 13848
Date: 2009-09-25 12:35:14 +0200
From: @grobian
ALTER USER name RENAME TO newname
## Comment 13849
Date: 2009-09-25 14:54:40 +0200
From: @grobian
implemented for the next feature release
## Comment 13850
Date: 2009-10-10 04:22:10 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13851
Date: 2009-10-25 03:20:14 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13852
Date: 2009-11-09 03:20:52 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13853
Date: 2009-11-24 03:20:20 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13854
Date: 2009-11-24 08:43:22 +0100
From: @grobian
just a few more days...
## Comment 13855
Date: 2009-11-30 11:06:36 +0100
From: @grobian
This feature is available starting from Nov2009
## Comment 13856
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2830257 at http://sourceforge.net/support/tracker.php?aid=2830257
| SQL allow changing username | https://api.github.com/repos/MonetDB/MonetDB/issues/2501/comments | 0 | 2020-11-30T11:00:10Z | 2024-06-28T07:23:06Z | https://github.com/MonetDB/MonetDB/issues/2501 | 753,376,852 | 2,501 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-07-03 02:48:10 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2016-04-11 11:44:44 +0200
## Comment 13842
Date: 2009-07-03 02:48:10 +0200
From: @skinkie
I received a take down notice of the KVK, until that is resolved I have to keep this academic...
With a new data provider aka real partner without legal battles; new bugs arise:
CREATE TABLE "sys"."faillissementen" (
"kvk" bigint,
"bedrijfsnaam" varchar(255),
"plaats" varchar(32),
"rechtbank" varchar(64),
"rolnummer" varchar(16),
"status" varchar(16),
"datum" date
);
CREATE TABLE "sys"."kvk" (
"kvk" bigint,
"bedrijfsnaam" varchar(255),
"adres" varchar(64),
"postcode" varchar(6),
"plaats" varchar(32),
"type" varchar(16)
);
select * from faillissementen, kvk where faillissementen.kvk = kvk.kvk;
0 tuples
Timer 25.549 msec 0 rows
sql>select * from faillissementen, kvk where (faillissementen.kvk*10000) = kvk.kvk;
22 tuples
Timer 44.483 msec 22 rows
So far fine; now... for some reason nobody really cares to normalise their numbers so I want to join the two.
select * from faillissementen, kvk where (faillissementen.kvk*10000) = kvk.kvk or faillissementen.kvk = kvk.kvk;
Query gets totally out of bounds.
Now you may ask; so what if you do it like first year students?
select * from faillissementen, kvk where (faillissementen.kvk*10000) = kvk.kvk union select * from faillissementen, kvk where faillissementen.kvk = kvk.kvk;
20 tuples
Timer 57.804 msec 20 rows
Now you see that there are 20 results instead of 22. Basically what misses are the multiple results of the kvk side matching the number; The most trivial example:
sql>select kvk, adres from kvk where kvk = 10926910000;
+-------------+-------------------------+
| kvk | adres |
+=============+=========================+
| 10926910000 | Anthonie Fokkerstraat 5 |
| 10926910000 | Anthonie Fokkerstraat 5 |
+-------------+-------------------------+
sql>select kvk.kvk, kvk.adres from faillissementen, kvk where (faillissementen.kvk*10000) = kvk.kvk union select kvk.kvk, kvk.adres from faillissementen, kvk where faillissementen.kvk = kvk.kvk;
+--------------+---------------------------+
| kvk | adres |
+==============+===========================+
| 10926910000 | Anthonie Fokkerstraat 5 |
| 60680140000 | Spanjaardslaan 25 -1 |
sql>select kvk.kvk, kvk.adres from faillissementen, kvk where (faillissementen.kvk*10000) = kvk.kvk;
+--------------+---------------------------+
| kvk | adres |
+==============+===========================+
| 10926910000 | Anthonie Fokkerstraat 5 |
| 10926910000 | Anthonie Fokkerstraat 5 |
## Comment 13843
Date: 2009-07-03 13:40:29 +0200
From: @skinkie
As Niels suggested, union all solves the last part; didn't know union had a build in distinct.
## Comment 13844
Date: 2009-08-06 15:45:09 +0200
From: @mlkersten
stacked for possible future improvements of the query optimizer
## Comment 13845
Date: 2009-08-14 13:58:23 +0200
From: @drstmane
tagged subject
## Comment 13846
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2816059 at http://sourceforge.net/support/tracker.php?aid=2816059
| SQL: Performance slowdown on or | https://api.github.com/repos/MonetDB/MonetDB/issues/2500/comments | 0 | 2020-11-30T10:47:16Z | 2024-06-28T07:19:38Z | https://github.com/MonetDB/MonetDB/issues/2500 | 753,368,114 | 2,500 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-06-28 08:00:04 +0200
From: @skinkie
To: clients devs <<bugs-clients>>
Version: -- development
Last updated: 2009-12-01 10:00:17 +0100
## Comment 13833
Date: 2009-06-28 20:00:04 +0200
From: @skinkie
The error value for an access denied (for example when a user is not granted to access a table) is set. But the error message is not. My suggestion is to set the in mapilib message, my strlen crashed, so I expect it was returning a null value. I have fixed my code by adding a null check.
## Comment 13834
Date: 2009-06-28 21:09:42 +0200
From: @grobian
can you give us (me) the slightest clue what you're doing, what happens and what you expect?
## Comment 13835
Date: 2009-06-28 21:22:58 +0200
From: @skinkie
> can you give us (me) the slightest clue what you're doing
I'm trying to do a query on a table I have not been granted access to. And trying to render the error message notifying my failure to do so.
> what happens
The error message is not set, since strlen crashed on the result value I expect it returned null.
> and what you expect?
My suggestion is to set the in mapilib message.
If the words cannot give you a clue I have the compiled C code here...
/* Render error reply
*/
cherokee_dwriter_dict_open (&hdl->writer);
add_cstr_int (&hdl->writer, "SUCCESS", false);
add_cstr_int (&hdl->writer, "MONETDB_ERRNO", mapi_error(dbh));
tmp = mapi_result_error(resHdl);
if (tmp)
add_cstr_str (&hdl->writer, "MONETDB_ERROR", tmp, strlen(tmp));
## Comment 13836
Date: 2009-06-28 21:29:23 +0200
From: @grobian
/* Render error reply
*/
cherokee_dwriter_dict_open (&hdl->writer);
add_cstr_int (&hdl->writer, "SUCCESS", false);
add_cstr_int (&hdl->writer, "MONETDB_ERRNO",
mapi_error(dbh));
tmp = mapi_result_error(resHdl);
if (tmp)
add_cstr_str (&hdl->writer,
"MONETDB_ERROR", tmp, strlen(tmp));
Can you shed some light on how you would describe the difference between "error value" and "error message"? In particular what the "error value" for an "access denied" is.
## Comment 13837
Date: 2009-06-28 21:38:55 +0200
From: @skinkie
> Can you shed some light on how you would describe the difference between "error value" and "error message"?
Ofcourse; as the documentation on <http://monetdb.cwi.nl/projects/monetdb/SQL/Documentation/Mapi-Function-Reference.html> explains the user:
MapiMsg mapi_error(Mapi mid)
Return the last error code or 0 if there is no error.
> In particular what the "error value" for an "access denied" is.
While:
char *mapi_error_str(Mapi mid)
Return a pointer to the last error message.
or
char *mapi_result_error(MapiHdl hdl)
Return a pointer to the last error message from the server.
So what a user of the mapilib expects here is that MapiMsg is a value upon error or 0 if there is no error. Now if I should use mapi_error_str or mapi_result_error, or as suggested <http://monetdb.cwi.nl/XQuery/Documentation/Library-Synopsis.html> (notice the typo) map_explain_result() is open to me.
## Comment 13838
Date: 2009-06-28 21:42:56 +0200
From: @grobian
I agree, it's almost impossible to pick the right function for the error at hand.
## Comment 13839
Date: 2009-07-13 10:41:45 +0200
From: @grobian
this is more of a feature request that gijs is going to tackle
## Comment 13840
Date: 2009-12-01 10:00:17 +0100
From: @grobian
we'll stick with plain old Mapi, so this feature request boils down to a RTFC
## Comment 13841
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2813606 at http://sourceforge.net/support/tracker.php?aid=2813606
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-clients@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| MAPI: Error message for access denied | https://api.github.com/repos/MonetDB/MonetDB/issues/2499/comments | 0 | 2020-11-30T10:47:13Z | 2024-06-28T07:23:06Z | https://github.com/MonetDB/MonetDB/issues/2499 | 753,368,079 | 2,499 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-06-04 10:09:09 +0200
From: @grobian
To: Merovingian devs <<bugs-merovingian>>
Version: -- development
Last updated: 2016-04-11 11:44:08 +0200
## Comment 13831
Date: 2009-06-04 22:09:09 +0200
From: @grobian
Merovingian currently only "discovers" others that run on the same network and port. In some situations this makes it impossible for two Merovingian to "see" each other. Instead, a fixed peer setting is necessary, where Merovingian can be configured to "see" another Merovingian at some IP and port.
## Comment 13832
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2801350 at http://sourceforge.net/support/tracker.php?aid=2801350
| Merovingian: allow for fixed peers | https://api.github.com/repos/MonetDB/MonetDB/issues/2498/comments | 0 | 2020-11-30T10:47:10Z | 2024-06-28T07:19:38Z | https://github.com/MonetDB/MonetDB/issues/2498 | 753,368,042 | 2,498 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-06-04 11:09:16 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2009-08-25 02:36:08 +0200
## Comment 13816
Date: 2009-06-04 11:09:16 +0200
From: @grobian
Currently the discovery service is hardwired to the merovingian (mapi) port. The discovery service port should be a separate option, such that it becomes possible to run multiple merovingians on the same machine still communicating to each other.
This would also alloow to disable the discovery service, by setting the port to 0.
## Comment 13817
Date: 2009-06-04 21:52:27 +0200
From: @grobian
the communicating to each other on the same machine part is bogus
## Comment 13818
Date: 2009-06-04 22:03:06 +0200
From: @grobian
implemented for the next release
## Comment 13819
Date: 2009-06-19 04:20:32 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13820
Date: 2009-06-19 08:10:21 +0200
From: @grobian
bad robot
## Comment 13821
Date: 2009-07-04 04:20:24 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13822
Date: 2009-07-04 12:30:45 +0200
From: @grobian
robot overrule
## Comment 13823
Date: 2009-07-19 04:20:13 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13824
Date: 2009-07-19 20:27:29 +0200
From: @grobian
robot re-overrule
## Comment 13825
Date: 2009-08-03 04:20:18 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13826
Date: 2009-08-03 09:13:55 +0200
From: @grobian
robot re-re-overrule
## Comment 13827
Date: 2009-08-18 04:20:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13828
Date: 2009-08-18 09:36:58 +0200
From: @grobian
boe
## Comment 13829
Date: 2009-08-25 14:36:07 +0200
From: @grobian
feature was delivered in Aug2008 release
## Comment 13830
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2800964 at http://sourceforge.net/support/tracker.php?aid=2800964
| Merovingian: have configurable discovery service port | https://api.github.com/repos/MonetDB/MonetDB/issues/2497/comments | 0 | 2020-11-30T10:47:06Z | 2024-06-28T07:23:06Z | https://github.com/MonetDB/MonetDB/issues/2497 | 753,367,999 | 2,497 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-05-16 02:54:07 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
CC: dennis, martin.van.dinther
Last updated: 2019-04-30 12:36:02 +0200
## Comment 13812
Date: 2009-05-16 02:54:07 +0200
From: @skinkie
For some reason this doesn't work:
sql>select GREATEST(5,3);
!SELECT: no such binary operator 'greatest(tinyint,tinyint)'
I didn't see GREATEST or LEAST in SQL92, but for some reason it seems to be defined in MonetDB anyway.
## Comment 13813
Date: 2009-07-09 20:54:00 +0200
From: @grobian
MAX/MIN don't work by chance?
where do GREATEST/LEAST come from?
## Comment 13814
Date: 2009-07-09 21:33:11 +0200
From: @skinkie
max/min operate on aggregated results. Greatest/Least results into a value choosen from two (or more) columns.
## Comment 13815
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2792535 at http://sourceforge.net/support/tracker.php?aid=2792535
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 19907
Date: 2014-07-16 11:12:40 +0200
From: Dennis Pallett <<dennis>>
I'm seconding this request for a GREATEST and a LEAST function. Seems like it should be fairly easy to implement (?).
## Comment 21950
Date: 2016-03-24 15:32:40 +0100
From: Martin van Dinther <<martin.van.dinther>>
MonetDB supports the sys.sql_min() and sys.sql_max() scalar functions to determine the greatest or least of two values (or column expressions), e.g:
select sql_min(12345, 5789); // the least of 2 smallint values
select sql_min(12345.12, 5789.12); // the least of 2 decimal values
select sql_min('ab', 'cb'); // the least of 2 char values
select sql_max(12345, 5789); // the greatest of 2 smallint values
select sql_max(12345.12, 5789.12); // the greatest of 2 decimal values
select sql_max('ab', 'cb'); // the greatest of 2 char values
Please note that sql_min() and sql_max():
- only support 2 arguments, not more (GREATEST and LEAST may have more than 2 arguments).
Use sql_min(arg1, sql_min(arg2, arg3)) to simulate 3 args equivalent.
- both arguments must be of the exact same type!
Use a cast() to make the arguments the same (largest) type, e.g.:
select sql_min(cast(12 as smallint), 5789);
- when the first argument is NULL (or expression evaluates to NULL) it returns an error!
- when the second argument is NULL (or expression evaluates to NULL) it always returns NULL!
## Comment 26853
Date: 2019-01-28 10:02:02 +0100
From: MonetDB Mercurial Repository <<hg>>
Changeset [cac83dfdf816](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=cac83dfdf816) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=cac83dfdf816](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=cac83dfdf816)
Changeset description:
Implemented two-argument least/greatest functions.
This fixes bug #2496.
| SQL: implement greatest / least | https://api.github.com/repos/MonetDB/MonetDB/issues/2496/comments | 0 | 2020-11-30T10:47:03Z | 2024-06-27T11:46:41Z | https://github.com/MonetDB/MonetDB/issues/2496 | 753,367,961 | 2,496 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-05-06 01:54:55 +0200
From: Wouter Alink <<alink>>
To: clients devs <<bugs-clients>>
Version: -- development
Last updated: 2009-05-20 05:12:16 +0200
## Comment 13808
Date: 2009-05-06 13:54:55 +0200
From: Wouter Alink <<alink>>
A little experiment (on the head) reveals a bug in mclient (it does
not handle correctly the optional BOM at the beginning of the input):
$ cat selectWithBOM.py
print "\xEF\xBB\xBFSELECT 1;"
$ python selectWithBOM.py > queryWithBOM.sql
$ xxd queryWithBOM.sql
0000000: efbb bf53 454c 4543 5420 313b 0a ...SELECT 1;.
$ cat queryWithBOM.sql
SELECT 1;
$ echo "SELECT 1;" | mclient -lsql
% . table_name
% single_value name
% tinyint type
% 1 length
[ 1 ]
$ cat queryWithBOM.sql | mclient -lsql
(Hangs)
## Comment 13809
Date: 2009-05-07 14:28:31 +0200
From: @sjoerdmullender
As far as I know, Unicode does not require the BOM to be supported in the UTF-8 encoding. If you can point me to the place in any appropriate document that support is required, this is a bug, otherwise a feature request.
## Comment 13810
Date: 2009-05-20 17:12:16 +0200
From: @sjoerdmullender
Implemented.
That is to say, only if the BOM occurs at the beginning of a file.
Test added: sql/src/test/BugTracker-2009/Tests/BOM.SF-2787813.sql
## Comment 13811
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2787813 at http://sourceforge.net/support/tracker.php?aid=2787813
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-clients@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| mclient: BOM on stdin not accepted | https://api.github.com/repos/MonetDB/MonetDB/issues/2495/comments | 0 | 2020-11-30T10:47:00Z | 2024-06-28T07:23:05Z | https://github.com/MonetDB/MonetDB/issues/2495 | 753,367,910 | 2,495 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-04-28 02:36:07 +0200
From: @grobian
To: MonetDB5 devs <<bugs-monetdb5>>
Version: -- development
Last updated: 2009-06-04 10:54:27 +0200
## Comment 13805
Date: 2009-04-28 14:36:07 +0200
From: @grobian
The current authorisation system needs a change to be able to dump and restore users from SQL.
- we need a SHA-backed password storage (windhoos SHA/openssl support required) iso current XOR engine
- we need SQL create user extension to use the SHA hash iso plain password
- we need mdump to be able to read the password BATs (currently protected)
## Comment 13806
Date: 2009-06-04 10:54:26 +0200
From: @grobian
implemented in May2009 release
## Comment 13807
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2782807 at http://sourceforge.net/support/tracker.php?aid=2782807
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-monetdb5@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| M5: change user management for dump/restore | https://api.github.com/repos/MonetDB/MonetDB/issues/2494/comments | 0 | 2020-11-30T10:46:57Z | 2024-06-28T07:23:05Z | https://github.com/MonetDB/MonetDB/issues/2494 | 753,367,883 | 2,494 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-04-28 01:29:35 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2009-05-28 04:20:26 +0200
## Comment 13801
Date: 2009-04-28 13:29:35 +0200
From: @grobian
Currently Merovingian inserts timestamps every once in a while in the logfile to indicate approximately when something happens. However, for merovingian self originated messages, the time is well known. Hence, it is better to just prefix all log messages with a timestamp, like e.g. syslog does.
## Comment 13802
Date: 2009-05-13 15:59:37 +0200
From: @grobian
implemented in Current
## Comment 13803
Date: 2009-05-28 04:20:26 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13804
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2782776 at http://sourceforge.net/support/tracker.php?aid=2782776
| Merovingian: prefix all log messages with timestamp | https://api.github.com/repos/MonetDB/MonetDB/issues/2493/comments | 0 | 2020-11-30T10:46:53Z | 2024-06-28T07:23:05Z | https://github.com/MonetDB/MonetDB/issues/2493 | 753,367,845 | 2,493 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-04-24 03:04:22 +0200
From: @skinkie
To: @njnes
Version: -- development
Last updated: 2009-12-01 10:03:19 +0100
## Comment 13797
Date: 2009-04-24 03:04:22 +0200
From: @skinkie
Upon a search clause such as lower(...) or upper(...) it would probably be beneficial to create an extra hash index to represent the new compared strings. Next to this, I don't know if Monet already uses tricks such to directly quit if lower(something) = 'HELLO', but this can be included.
## Comment 13798
Date: 2009-04-24 03:26:46 +0200
From: @skinkie
I wonder now if it wouldn't be more interesting to make it by default caseless, since this will allow all operations {upper, lower, none}, and would only act as prefilter.
## Comment 13799
Date: 2009-12-01 10:03:19 +0100
From: @grobian
Please explain what "active string modification" is and in which apparent cases this would cause a problem.
## Comment 13800
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2780037 at http://sourceforge.net/support/tracker.php?aid=2780037
| Create new hash index for active string modification | https://api.github.com/repos/MonetDB/MonetDB/issues/2492/comments | 0 | 2020-11-30T10:46:50Z | 2024-06-28T07:23:04Z | https://github.com/MonetDB/MonetDB/issues/2492 | 753,367,800 | 2,492 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-04-22 01:56:58 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2016-04-11 11:45:11 +0200
## Comment 13794
Date: 2009-04-22 13:56:58 +0200
From: @skinkie
update(sys.node_tags,
| project (
| | left outer join (
| | | table(sys.node_tags) [ node_tags.node, node_tags.k, node_tags.v, node_tags.vlen, node_tags.%TID% NOT NULL ],
| | | group by (
| | | | table(sys.node_tags) [ node_tags.node as y.node, node_tags.k as y.k, node_tags.v as y.v, node_tags.vlen as y.vlen, node_tags.%TID% NOT NULL as y.%TID% ]
| | | ) [ y.node as L1.node, y.k as L1.k, y.v as L1.v ] [ y.node as L1.node, y.k as L1.k, y.v as L1.v, zero_or_one(length(y.v)) NOT NULL as L2 ]
| | ) [ node_tags.node = L1.node, node_tags.k = L1.k, node_tags.v = L1.v ]
| ) [ node_tags.%TID% NOT NULL, L2 NOT NULL as L3.L3 ]
) [ node_tags.vlen NOT NULL ]
This could use the unique property instead of group by.
## Comment 13795
Date: 2009-12-01 10:05:35 +0100
From: @grobian
does it make a notable difference?
## Comment 13796
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2778533 at http://sourceforge.net/support/tracker.php?aid=2778533
| SQL: use UNIQUE instead of GROUP BY | https://api.github.com/repos/MonetDB/MonetDB/issues/2491/comments | 0 | 2020-11-30T10:46:48Z | 2024-06-28T07:19:38Z | https://github.com/MonetDB/MonetDB/issues/2491 | 753,367,766 | 2,491 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-04-20 03:09:19 +0200
From: @skinkie
To: @njnes
Version: -- development
Last updated: 2009-04-20 03:39:19 +0200
## Comment 13791
Date: 2009-04-20 15:09:19 +0200
From: @skinkie
UPDATE p
SET
p.col1 = p.col1 + q.otherCol
FROM table1 AS p join table2 AS q
ON p.fk_table2_pk = q.pk
WHERE
q.filterColumn = 'something'
The above should allow bulk processing ot updates. I have tranlated it to the following query;
create table test3(id integer, hoi integer);
create table test4(id integer, hoi integer);
insert into test3 values (1, 2);
insert into test4 values (1, 3);
This works: update test3 set hoi = (select hoi from test4 where test3.id = test4.id);
I expect this should have better performance, if it would be supported:
update test3 set test3.hoi = test4.hoi from test3 join test4 on test3.id = test4.id;
## Comment 13792
Date: 2009-04-20 15:39:19 +0200
From: @skinkie
Shame on me... it is already by design;
update(sys.test3a,
| project (
| | left outer join (
| | | table(sys.test3a) [ test3a.id, test3a.hoi, test3a.%TID% NOT NULL ],
| | | group by (
| | | | table(sys.test4a) [ test4a.id, test4a.hoi, test4a.%TID% NOT NULL ]
| | | ) [ test4a.id as L1.id ] [ test4a.id as L1.id, zero_or_one(test4a.hoi) NOT NULL as L2 ]
| | ) [ test3a.id = L1.id ]
| ) [ test3a.%TID% NOT NULL, L2 NOT NULL as L3.L3 ]
) [ test3a.hoi NOT NULL ]
## Comment 13793
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2776184 at http://sourceforge.net/support/tracker.php?aid=2776184
| Update using JOIN | https://api.github.com/repos/MonetDB/MonetDB/issues/2490/comments | 0 | 2020-11-30T10:46:44Z | 2024-06-28T07:23:04Z | https://github.com/MonetDB/MonetDB/issues/2490 | 753,367,730 | 2,490 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-03-23 10:46:15 +0100
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2009-05-30 04:20:24 +0200
## Comment 13787
Date: 2009-03-23 10:46:15 +0100
From: @grobian
For utilities like logrotate, it is necessary to be able to reopen open logfiles such that old logfiles can be moved out of the way, and a new logfile written.
## Comment 13788
Date: 2009-05-15 18:35:24 +0200
From: @grobian
scheduled for the May feature release.
## Comment 13789
Date: 2009-05-30 04:20:24 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13790
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2705670 at http://sourceforge.net/support/tracker.php?aid=2705670
| Merovingian: reopen logfiles on SIGHUP | https://api.github.com/repos/MonetDB/MonetDB/issues/2489/comments | 0 | 2020-11-30T10:46:41Z | 2024-06-28T07:23:04Z | https://github.com/MonetDB/MonetDB/issues/2489 | 753,367,686 | 2,489 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-03-22 08:42:21 +0100
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2016-04-11 11:46:02 +0200
## Comment 13784
Date: 2009-03-22 20:42:21 +0100
From: @skinkie
Using the OSM dataset I noticed there is a significant difference in the following queries:
select * from way_tags where way = 9168360 and k = 'created_by';
Timer 10583.483 msec 2 rows
vs
select * from way_tags where k = 'created_by' and way = 9168360;
Timer 331.086 msec 2 rows
(this is not about hot/cold performance, the next try of the first query results in 10286.816 msec)
Niels noted that this could be solved automatically using a dynamic selectivity optimiser. I first noticed this issue using delete from.
## Comment 13785
Date: 2009-03-22 20:51:23 +0100
From: @mlkersten
such an optimizer is a feature request, rather then a bug.
## Comment 13786
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2704025 at http://sourceforge.net/support/tracker.php?aid=2704025
| SQL: dynamic selectivity optimiser required | https://api.github.com/repos/MonetDB/MonetDB/issues/2488/comments | 0 | 2020-11-30T10:46:38Z | 2024-06-28T07:19:37Z | https://github.com/MonetDB/MonetDB/issues/2488 | 753,367,656 | 2,488 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-03-15 12:05:54 +0100
From: @grobian
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2009-03-23 10:44:50 +0100
## Comment 13776
Date: 2009-03-15 12:05:54 +0100
From: @grobian
While trying to implement --enable-testing and --check in monetdb-install.sh, I came to the unpleasant discovery that the testing repository is not made available anywhere, but in CVS.
Without testing, it is hard to run Mtest.py on machines that aren't setup in a developer fashion, but instead simply used for a one-time-shot using monetdb-install.sh/SuperSourceTarball.
## Comment 13777
Date: 2009-03-15 12:25:04 +0100
From: @drstmane
MonetDB-testing appears to be available both as tarball and as binary (Linux) packages (see below).
It does not seems to be part of MonetDB-Feb2009-SuperBall.tar.bz2, though:
========
$ wget http://monetdb.cwi.nl/downloads/sources/Feb2009/MonetDB-Feb2009-SuperBall.tar.bz2 && tar -tvpf MonetDB-Feb2009-SuperBall.tar.bz2 | grep -c MonetDB-testing
--2009-03-15 12:24:19-- http://monetdb.cwi.nl/downloads/sources/Feb2009/MonetDB-Feb2009-SuperBall.tar.bz2
Resolving monetdb.cwi.nl... 192.16.196.21
Connecting to monetdb.cwi.nl|192.16.196.21|:80... connected.
HTTP request sent, awaiting response... 200 OK
Length: 54892469 (52M) [application/x-bzip2]
Saving to: `MonetDB-Feb2009-SuperBall.tar.bz2.2'
100%[============================================================================================================================================================================================================>] 54,892,469 33.2M/s in 1.6s
2009-03-15 12:24:21 (33.2 MB/s) - `MonetDB-Feb2009-SuperBall.tar.bz2.2' saved [54892469/54892469]
0
========
http://monetdb.cwi.nl/downloads/sources/Feb2009/MonetDB-testing-1.28.0.tar.gz
http://monetdb.cwi.nl/downloads/Ubuntu/pool/monetdb/m/monetdb-testing/monetdb-testing_1.28.0-1.tar.gz
http://monetdb.cwi.nl/downloads/Ubuntu/pool/monetdb/m/monetdb-testing/monetdb-testing_1.28.0-1_amd64.deb
http://monetdb.cwi.nl/downloads/Ubuntu/pool/monetdb/m/monetdb-testing/monetdb-testing_1.28.0-1.dsc
http://monetdb.cwi.nl/downloads/Ubuntu/pool/monetdb/m/monetdb-client/monetdb-client-testing_1.28.0-1_amd64.deb
http://monetdb.cwi.nl/downloads/Debian/pool/monetdb/m/monetdb-testing/monetdb-testing_1.28.0-1.tar.gz
http://monetdb.cwi.nl/downloads/Debian/pool/monetdb/m/monetdb-testing/monetdb-testing_1.28.0-1_i386.deb
http://monetdb.cwi.nl/downloads/Debian/pool/monetdb/m/monetdb-testing/monetdb-testing_1.28.0-1_amd64.deb
http://monetdb.cwi.nl/downloads/Debian/pool/monetdb/m/monetdb-testing/monetdb-testing_1.28.0-1.dsc
http://monetdb.cwi.nl/downloads/Debian/pool/monetdb/m/monetdb-client/monetdb-client-testing_1.28.0-1_i386.deb
http://monetdb.cwi.nl/downloads/Debian/pool/monetdb/m/monetdb-client/monetdb-client-testing_1.28.0-1_amd64.deb
http://monetdb.cwi.nl/downloads/Fedora/debug/10/i386/MonetDB-testing-debuginfo-1.28.0-1.fc10.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/debug/10/x86_64/MonetDB-testing-debuginfo-1.28.0-1.fc10.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/debug/9/i386/MonetDB-testing-debuginfo-1.28.0-1.fc9.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/debug/9/x86_64/MonetDB-testing-debuginfo-1.28.0-1.fc9.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/10/i386/MonetDB-testing-1.26.2-1.fc10.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/10/i386/MonetDB-testing-1.28.0-1.fc10.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/10/i386/MonetDB-testing-1.26.0-1.fc10.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/10/i386/MonetDB-testing-1.26.4-1.fc10.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/10/x86_64/MonetDB-testing-1.26.0-1.fc10.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/10/x86_64/MonetDB-testing-1.26.2-1.fc10.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/10/x86_64/MonetDB-testing-1.26.4-1.fc10.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/10/x86_64/MonetDB-testing-1.28.0-1.fc10.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/source/MonetDB-testing-1.28.0-1.src.rpm
http://monetdb.cwi.nl/downloads/Fedora/9/i386/MonetDB-testing-1.26.4-1.fc9.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/9/i386/MonetDB-testing-1.26.0-1.fc9.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/9/i386/MonetDB-testing-1.26.2-1.fc9.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/9/i386/MonetDB-testing-1.28.0-1.fc9.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/9/x86_64/MonetDB-testing-1.26.0-1.fc9.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/9/x86_64/MonetDB-testing-1.28.0-1.fc9.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/9/x86_64/MonetDB-testing-1.26.4-1.fc9.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/9/x86_64/MonetDB-testing-1.26.2-1.fc9.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/8/i386/MonetDB-testing-1.26.0-1.fc8.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/8/i386/MonetDB-testing-1.26.4-1.fc8.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/8/i386/MonetDB-testing-1.26.2-1.fc8.oid32.i386.rpm
http://monetdb.cwi.nl/downloads/Fedora/8/x86_64/MonetDB-testing-1.26.2-1.fc8.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/8/x86_64/MonetDB-testing-1.26.0-1.fc8.oid64.x86_64.rpm
http://monetdb.cwi.nl/downloads/Fedora/8/x86_64/MonetDB-testing-1.26.4-1.fc8.oid64.x86_64.rpm
## Comment 13778
Date: 2009-03-15 12:41:25 +0100
From: @grobian
My bad, it's not in http://monetdb.cwi.nl/downloads/sources/Nov2008-SP2/ (which is what monetdb-install.sh uses AFAICT)
## Comment 13779
Date: 2009-03-15 12:44:09 +0100
From: @drstmane
With Nov2008, testing is/was still in MonetDB.
## Comment 13780
Date: 2009-03-15 16:01:15 +0100
From: @grobian
(pegasus:tars/latest/current) fabian% ls /net/homer.ins.cwi.nl/export/scratch2//monet/Testing/Current/.DailyBuilds./1____all/SourceTarBalls/*testing*
ls: No match.
is it supposed to be in the nightlies?
## Comment 13781
Date: 2009-03-15 16:08:26 +0100
From: @drstmane
MonetDB-testing will be available in the nightly builds as of tomorrow (Mon Mar 16 2009).
## Comment 13782
Date: 2009-03-23 10:44:50 +0100
From: @grobian
available now. Thanks!
## Comment 13783
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2687560 at http://sourceforge.net/support/tracker.php?aid=2687560
| make MonetDB testing available as tarballs | https://api.github.com/repos/MonetDB/MonetDB/issues/2487/comments | 0 | 2020-11-30T10:46:35Z | 2024-06-28T07:23:04Z | https://github.com/MonetDB/MonetDB/issues/2487 | 753,367,617 | 2,487 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-03-12 10:18:49 +0100
From: @skinkie
To: @njnes
Version: -- development
Last updated: 2009-12-01 10:09:13 +0100
## Comment 13773
Date: 2009-03-12 10:18:49 +0100
From: @skinkie
as $subject
## Comment 13774
Date: 2009-12-01 10:09:13 +0100
From: @grobian
sql>select XOR(5,3);
+------------------+
| xor_single_value |
+==================+
| false |
+------------------+
1 tuple
sql>
WFM
## Comment 13775
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2684760 at http://sourceforge.net/support/tracker.php?aid=2684760
| XOR | https://api.github.com/repos/MonetDB/MonetDB/issues/2486/comments | 0 | 2020-11-30T10:46:32Z | 2024-06-28T07:23:03Z | https://github.com/MonetDB/MonetDB/issues/2486 | 753,367,585 | 2,486 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-03-11 02:31:28 +0100
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
CC: duc
Last updated: 2016-04-11 11:46:12 +0200
## Comment 13767
Date: 2009-03-11 14:31:28 +0100
From: @skinkie
Upon inserting it would be nice to have a default value based on other columns in that same row.
## Comment 13768
Date: 2009-12-01 10:10:38 +0100
From: @grobian
Why can't you write that using the current DEFAULT implementation?
NEEDINFO
## Comment 13769
Date: 2009-12-01 10:35:03 +0100
From: @grobian
reopen for skinkie, as it seems he doesn't want to clarify otherwise
## Comment 13770
Date: 2009-12-01 10:44:11 +0100
From: @skinkie
;)
Other database support 'dynamic' DEFAULTs. Hence you are allowed to place an expression into the default. This is currently not possible in MonetDB where it is static.
I don't see the big problem (performance) problem in this as long the default is added on the end of a bulk transaction. Instead of foreach record. S in general it would be an 'on-insert-update-trigger', opposed to the other request 'on-update-update-trigger'.
## Comment 13771
Date: 2009-12-01 10:47:45 +0100
From: @grobian
as far as I know "DEFAULT NEXT VALUE FOR seq_X" is pretty dynamic
## Comment 13772
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2682714 at http://sourceforge.net/support/tracker.php?aid=2682714
| SQL: default with expression | https://api.github.com/repos/MonetDB/MonetDB/issues/2485/comments | 0 | 2020-11-30T10:46:29Z | 2024-06-28T07:19:37Z | https://github.com/MonetDB/MonetDB/issues/2485 | 753,367,551 | 2,485 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-03-11 02:29:55 +0100
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
CC: duc
Last updated: 2016-04-11 11:46:36 +0200
## Comment 13764
Date: 2009-03-11 14:29:55 +0100
From: @skinkie
I would like to update a specific column after an update to its row. The problem that will occur is a recursion. The feature request is the abilitity to put triggers upon column updates instead of the row update.
## Comment 13765
Date: 2009-12-01 10:12:35 +0100
From: @grobian
decomplicated the subject
## Comment 13766
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2682710 at http://sourceforge.net/support/tracker.php?aid=2682710
| SQL: add trigger support for columns | https://api.github.com/repos/MonetDB/MonetDB/issues/2484/comments | 0 | 2020-11-30T10:46:26Z | 2024-06-28T07:19:37Z | https://github.com/MonetDB/MonetDB/issues/2484 | 753,367,502 | 2,484 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-03-07 07:21:21 +0100
From: @skinkie
To: @njnes
Version: -- development
Last updated: 2009-03-12 10:19:18 +0100
## Comment 13761
Date: 2009-03-07 19:21:21 +0100
From: @skinkie
This implements 'in SQL' access to the last_id.
Index: src/backends/monet5/sql.mx
===================================================================
RCS file: /cvsroot/monetdb/sql/src/backends/monet5/sql.mx,v
retrieving revision 1.319.2.1
diff -u -r1.319.2.1 sql.mx
--- src/backends/monet5/sql.mx 15 Feb 2009 10:44:34 -0000 1.319.2.1
+++ src/backends/monet5/sql.mx 7 Mar 2009 18:19:37 -0000
@@ -1330,6 +1330,7 @@
if (seq && seq_next_value(seq, res)) {
m->last_id = *res;
+ stack_set_number(m, "last_id", m->last_id); /* MSSQL defined extension @@IDENTITY */
return MAL_SUCCEED;
}
}
Index: src/backends/monet5/sql_scenario.mx
===================================================================
RCS file: /cvsroot/monetdb/sql/src/backends/monet5/sql_scenario.mx,v
retrieving revision 1.334.2.2
diff -u -r1.334.2.2 sql_scenario.mx
--- src/backends/monet5/sql_scenario.mx 27 Feb 2009 19:26:08 -0000 1.334.2.2
+++ src/backends/monet5/sql_scenario.mx 7 Mar 2009 18:19:38 -0000
@@ -273,7 +273,7 @@
sql_subtype ctype;
char *typename;
stmt *r = NULL;
- lng sec = 0;
+ lng sec = 0, last_id = -1;
bit T = TRUE;
bit F = FALSE;
ValRecord src;
@@ -281,6 +281,7 @@
typename = "int";
sql_find_subtype(&ctype, typename, 0, 0);
SQLglobal("debug", &sql->debug);
+ SQLglobal("last_id", &last_id);
typename = "varchar";
sql_find_subtype(&ctype, typename, 1024, 0);
## Comment 13762
Date: 2009-03-12 10:19:17 +0100
From: @skinkie
It is in.
## Comment 13763
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2671706 at http://sourceforge.net/support/tracker.php?aid=2671706
| @@IDENTITY like extension for MonetDB | https://api.github.com/repos/MonetDB/MonetDB/issues/2483/comments | 0 | 2020-11-30T10:46:23Z | 2024-06-28T07:23:03Z | https://github.com/MonetDB/MonetDB/issues/2483 | 753,367,455 | 2,483 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-02-24 01:56:20 +0100
From: @skinkie
To: @njnes
Version: -- development
Last updated: 2009-03-12 10:21:54 +0100
## Comment 13757
Date: 2009-02-24 01:56:20 +0100
From: @skinkie
MonetDB currently does not support <http://monetdb.cwi.nl/SQL/Documentation/SQL-Features.html> E121-03 Value expressions in ORDER BY clause. This SQL99 syntax consist of for example:
ORDER BY count(something) ASC
## Comment 13758
Date: 2009-02-24 08:43:44 +0100
From: @skinkie
Oh, I see...the TopCount function is generating this.
Yes, E121-03 (Value expressions in ORDER BY clause) is the relevant feature.
The <order by> scoping rules are subtle because of this feature; you can find them in SQL:2003 Part 2, Section 14.1 <declare cursor>. We think we have them correctly implemented in the Eigenbase validator.
JVS
## Comment 13759
Date: 2009-03-12 10:21:54 +0100
From: @skinkie
It is in.
## Comment 13760
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2632114 at http://sourceforge.net/support/tracker.php?aid=2632114
| E121-03 Value expressions in ORDER BY clause | https://api.github.com/repos/MonetDB/MonetDB/issues/2482/comments | 0 | 2020-11-30T10:46:19Z | 2024-06-28T13:42:12Z | https://github.com/MonetDB/MonetDB/issues/2482 | 753,367,413 | 2,482 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-02-17 04:58:15 +0100
From: @skinkie
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2009-03-12 10:22:24 +0100
## Comment 13754
Date: 2009-02-17 16:58:15 +0100
From: @skinkie
sql>select '1' as "(Key)";
!Invalid identifier '(Key)'
0 tuples
sql>select '1' as "Key";
+-----+
| Key |
+=====+
| 1 |
+-----+
1 tuple
## Comment 13755
Date: 2009-03-12 10:22:24 +0100
From: @skinkie
It is in.
## Comment 13756
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2609477 at http://sourceforge.net/support/tracker.php?aid=2609477
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Increase valid column identifiers | https://api.github.com/repos/MonetDB/MonetDB/issues/2481/comments | 0 | 2020-11-30T10:46:16Z | 2024-06-28T13:42:12Z | https://github.com/MonetDB/MonetDB/issues/2481 | 753,367,378 | 2,481 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-02-17 04:31:55 +0100
From: @skinkie
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2009-03-12 10:20:19 +0100
## Comment 13751
Date: 2009-02-17 16:31:55 +0100
From: @skinkie
SELECT * FROM (VALUES ('USA', 'US'), ('Mexico', 'MX'), ('Canada', 'CA')) AS "t";
Not supported yet.
## Comment 13752
Date: 2009-03-12 10:20:19 +0100
From: @skinkie
It is in.
## Comment 13753
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2609425 at http://sourceforge.net/support/tracker.php?aid=2609425
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| VALUES (...) as t; | https://api.github.com/repos/MonetDB/MonetDB/issues/2480/comments | 0 | 2020-11-30T10:46:13Z | 2024-06-28T13:42:11Z | https://github.com/MonetDB/MonetDB/issues/2480 | 753,367,337 | 2,480 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-02-15 10:26:18 +0100
From: @drstmane
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2011-04-28 16:38:54 +0200
## Comment 13749
Date: 2009-02-15 22:26:18 +0100
From: @drstmane
Like child and descendant steps, also preceding and following steps should allow name-(and other) -test push-downs to avoid unnecessary large intermediate results.
## Comment 13750
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2603533 at http://sourceforge.net/support/tracker.php?aid=2603533
## Comment 15760
Date: 2011-04-28 16:38:54 +0200
From: @sjoerdmullender
Development on MonetDB4 and MonetDB/XQuery has ceased. It is unlikely that this bug will be fixed.
The last release of MonetDB/XQuery and MonetDB4 is Mar2011.
| XQ: name-test push down for preceeding and following step | https://api.github.com/repos/MonetDB/MonetDB/issues/2479/comments | 0 | 2020-11-30T10:46:09Z | 2024-06-28T13:42:10Z | https://github.com/MonetDB/MonetDB/issues/2479 | 753,367,299 | 2,479 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-02-03 08:00:33 +0100
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: duc, @eyalroz, irares, @PedroTadim, @drstmane
Last updated: 2019-04-30 12:36:02 +0200
## Comment 13745
Date: 2009-02-03 20:00:33 +0100
From: @mlkersten
RENAME COLUMN through SQL
## Comment 13746
Date: 2009-04-28 13:33:43 +0200
From: @grobian
ALTER TABLE xxx RENAME [ COLUMN ] yyy TO zzz
## Comment 13747
Date: 2009-12-04 11:33:54 +0100
From: @mlkersten
seems long hanging fruit.
## Comment 13748
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2561693 at http://sourceforge.net/support/tracker.php?aid=2561693
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 15445
Date: 2011-01-31 14:26:48 +0100
From: @grobian
Created attachment 50
partial rename column patch
This patch contains the work done so far, it actually breaks on the gdk logger, where I cannot do the rename thing.
> Attached file: [rename-column.patch](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_2478_rename-column.patch_50) (text/plain, 8163 bytes)
> Description: partial rename column patch
## Comment 17071
Date: 2012-03-15 08:00:29 +0100
From: irares
Hello guys,
I have made yesterday the second serious attempt to move our application from Postgres to Monet. I have met two show-stopper issues which put an end to this:
1. Lack of RENAME TABLE and RENAME COLUMN functionality during the run of our SQL migration scripts. This is serious. For any moderately complex database schema, dropping and recreating tables just to change the name of a column is hard work.
2. Extremely slow INSERT speed through JDBC (200 rows / second). Postgres does 6000 rows / sec.
It is nice that Monet is 20 times faster than Postgres for reads (and 30 times slower for writes), but the SQL is just too limited.
## Comment 17074
Date: 2012-03-15 09:06:13 +0100
From: @drstmane
This rather looks like a feature request that a bug report to me.
## Comment 17075
Date: 2012-03-15 09:09:24 +0100
From: @drstmane
As an alternative to renaming tables or columns, using views and handling name mappings in the application could be considered.
## Comment 17076
Date: 2012-03-15 13:11:39 +0100
From: irares
(In reply to comment 7)
> As an alternative to renaming tables or columns, using views and handling name
> mappings in the application could be considered.
This enhancement request was first filled Feb 2009, 3 years and 1 month ago. Wouldn't you say it is more effective that you add a small feature implemented by most common databases, such as Postgres, MySQL, MSSQL, etc, rather than force your users to use difficult and kludgy workarounds? Renaming columns and tables are basic operations, not rocket science.
## Comment 17077
Date: 2012-03-15 13:14:32 +0100
From: @grobian
Problem is that even though it seems easy to do, in fact it is not, since at a critical level columns cannot be renamed. The partial patch basically went that way and it won't work. There might be more luck in an approach that drops the column, but sneakily keeps the data around, then creates a new column, and sneakily attaches the saved data again.
## Comment 24616
Date: 2016-10-24 23:16:07 +0200
From: @eyalroz
I am just now sorely missing this feature. Scenario: You've loaded a very large amount of data into a DB table, and the original files are gone. Do you really have to recreate the entire table just for a column rename? Surely that can't be right.
By the way...
- What was missing/unacceptable in Fabian's approach? He made it seem kind of easy, essentially just updating the columns table.
- Can this be done "manually", by a humanly-feasible manipulating the BBP data while the DBMS is stopped? That could be a kind of a workaround.
## Comment 26672
Date: 2018-11-07 11:56:29 +0100
From: @PedroTadim
A new SQL statement will be available in the next feature release of MonetDB for renaming columns:
ALTER TABLE [IF EXISTS] [schemaname.]tablename RENAME [COLUMN] oldname TO newname;
We did this by adding a new entry in the GDK logger referencing the name change of a SQL object, thus no moving data occurs.
However if the column as dependencies (i.g. used in a UDF or SQL key), it can't be renamed. This check is enforced because renaming while the column has references can potentially break users SQL scripts.
## Comment 26821
Date: 2019-01-21 15:38:15 +0100
From: @sjoerdmullender
This should be in the next feature release.
| SQL: rename columns | https://api.github.com/repos/MonetDB/MonetDB/issues/2478/comments | 0 | 2020-11-30T10:46:05Z | 2024-06-27T11:46:24Z | https://github.com/MonetDB/MonetDB/issues/2478 | 753,367,255 | 2,478 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-01-28 05:35:49 +0100
From: dbx5000 <<dbx5000>>
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2009-01-28 06:47:34 +0100
## Comment 13741
Date: 2009-01-28 17:35:49 +0100
From: dbx5000 <<dbx5000>>
Hi MonetDB development team,
in the documentation of MonetDB is written that the timing can be toggled on or off with the \t command.
But with this command The results look like total time(not in detail view)
How can I get the Time in seperate Form (Trans, Shred,Query,Print..)?
regards,
## Comment 13742
Date: 2009-01-28 18:47:34 +0100
From: @drstmane
in the current Nov2008 (SP2) release, detailed timing is not yet supported with the default "Algebra" back-end but only with the old "milprint_summer" back-end:
========
$ mclient -lxquery
xquery>\t
xquery>1+2
more><>
3
Timer 7.534 msec
xquery>\g
xquery>1+2
more><>
3
Trans 4.361 msec
Shred 0.000 msec
Query 2.395 msec
Print 2.009 msec
Timer 11.686 msec
xquery>\G
xquery>1+2
more><>
3
Timer 7.509 msec
xquery>
========
In the Current CVS HEAD, and hence in the next release, also the "Algebra" back-end will feature detailed timing:
========
$ mclient -lxquery
xquery>\t
xquery>1+2
more><>
3
Trans 15.411 msec
Shred 0.000 msec
Query 3.620 msec
Print 0.344 msec
Timer 24.118 msec
xquery>\g
xquery>1+2
more><>
3
Trans 11.188 msec
Shred 0.000 msec
Query 2.496 msec
Print 2.244 msec
Timer 19.698 msec
xquery>\G
xquery>1+2
more><>
3
Trans 1.611 msec
Shred 0.000 msec
Query 2.864 msec
Print 0.246 msec
Timer 7.959 msec
xquery>
========
## Comment 13743
Date: 2009-01-29 09:12:37 +0100
From: dbx5000 <<dbx5000>>
Thank you very much for your quickly answer
## Comment 13744
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2543381 at http://sourceforge.net/support/tracker.php?aid=2543381
| Timing | https://api.github.com/repos/MonetDB/MonetDB/issues/2477/comments | 0 | 2020-11-30T10:45:59Z | 2024-06-28T13:42:09Z | https://github.com/MonetDB/MonetDB/issues/2477 | 753,367,176 | 2,477 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-01-27 08:47:35 +0100
From: Hans <<hans_0_>>
To: MonetDB4 devs <<bugs-monetdb4>>
Version: -- development
Last updated: 2009-06-04 10:09:54 +0200
## Comment 13737
Date: 2009-01-27 08:47:35 +0100
From: Hans <<hans_0_>>
Please add a --version option to Mserver (v4) so it can display its version information, e.g. 4.26.5 and maybe a build date. The output format shall be so that it can be easily parsed, e.g. for version number comparison. This option should be independent of other options and environment, so Mserver should not try to read the config file.
I already requested this by email to MonetDB-users on monday 20090126 16:34.
## Comment 13738
Date: 2009-04-28 14:30:47 +0200
From: @grobian
Mserver --version now returns output like mserver5 --version. Is that enough?
## Comment 13739
Date: 2009-06-04 22:09:48 +0200
From: @grobian
assuming yes
## Comment 13740
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2539718 at http://sourceforge.net/support/tracker.php?aid=2539718
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-monetdb4@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Mserver should have a version display option | https://api.github.com/repos/MonetDB/MonetDB/issues/2476/comments | 0 | 2020-11-30T10:45:56Z | 2024-06-28T13:42:09Z | https://github.com/MonetDB/MonetDB/issues/2476 | 753,367,146 | 2,476 |
[
"MonetDB",
"MonetDB"
] | Date: 2009-01-27 04:52:36 +0100
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2011-01-25 15:31:16 +0100
## Comment 13730
Date: 2009-01-27 04:52:36 +0100
From: @skinkie
CREATE TABLE nodes_legacy (id integer, long double, lat double, username varchar(255), timestamp timestamptz);
COPY 6833589 RECORDS INTO nodes_legacy from '/mnt/dv/adt/temp/tmp/nodes.csv' USING DELIMITERS ',', '\n', '''';
!SQLException:sql:value '2008-10-11T06:54:34Z' while parsing '2008-10-11T06:54:34Z'' from line 0 field 4 not inserted, expecting type timestamp
!SQLException:importTable:failed to import table
35835479, 51.9745073, 7.2971406, 'Fam-Os', '2008-10-11T06:54:34Z'
53989466, 51.974842, 7.2967195, 'Fam-Os', '2008-10-11T06:54:34Z'
53989477, 51.9781241, 7.2954519, 'Fam-Os', '2008-10-11T06:54:34Z'
53989479, 51.9789966, 7.2952828, 'Fam-Os', '2008-10-11T06:54:34Z'
53989481, 51.9795392, 7.2952676, 'Fam-Os', '2008-10-11T06:54:34Z'
53989483, 51.9805516, 7.294534, 'Fam-Os', '2008-10-11T06:54:34Z'
53989485, 51.981326, 7.2940894, 'Fam-Os', '2008-10-11T06:54:34Z'
53989546, 51.9750623, 7.2964489, 'Fam-Os', '2008-10-11T06:54:34Z'
53989558, 51.9785337, 7.295313, 'Fam-Os', '2008-10-11T06:54:34Z'
53989562, 51.9778826, 7.2955678, 'Fam-Os', '2008-10-11T06:54:34Z'
54354966, 51.9791179, 7.2952621, 'Fam-Os', '2008-10-11T06:54:34Z'
291835846, 51.9782565, 7.2953858, 'Fam-Os', '2008-10-11T06:54:34Z'
301779697, 51.9786924, 7.2953026, 'Fam-Os', '2008-10-11T06:54:34Z'
301779712, 51.9786557, 7.295305, 'Fam-Os', '2008-10-11T06:54:34Z'
303693820, 51.9803148, 7.2947056, 'Fam-Os', '2008-10-11T06:54:34Z'
(likewise without '')
## Comment 13731
Date: 2009-01-27 05:10:58 +0100
From: @skinkie
Seems the problem fixes it self by removing the 'Z' from the data time...
## Comment 13732
Date: 2009-01-27 08:49:42 +0100
From: @lsidir
Indeed the time module does not parse the Z (at least that was the case in M4 and I am guessing that is the case also in M5). I would suggest to add the Z in the parsing of timestamps, also because the XQuery dateTimes use Z too:)
## Comment 13733
Date: 2009-02-09 15:22:24 +0100
From: @romulogoncalves
Niels,
Is this a feature request or bug?
## Comment 13734
Date: 2009-02-09 15:44:11 +0100
From: @romulogoncalves
I am making this report a feature request.
Niels (SQL god father gave green light for that)....
we need to support the Z we need to follow the Z semantics too, Z == time zone Zero).
## Comment 13735
Date: 2009-12-01 10:15:19 +0100
From: @grobian
made subject reflect what the request is about, maybe helps to get it implemented as it doesn't seem awfully hard to implement this
## Comment 13736
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2539357 at http://sourceforge.net/support/tracker.php?aid=2539357
## Comment 15433
Date: 2011-01-25 15:31:16 +0100
From: @grobian
Seems to work these days:
sql>select timestamp '2008-10-11T06:54:34Z';
+----------------------------+
| single_value |
+============================+
| 2008-10-11 06:54:34.000000 |
+----------------------------+
1 tuple (0.485ms)
| SQL: add timestamp parsing support for POSIX(?) notation (Z) | https://api.github.com/repos/MonetDB/MonetDB/issues/2475/comments | 0 | 2020-11-30T10:45:53Z | 2024-06-27T11:46:22Z | https://github.com/MonetDB/MonetDB/issues/2475 | 753,367,115 | 2,475 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-11-11 10:07:02 +0100
From: Thomas Neumann <<tneumann>>
To: @njnes
Version: -- development
Last updated: 2009-12-04 02:35:52 +0100
## Comment 13722
Date: 2008-11-11 10:07:02 +0100
From: Thomas Neumann <<tneumann>>
Bulkloading a large string table of the form (id,value) kills mserver. I attached a data generator that generates 21GB of input to illustrate the problem.
How to reproduce:
- g++ -odatagen -O2 datagen.cpp
- ./datagen > test.sql
- monetdb create foo
- monetdb start foo
- mclient -l sql -d foo test.sql
results in:
ERROR = Connection terminated
and the server crashed (on Linux x86_64).
Note that there seem to be two problems in one: The primary crash seems to be due to the "primary key" constraints in the table.
Dropping it changes the behavior: The bulkload succeeds on a machine with 32GB of main memory, but effectively stops (i.e., swap excessively) on a machine with 2GB of main memory.
Adding the primary key constrants after successful bulkloading without it crashes the server, too.
## Comment 13723
Date: 2009-02-12 08:02:49 +0100
From: @njnes
The 2GB machine was a 32 or 64 bit system? Loading on a 64 bit machine with 8G seems to work okay (including primary key).
## Comment 13724
Date: 2009-02-12 08:43:03 +0100
From: Thomas Neumann <<tneumann>>
Both machines were 64 bit systems. Note that I reported the bug before the latest release, so it might have been fixed in between (in particular since the changelog mentions changes to the data loader). I did not try the last release yet.
## Comment 13725
Date: 2009-02-12 08:47:39 +0100
From: @njnes
I'll try on a 2G machine later.
## Comment 13726
Date: 2009-02-13 08:24:45 +0100
From: @njnes
indeed on a limited machine there still is a crash. Probably when the memory (or swap) run out. Changing into a feature request as its a long term goal to improve per client resource handling and make it more safe.
## Comment 13727
Date: 2009-12-01 10:18:46 +0100
From: @grobian
a crash report can't ever be a feature request, so changed subject accordingly
## Comment 13728
Date: 2009-12-04 14:35:51 +0100
From: @mlkersten
This falls under the category hardening. Taking care of in many places.
## Comment 13729
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2261241 at http://sourceforge.net/support/tracker.php?aid=2261241
| M5: add per client resource constraints | https://api.github.com/repos/MonetDB/MonetDB/issues/2474/comments | 0 | 2020-11-30T10:45:50Z | 2024-06-28T13:42:08Z | https://github.com/MonetDB/MonetDB/issues/2474 | 753,367,072 | 2,474 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-11-10 05:53:32 +0100
From: @sjoerdmullender
To: XQuery devs <<bugs-xquery>>
Version: -- development
Last updated: 2011-04-28 16:38:53 +0200
## Comment 13719
Date: 2008-11-10 17:53:32 +0100
From: @sjoerdmullender
We need an interface in XQuery so that we can add a document which is on the client side to a collection without the client having to implement an http server.
Currently pf:add-doc takes a URI (or filename) which means that the server needs to go and read from some http server.
One possible solution might be to introduce an XQuery level function
declare document management function pf:add-doc-from-stream($name as string, $collection as string, $percentage as integer)
which reads a document from the input stream. This function cannot be loop-lifted. The query file might look something like this:
pf:add-doc-from-stream("document.xml","collection",0)
<>
<document>
...
</document>
<>
mclient would probably have to be changed somewhat to support this, but there are probably more changes needed in the server.
## Comment 13720
Date: 2008-11-17 18:07:35 +0100
From: @sjoerdmullender
Peter's idea for solving the dump/restore issue is something like:
- dump database into one huge XML document describing all relevant parameters (somewhat like the current XRPC-based dump, except everything into one document);
- restore this using mclient -I... -C... into one document;
- use XQuery to split into multiple files on the server;
- shred these files into their final destination;
- clean up by restarting server.
## Comment 13721
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2256626 at http://sourceforge.net/support/tracker.php?aid=2256626
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-xquery@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 15757
Date: 2011-04-28 16:38:53 +0200
From: @sjoerdmullender
Development on MonetDB4 and MonetDB/XQuery has ceased. It is unlikely that this bug will be fixed.
The last release of MonetDB/XQuery and MonetDB4 is Mar2011.
| PF: add document from mclient | https://api.github.com/repos/MonetDB/MonetDB/issues/2473/comments | 0 | 2020-11-30T10:45:47Z | 2024-06-28T13:42:07Z | https://github.com/MonetDB/MonetDB/issues/2473 | 753,367,032 | 2,473 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-11-06 02:34:19 +0100
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
CC: @njnes
Last updated: 2012-05-09 17:24:21 +0200
## Comment 13712
Date: 2008-11-06 14:34:19 +0100
From: @skinkie
Currently it looks like any operation on a view is stacked; the view-query is executed along the query on the view. I wonder, especially for aggregated values such as created with GROUP BY, if it isn't a better solution to materialize them instead of recalculating them.
## Comment 13713
Date: 2008-11-18 16:09:45 +0100
From: @mlkersten
Observation confirmed. The commonTerm optimizer should
have removed the duplicate code.
## Comment 13714
Date: 2008-11-18 16:22:18 +0100
From: @skinkie
Right comment to wrong bug?
## Comment 13715
Date: 2008-11-19 08:30:47 +0100
From: @mlkersten
This can be considered a bug, because the implementation
was meant to remove all such duplicate subqueries.
A new commonterm optimizer has been created, but will
be checked in after the SIGMOD deadline to avoid
possible interference with recycler experimentation.
## Comment 13716
Date: 2008-11-19 08:38:50 +0100
From: @mlkersten
A test file has been added to test/BugTracker
## Comment 13717
Date: 2009-02-11 21:06:50 +0100
From: @njnes
This is a feature request
## Comment 13718
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2229591 at http://sourceforge.net/support/tracker.php?aid=2229591
| SQL: materialize aggregate results in a view | https://api.github.com/repos/MonetDB/MonetDB/issues/2472/comments | 0 | 2020-11-30T10:45:44Z | 2024-06-28T13:42:07Z | https://github.com/MonetDB/MonetDB/issues/2472 | 753,367,002 | 2,472 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-11-04 02:59:22 +0100
From: @grobian
To: Documentation maintainers <<bugs-docs>>
Version: unspecified
Last updated: 2009-12-01 10:21:06 +0100
## Comment 13709
Date: 2008-11-04 14:59:22 +0100
From: @grobian
Some of our tools have man-pages. We should put them in our webpage, keeping them up-to-date (autogenerated from the manpage sources, I guess).
I put the manpages for merovingian and monetdb here:
http://homepages.cwi.nl/~fabian/MonetDB/Man%20Pages.html/
better HTML generation is required, and also the man page for mclient needs to be included.
## Comment 13710
Date: 2009-12-01 10:21:06 +0100
From: @grobian
maybe in our new website, won't solve it for the current incarnation.
## Comment 13711
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2221032 at http://sourceforge.net/support/tracker.php?aid=2221032
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-docs@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| website: include html versions of man-pages | https://api.github.com/repos/MonetDB/MonetDB/issues/2471/comments | 0 | 2020-11-30T10:45:41Z | 2024-06-28T13:42:06Z | https://github.com/MonetDB/MonetDB/issues/2471 | 753,366,965 | 2,471 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-11-04 02:52:51 +0100
From: @grobian
To: Documentation maintainers <<bugs-docs>>
Version: unspecified
Last updated: 2012-02-27 11:33:41 +0100
## Comment 13707
Date: 2008-11-04 14:52:51 +0100
From: @grobian
The (API) documentation that is stored in the Java sources can easily be extracted using `ant doc` (when run in the java repository). The output of this (the doc directory) should be put into the website, e.g. under monetdb.cwi.nl/javadoc (I don't really care where) and linked from our documentation part on APIs.
The documentation should be refreshed regularly. The documentation currently lives at http://homepages.cwi.nl/~fabian/MonetDB/Java%20Package.html/ but it is not always as up-to-date as would be optimal.
## Comment 13708
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2221019 at http://sourceforge.net/support/tracker.php?aid=2221019
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-docs@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| website: include Javadoc documentation | https://api.github.com/repos/MonetDB/MonetDB/issues/2470/comments | 0 | 2020-11-30T10:45:38Z | 2024-06-28T13:42:05Z | https://github.com/MonetDB/MonetDB/issues/2470 | 753,366,919 | 2,470 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-11-04 01:07:28 +0100
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2009-12-04 11:37:47 +0100
## Comment 13698
Date: 2008-11-04 13:07:28 +0100
From: @skinkie
I think a significant performance gain can be achieved if the SQL parser would 'deduplicate' equal subqueries. The most trivial scenario would be a 'union all' with 10 equal queries next to eachother. The amount of time taken for processing this query (measured by more than 450 attributes) seems to be equal to executing each query individually. With my naive knowledge I think a programmer would 'assume' an atomic query. Hence all different subqueries should result in the same result under the hood.
Now one could argue, deduplicate at layer 8 ;) True; But it would a nifty feature if one could just get the optimization from SQL without changing queries that were explicitly executed in this sequence to align with previous results.
## Comment 13699
Date: 2008-11-18 09:39:11 +0100
From: @grobian
Perhaps it is easier to use the
USING SELECT xxx AS y
SELECT * FROM y UNION SELECT * FROM y;
or similar syntax. (I'm not sure if I got the syntax right.)
## Comment 13700
Date: 2008-11-18 09:55:17 +0100
From: @mlkersten
if the subquery is really identical then the MAL optimizer
will/should catch it. And the recycler would further reduce
the work as well.
## Comment 13701
Date: 2008-11-18 11:08:57 +0100
From: @skinkie
@Martin; Should it be explicitly enabled? Because I highly doubt it is working.
@Fabian; Of course I could catch it in the program that is actually dynamically generating the mess. But the thing is I want it to be simple to process the output.
## Comment 13702
Date: 2008-11-18 14:13:21 +0100
From: @mlkersten
The commenterm expression optimizer is on by default.
A snippet of your code to ensure that the plans are
indeed not recognized are welcome.
## Comment 13703
Date: 2009-02-12 14:25:25 +0100
From: @skinkie
create table unitest (one integer);
insert into unitest (one) values (1);
insert into unitest (one) values (2);
insert into unitest (one) values (3);
insert into unitest (one) values (4);
insert into unitest (one) values (5);
sql>explain select * from unitest;
+--------------------------------------------------------------------------------------------------------+
| function user.s22_1():void; |
| _1:bat[:oid,:int] := sql.bind("sys","unitest","one",0); |
| _6 := sql.resultSet(1,1,_1); |
| sql.rsColumn(_6,"sys.unitest","one","int",32,0,_1); |
| _1:bat[:oid,:int] := nil; |
| _12 := io.stdout(); |
| sql.exportResult(_12,_6); |
| end s22_1; |
+--------------------------------------------------------------------------------------------------------+
Timer 0.864 msec 10 rows
sql>explain select * from unitest union all select * from unitest;
+--------------------------------------------------------------------------------------------------------+
| function user.s23_1():void; |
| _1:bat[:oid,:int] := sql.bind("sys","unitest","one",0); |
| _6:bat[:oid,:int] := bat.new(nil:oid,nil:int); |
| _9 := bat.append(_6,_1,true); |
| _6:bat[:oid,:int] := nil; |
| _11 := bat.append(_9,_1,true); |
| _9 := nil; |
| _1:bat[:oid,:int] := nil; |
| _12 := sql.resultSet(1,1,_11); |
| sql.rsColumn(_12,".unitest","one","int",32,0,_11); |
| _11 := nil; |
| _18 := io.stdout(); |
| sql.exportResult(_18,_12); |
| end s23_1; |
+--------------------------------------------------------------------------------------------------------+
Timer 1.115 msec 10 rows
sql>explain select * from unitest union all select * from unitest union all select * from unitest;
+--------------------------------------------------------------------------------------------------------+
| function user.s24_1():void; |
| _1:bat[:oid,:int] := sql.bind("sys","unitest","one",0); |
| _6:bat[:oid,:int] := bat.new(nil:oid,nil:int); |
| _9 := bat.append(_6,_1,true); |
| _6:bat[:oid,:int] := nil; |
| _11 := bat.append(_9,_1,true); |
| _9 := nil; |
| _12:bat[:oid,:int] := bat.new(nil:oid,nil:int); |
| _13 := bat.append(_12,_11,true); |
| _12:bat[:oid,:int] := nil; |
| _11 := nil; |
| _14 := bat.append(_13,_1,true); |
| _13 := nil; |
| _1:bat[:oid,:int] := nil; |
| _15 := sql.resultSet(1,1,_14); |
| sql.rsColumn(_15,".unitest","one","int",32,0,_14); |
| _14 := nil; |
| _21 := io.stdout(); |
| sql.exportResult(_21,_15); |
| end s24_1; |
+--------------------------------------------------------------------------------------------------------+
Timer 1.339 msec 10 rows
## Comment 13704
Date: 2009-12-01 10:23:26 +0100
From: @grobian
apparently this is not a bug in the common term expression optimiser then?
## Comment 13705
Date: 2009-12-04 11:37:47 +0100
From: @mlkersten
Done and others are captured by the recycler.
>explain select * from unitest union all select * from unitest;
&1 0 32 1 32
% .explain table_name
% mal name
% clob type
% 0 length
=function user.s2_0{autoCommit=true}():void;
=barrier _52 := language.dataflow();
= _2:bat[:oid,:int] := sql.bind("sys","unitest","one",0);
= _7:bat[:oid,:int] := sql.bind("sys","unitest","one",2);
= _9 := algebra.kdifference(_2,_7);
= _2:bat[:oid,:int] := nil:BAT;
= _10 := algebra.kunion(_9,_7);
= _9 := nil:BAT;
= _7:bat[:oid,:int] := nil:BAT;
= _11:bat[:oid,:int] := sql.bind("sys","unitest","one",1);
= _13 := algebra.kunion(_10,_11);
= _10 := nil:BAT;
= _11:bat[:oid,:int] := nil:BAT;
= _14:bat[:oid,:oid] := sql.bind_dbat("sys","unitest",1);
= _15 := bat.reverse(_14);
= _14:bat[:oid,:oid] := nil:BAT;
= _16 := algebra.kdifference(_13,_15);
= _13 := nil:BAT;
= _15 := nil:BAT;
=exit _52;
= _17:bat[:oid,:int] := bat.new(nil:oid,nil:int);
= _20 := bat.append(_17,_16,true);
= _17:bat[:oid,:int] := nil:BAT;
= _21 := bat.append(_20,_16,true);
= _20 := nil:BAT;
= _16 := nil:BAT;
= _22 := sql.resultSet(1,1,_21);
= sql.rsColumn(_22,".unitest","one","int",32,0,_21);
= _21 := nil:BAT;
= _27 := io.stdout();
= sql.exportResult(_27,_22);
=end s2_0;
## Comment 13706
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2220597 at http://sourceforge.net/support/tracker.php?aid=2220597
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| SQL: reduce duplicate subqueries in one query | https://api.github.com/repos/MonetDB/MonetDB/issues/2469/comments | 0 | 2020-11-30T10:45:34Z | 2024-06-28T13:42:04Z | https://github.com/MonetDB/MonetDB/issues/2469 | 753,366,877 | 2,469 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-10-21 06:40:12 +0200
From: @skinkie
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2008-10-22 12:17:56 +0200
## Comment 13691
Date: 2008-10-21 18:40:12 +0200
From: @skinkie
I think it would be a great improvement if the version string (in svn it is usually named the rXXXX) plus compilation date after the actual named version. In this case when bugs are reported it is clear to see if the reported is using the latest CVS version; the assignee knows what version of the code is running. There are reasons to say that a user should always upgrade to cvs head, for me this would mean that after every problem I find I need to spend 30 minutes on upgrading the code to find it out. Might be better to just report the issue, and get it marked as fixed, if the problem is already fixed.
This will probably also make a difference for users (like me) that are compiling new code and expect a new version. If they see that the actual revision/date the code was changed for a specific subpart has not been changed between releases this could point out that something went wrong with the compilation and the user might be still running an older version.
## Comment 13692
Date: 2008-10-21 22:29:17 +0200
From: @grobian
Posting the output of `mserver5 --version` comes sort of close, as it gives us answers to some questions regarding your setup.
## Comment 13693
Date: 2008-10-21 22:39:05 +0200
From: @skinkie
MonetDB server v5.7.0 (32-bit), based on kernel v1.25.0 (32-bit oids)
Copyright (c) 1993-July 2008 CWI
Copyright (c) August 2008- MonetDB B.V., all rights reserved
Visit http://monetdb.cwi.nl/ for further information
Configured for prefix: /opt/monetdb
Libraries:
libpcre: 7.8 2008-09-05 (compiled with 7.8)
openssl: OpenSSL 0.9.8i 15 Sep 2008 (compiled with OpenSSL 0.9.8i 15 Sep 2008)
Compiled by: root@handler_osm
Compilation: gcc -Wall -Wextra -std=c99 -g -Werror-implicit-function-declaration -Werror -Wpointer-arith -Wdeclaration-after-statement -Wundef -Wp,-D_FORTIFY_SOURCE=2
Linking : /usr/i686-pc-linux-gnu/bin/ld
This would still not allow you to say what CVS I'm running night?
What I would suggest is:
MonetDB server v5.7.0-CVS XXXXXX (32-bit), based on kernel v1.25.0-CVS XXXXXX (32-bit oids)
MonetDB/SQL module v2.25.0-CVS XXXXXXX loaded
## Comment 13694
Date: 2008-10-21 23:16:54 +0200
From: @sjoerdmullender
Adding the CVS version information is not feasible: every source file has its own CVS version number and there are hundreds of source files. In CVS there is no such thing as a global version.
## Comment 13695
Date: 2008-10-21 23:31:27 +0200
From: @drstmane
In contrast to SVN, CVS has not global revision number, but each file has its own revision number; and since CVS allows updates of individual files to arbitrary revisions, the only "water proof" and hence "useful" solution would be to add and list the revision number of all (code) files --- that is a lot in MonetDB, and hence hardly feasible ...
Maybe, we could consider creating a checksum of all code files or so?
Apart from that, users are welcome to report bugs for whatever version they are using, preferably the latest "Stable" release.
In that case, the can simply say "using latest Stable release version X.Y.Z" or "CVS checkout/update from latest Stable release branches on <date>, <time>, preferably adding the output of `mserver5 --version` and `mserver5 --dbinit='include sql;'` or `Mserver --dbinit='module(pathfinder);'`.
Provided the bug report comes with sufficient instructions to (easily) reproduce the steps that triggered the bug with the user, the developers will be able to check whether it does or does not work with both the latest CVS version of the latest release branches (or the development trunk) and only once the developers have either confirmed that the problem has already been fixed, or once it is fixed, the user needs to upgrade to the latest CVS version (preferably from the latest release branches, where the problem has been fixed).
For most users, using the "Current" development trunk is only "necessary", if they indeed require the latest (not necessarily stable) "cutting-edge" features.
## Comment 13696
Date: 2008-10-22 00:17:42 +0200
From: @skinkie
Thanks for the explanation; in this case a checkout/update date doesn't make much sense nor does the actual time of compiling. Although I think it would be a useful add on; I understand its problems for implementation now. Will close then.
## Comment 13697
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2184885 at http://sourceforge.net/support/tracker.php?aid=2184885
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Add CVS version information too all versions | https://api.github.com/repos/MonetDB/MonetDB/issues/2468/comments | 0 | 2020-11-30T10:45:31Z | 2024-06-28T13:42:03Z | https://github.com/MonetDB/MonetDB/issues/2468 | 753,366,840 | 2,468 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-10-20 12:41:33 +0200
From: @skinkie
To: MonetDB4 devs <<bugs-monetdb4>>
Version: -- development
Last updated: 2009-06-04 10:11:11 +0200
## Comment 13688
Date: 2008-10-20 12:41:33 +0200
From: @skinkie
As mentioned in the initial comment by sjoerd in:
https://sourceforge.net/tracker2/?func=detail&aid=2171063&group_id=56967&atid=482468
[quote]
If you think that it should do more (such as if (fork() > 0) exit(0);), it is a feature request.
[/quote]
I think.
## Comment 13689
Date: 2009-04-28 14:31:32 +0200
From: @grobian
use merovingian
## Comment 13690
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2182087 at http://sourceforge.net/support/tracker.php?aid=2182087
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-monetdb4@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Mserver5 to daemonize | https://api.github.com/repos/MonetDB/MonetDB/issues/2467/comments | 0 | 2020-11-30T10:45:27Z | 2024-06-28T13:42:03Z | https://github.com/MonetDB/MonetDB/issues/2467 | 753,366,796 | 2,467 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-10-18 05:37:31 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2011-01-25 15:29:48 +0100
## Comment 13683
Date: 2008-10-18 05:37:31 +0200
From: @skinkie
It would be a really nice addiction to the 'helpful examples' base if some examples could be added with respect to the processing of functions. I took a peak in the sql test directory but non of the examples really fits the bill.
What I want is an equivalent of the http://dev.mysql.com/doc/refman/5.0/en/group-by-functions.htmlfunction_group-concat MySQL function, or STRAGG in Orace.
The use case is on the fly GIS creation polyline creation.
SELECT '(' || x || ',' || y || ')' INTO out FROM way_nds, nodes_legacy WHERE thisway = way AND nodes_legacy.id = to_node;
Would select a list of (x,y) that can be concatted to polyline((..), (..)) this would allow for example OGR/MapServer to make use of my dataset directly without going the way of making a hybride dataset that stores the POLYLINE(...) in my way column. Quick and dirty solutions are willingly accepted.
## Comment 13684
Date: 2008-10-18 06:05:41 +0200
From: @skinkie
I'll just make it easyer...
CREATE AGGREGATE group_concat (
BASETYPE = text,
SFUNC = _group_concat,
STYPE = text
);
How to get the above running inside Monet?
Below is what already works:
CREATE FUNCTION _group_concat(input1 text, input2 text)
RETURNS text
BEGIN
RETURN SELECT CASE
WHEN input2 IS NULL THEN input1
WHEN input1 IS NULL THEN input2
ELSE input1 || ',' || input2
END;
END;
## Comment 13685
Date: 2008-10-19 05:10:36 +0200
From: @skinkie
Application layer solution suffices for now.
## Comment 13686
Date: 2009-12-01 10:27:22 +0100
From: @grobian
I don't see how this would be a MonetDB request, if it seems that this should be handled in the application layer.
## Comment 13687
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2175942 at http://sourceforge.net/support/tracker.php?aid=2175942
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 15432
Date: 2011-01-25 15:29:48 +0100
From: @grobian
wont fix for now. We accept patches though.
| SQL: string aggregate function (group_concat) | https://api.github.com/repos/MonetDB/MonetDB/issues/2466/comments | 0 | 2020-11-30T10:45:25Z | 2024-06-28T13:42:02Z | https://github.com/MonetDB/MonetDB/issues/2466 | 753,366,760 | 2,466 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-10-12 05:23:21 +0200
From: Carl J. Nobile <<cnobile>>
To: @njnes
Version: -- development
Last updated: 2009-02-11 09:09:09 +0100
## Comment 13680
Date: 2008-10-12 05:23:21 +0200
From: Carl J. Nobile <<cnobile>>
Neils, In connection with the bug I found where adding a column during updates then deleting caused the DB to loose track of the columns in subsequent selections, I realized that there is another issue which is maybe a feature request which is why it is here.
In a sequence where a column needs to be added, as of now, one needs to commit before data can be inserted into the db. This is not optimal since adding the column is really part of the transaction and at least two transactions are needed to do what should be done in one maybe the process non-atomic. The problem also comes into view when the sys.columns view is accessed and the newly added column is not there yet.
During the fix to the bug I found above could this also be fixed? I know this may not be easy, but I felt it should be stated non-the-less.
-Carl
## Comment 13681
Date: 2009-02-11 21:09:08 +0100
From: @njnes
fixed. New columns/indices created through alter are now always 'updates' (ie initial bats are created with equal lengths as the other bats in the table). This make sure all bats stay aligned.
## Comment 13682
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2160624 at http://sourceforge.net/support/tracker.php?aid=2160624
| Transactions | https://api.github.com/repos/MonetDB/MonetDB/issues/2465/comments | 0 | 2020-11-30T10:45:21Z | 2024-06-28T13:42:01Z | https://github.com/MonetDB/MonetDB/issues/2465 | 753,366,727 | 2,465 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-10-12 03:34:49 +0200
From: Carl J. Nobile <<cnobile>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2008-11-14 05:12:22 +0100
## Comment 13677
Date: 2008-10-12 03:34:49 +0200
From: Carl J. Nobile <<cnobile>>
Neils, In connection with the bug I found where adding a column during updates then deleting caused the DB to loose track of the columns in subsequent selections, I realized that there is another issue which is maybe a feature request which is why it is here.
In a sequence where a column needs to be added, as of now, one needs to commit before data can be inserted into the db. This is not optimal since adding the column is really part of the transaction and at least two transactions are needed to do what should be done in one maybe the process non-atomic. The problem also comes into view when the sys.columns view is accessed and the newly added column is not there yet.
During the fix to the bug I found above could this also be fixed? I know this may not be easy, but I felt it should be stated non-the-less.
-Carl
## Comment 13678
Date: 2008-11-14 17:11:43 +0100
From: @sjoerdmullender
Duplicate of 2160624.
## Comment 13679
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2160475 at http://sourceforge.net/support/tracker.php?aid=2160475
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Transactions | https://api.github.com/repos/MonetDB/MonetDB/issues/2464/comments | 0 | 2020-11-30T10:45:18Z | 2024-06-28T13:42:52Z | https://github.com/MonetDB/MonetDB/issues/2464 | 753,366,692 | 2,464 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-10-09 03:02:48 +0200
From: @skinkie
To: @njnes
Version: -- development
Last updated: 2009-12-04 04:28:20 +0100
## Comment 13671
Date: 2008-10-09 03:02:48 +0200
From: @skinkie
I'm trying to debug an issue with probably a duplicate primary key. For that I reduce the actual amount of rows I want to insert using copy into.
insertme4.sql:
START TRANSACTION;
CREATE TABLE way_tags (way integer, k varchar(255), v varchar(1024), primary key (way, k), foreign key(way) references ways);
COPY 2000000 RECORDS INTO way_tags from '/mnt/dv/split/way_tags.csv' USING DELIMITERS ',', '\n', '''';
COMMIT;
If I execute this code multiple times I end up in 4000000 inserted rows. The strange thing is, when I actually try to insert 4000000 rows I end up in:
ERROR = !SQLException:assert:INSERT INTO: PRIMARY KEY constraint 'way_tags.way_tags_way_k_pkey' violated
(the same error as I had before)
Now because that number was so nice round and pretty I have tried to insert 2000001 tags where I end up in:
4000002
But this actually *matches* what is inserted in that table:
sql>select count(*) from way_tags;
+---------+
| count_ |
+=========+
| 4000002 |
+---------+
Feature request:
From debugging perspective it would be very nice if a copy into primary key problem also returned the actual 'line' or broken thing.
## Comment 13672
Date: 2008-12-01 16:14:26 +0100
From: @romulogoncalves
I propose to change this report to a feature request report....
Niels,
Do you agree?
## Comment 13673
Date: 2008-12-01 19:48:27 +0100
From: @romulogoncalves
This report became a feature request report instead of bug report
## Comment 13674
Date: 2009-12-01 10:29:37 +0100
From: @grobian
I think the new copy into code is a bit more friendly about reporting errors, but someone who worked on that should give the definite answer.
## Comment 13675
Date: 2009-12-04 16:28:20 +0100
From: @mlkersten
This feature has been implemented long time ago.
The erroneous place (one of them) is reported as line..field
combination.
## Comment 13676
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2154105 at http://sourceforge.net/support/tracker.php?aid=2154105
| SQL: COPY INTO report line number or row on failure | https://api.github.com/repos/MonetDB/MonetDB/issues/2463/comments | 0 | 2020-11-30T10:45:14Z | 2024-06-28T13:42:52Z | https://github.com/MonetDB/MonetDB/issues/2463 | 753,366,648 | 2,463 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-10-01 10:42:41 +0200
From: Loredana Afanasiev <<lafanasi>>
To: Jan Flokstra <<jflokstra>>
Version: -- development
Last updated: 2008-11-14 05:14:09 +0100
## Comment 13661
Date: 2008-10-01 22:42:41 +0200
From: Loredana Afanasiev <<lafanasi>>
Dear MonetDB/XQuery team,
When shredding a document that contains a document type declaration referring a dtd file given by a relative path, MonetDB returns a "could not find the dtd file" error.
For example, shredding a document that contains this declaration:
<!DOCTYPE handeling SYSTEM "handeling.dtd">,
generates this error:
MAPI = monetdb@localhost:50000QUERY = let $path := "/scratch/PolitiekeData/PoliticalMashup/HAN"ERROR = !ERROR: I/O warning : failed to load external entity "handeling.dtd" !ERROR: Start tag expected, '<' not found
!ERROR: [shred_url]: 1 times inserted nil due to errors at tuples 0@0.
!ERROR: [shred_url]: first error was:
!ERROR: shred_external_subset: WARNING: xmlParseDTD("handeling.dtd") FAILED, NO ID/IDREF QUERIES
!ERROR: shred_external_subset: NOTE : maybe using absolute filenames works, sorry!
!ERROR: CMDshred_url: operation failed.
Given a big collection of such documents, it is not pleasant/clean to change the dtd file into an absolute path. It would be great if MonetDB would accept relative dtd paths also.
Thank you for consideration,
l.
## Comment 13662
Date: 2008-10-01 22:58:35 +0200
From: @drstmane
Question: relative to what?
## Comment 13663
Date: 2008-10-01 23:22:06 +0200
From: @sjoerdmullender
The document being shredded.
## Comment 13664
Date: 2008-10-01 23:24:21 +0200
From: @drstmane
We rely on libxml2 to shred documents, and it could in fact be that libxml2 requires absolute paths (or actually URIs!) for DTDs.
Jan (F),
could check and comment?
Stefan
## Comment 13665
Date: 2008-10-02 08:57:39 +0200
From: Jan Flokstra <<jflokstra>>
I remember I had a look at this problem before. I think the problem was that libxml looks relative to the current directory and because we execute from within the database server the concept of "./" does not work. Also a problem is that DTD loading is performed in an 'automagic' way by libxml2 so you have to find out how to influence the search for dtd's in an indirect way,
JanF.
## Comment 13666
Date: 2008-10-02 09:21:05 +0200
From: Jan Flokstra <<jflokstra>>
I remember I had a look at this problem before. I think the problem was that libxml looks relative to the current directory and because we execute from within the database server the concept of "./" does not work. Also a problem is that DTD loading is performed in an 'automagic' way by libxml2 so you have to find out how to influence the search for dtd's in an indirect way,
JanF.
## Comment 13667
Date: 2008-10-15 15:25:18 +0200
From: @drstmane
Loredana,
JanF just implemented the following in the "Current" development trunk of pathfinder (CVS HEAD):
- implement request [ monetdb-Feature Requests-2141068 ] relative path to a DTD
When an external DTD has to be loaded in the current version only absolute
path's may be used. This new implementation allows to relative path for
external DTD;s. When a relative path is used the shredder will lookup the
DTD relative to the location of the shredded xml document. The implementation
also works when the xml document is a URL.
Could you please check, whether this fulfills your request, and report?
Thanks!
Stefan
## Comment 13668
Date: 2008-11-02 20:57:41 +0100
From: Loredana Afanasiev <<lafanasi>>
Hi Stefan, JanF,
sorry for taking so long. I installed the "current" pf and it loaded my collection without problems. Thank you a lot for implementing my request promptly!
l.
## Comment 13669
Date: 2008-11-14 17:14:26 +0100
From: @sjoerdmullender
Closed since this has been implemented.
## Comment 13670
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2141068 at http://sourceforge.net/support/tracker.php?aid=2141068
| relative path to a DTD | https://api.github.com/repos/MonetDB/MonetDB/issues/2462/comments | 0 | 2020-11-30T10:45:11Z | 2024-06-28T13:42:51Z | https://github.com/MonetDB/MonetDB/issues/2462 | 753,366,612 | 2,462 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-09-10 11:33:39 +0200
From: @skinkie
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2011-01-25 15:28:31 +0100
## Comment 13652
Date: 2008-09-10 23:33:39 +0200
From: @skinkie
How difficult would it be to accept the following format? A hackish solution would be accepted:
2008-07-25T15:54:39
The following is accepted in Monet:
2008-07-25 15:54:39
## Comment 13653
Date: 2008-09-10 23:39:20 +0200
From: @mlkersten
This is hackish.
What is the rationale? What standard requires this?
What system recognizes this?
## Comment 13654
Date: 2008-09-11 00:01:34 +0200
From: @skinkie
We call it XML... it is around us everywhere without any proofs it is an actual standard. But the acceptation of a space vs 'T' wouldn't be that hard, I just cannot find how sql_find_subtype/sql_init_subtype defines this...
## Comment 13655
Date: 2008-09-11 00:05:39 +0200
From: @skinkie
To be factually correct it is ISO 8601 combined time date.
* Date and time represents a specified time of a specified day. When use is made of the calendar date the representation is:
YYYY-MM-DDThh:mm:ss
where the capital letter T is used to separate the date and time components. Thus, for a very precise date and time, look at this:
Example: 2003-04-01T13:01:02 represents one minute and two seconds after one o'clock in the afternoon of 2003-04-01.
## Comment 13656
Date: 2008-09-11 00:31:12 +0200
From: @skinkie
File Added: iso8601.patch
## Comment 13657
Date: 2008-09-11 08:59:07 +0200
From: @grobian
Is this just because you do a COPY INTO? (How about sed somewhere in the pipeline in that case?) If not, can't you use an SQL string search-and-replace function to do it? We try to stick to the (ugly) SQL standard.
## Comment 13658
Date: 2009-12-01 10:32:47 +0100
From: @grobian
This seems like a duplicate of 2539357. It's just requesting the timestamp parser to be a bit more flexible.
## Comment 13659
Date: 2010-02-19 20:16:19 +0100
From: Mitja Kleider <<mkleider>>
I think this is working now.
## Comment 13660
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2104442 at http://sourceforge.net/support/tracker.php?aid=2104442
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 15431
Date: 2011-01-25 15:28:31 +0100
From: @grobian
this works like a charm these days:
sql>select timestamp '2008-07-25T15:54:39';
+----------------------------+
| single_value |
+============================+
| 2008-07-25 15:54:39.000000 |
+----------------------------+
1 tuple (0.175ms)
| SQL: different timestamptz format | https://api.github.com/repos/MonetDB/MonetDB/issues/2461/comments | 0 | 2020-11-30T10:45:07Z | 2024-06-27T11:46:08Z | https://github.com/MonetDB/MonetDB/issues/2461 | 753,366,568 | 2,461 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-08-26 07:57:33 +0200
From: Qian Qiao <<drizzlerain>>
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2011-01-25 15:26:24 +0100
## Comment 13647
Date: 2008-08-26 07:57:33 +0200
From: Qian Qiao <<drizzlerain>>
OS:RedHat Linux AS 4
GCC:3.4.6
1.setup:
CREATE TABLE T850 (
T850KEY INT NOT NULL UNIQUE,
T850C CHAR (11),
T850VC VARCHAR (10));
2.insert:
INSERT INTO T850 VALUES (10, '1234567890', '1234567890');
INSERT INTO T850 VALUES (5, '12345 ', '12345');
3.result:
SELECT COUNT(*) FROM T850 WHERE T850C = T850VC;
+--------+
| count_ |
+========+
| 1 |
+--------+
4.question:
count should be 2.
## Comment 13648
Date: 2008-08-26 09:12:21 +0200
From: @grobian
Logged In: YES
user_id=963970
Originator: NO
I think the output is correct. Trailing spaces /are/ included in the string if it is of type CHAR. If you want to compare to VARCHAR, use LTRIM or something.
## Comment 13649
Date: 2008-08-26 10:10:55 +0200
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
See also
[ 2075268 ] About CHAR
http://sourceforge.net/tracker/index.php?func=detail&aid=2075268&group_id=56967&atid=482468
## Comment 13650
Date: 2008-08-26 14:01:40 +0200
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
Test added to src/tests/BugTracker-2008/Tests/varchar_char_semantics.SF-2075085.sql
sql defines CHAR(x) as a fixed length type with implicit collation 'pad space'
which means its filled up till the x length with spaces.
MonetDB doesn't follow this definition, it puts char into varchar and takes
the varchar specification for it too. So no space padding.
This is done for 2 main reasons
1) we don't support 'collations'
2) we currently don't have a native fixed with character data type.
(so we move this bug to a feature request!)
## Comment 13651
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2075085 at http://sourceforge.net/support/tracker.php?aid=2075085
## Comment 15430
Date: 2011-01-25 15:26:24 +0100
From: @grobian
Seems unlikely to me we will ever fix/implement this.
| SQL: support native (fixed size) CHAR type | https://api.github.com/repos/MonetDB/MonetDB/issues/2460/comments | 0 | 2020-11-30T10:45:04Z | 2024-06-28T13:42:50Z | https://github.com/MonetDB/MonetDB/issues/2460 | 753,366,532 | 2,460 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-08-24 04:26:47 +0200
From: @skinkie
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2009-06-04 10:24:12 +0200
## Comment 13640
Date: 2008-08-24 16:26:47 +0200
From: @skinkie
Would it be possible to output early results instead of waiting for the complete output?
## Comment 13641
Date: 2008-08-24 16:29:51 +0200
From: @grobian
Logged In: YES
user_id=963970
Originator: NO
You're probably after a server change, not a client change.
## Comment 13642
Date: 2008-08-24 16:57:42 +0200
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
To be a bit more verbose: MonetDB uses (for good reasons; see our research literature) "bulk processing" (as opposed to pipelining in several other DBMSs); hence, there are now "early results" that the client could output; no single result has been produced before the last operator in the respective MIL/MAL (binary algebra) plan is starting to execute; the time span from "producing" the first to the last result item is only the time the very last operator in the generated MIL/MAL (binary algebra) plan takes to execute --- usually only a faction of the total query execution time ...
## Comment 13643
Date: 2008-08-24 16:59:12 +0200
From: @skinkie
Logged In: YES
user_id=533026
Originator: YES
I just reported a bug on that, the server output is broken.
## Comment 13644
Date: 2009-06-04 22:13:14 +0200
From: @grobian
Is this feasible at all to keep as request? I think we will never do this as it is contrary to our model.
## Comment 13645
Date: 2009-06-04 22:24:12 +0200
From: @sjoerdmullender
Results are returned as soon as they are available.
Closing as the bulk processing paradigm isn't going to be changed.
## Comment 13646
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2071611 at http://sourceforge.net/support/tracker.php?aid=2071611
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| mclient early results | https://api.github.com/repos/MonetDB/MonetDB/issues/2459/comments | 0 | 2020-11-30T10:45:00Z | 2024-06-28T13:42:50Z | https://github.com/MonetDB/MonetDB/issues/2459 | 753,366,498 | 2,459 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-08-21 01:26:31 +0200
From: @skinkie
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2010-08-20 14:03:55 +0200
## Comment 13637
Date: 2008-08-21 13:26:31 +0200
From: @skinkie
select * from nodes_legacy where x between -10 and 11 and y -1 and 1;
!predicate BINOP 95: time to implement some more
## Comment 13638
Date: 2008-08-21 13:31:46 +0200
From: @skinkie
Logged In: YES
user_id=533026
Originator: YES
nvm
## Comment 13639
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2064570 at http://sourceforge.net/support/tracker.php?aid=2064570
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
Bug has invalid status, setting status to "NEW".
Previous status was "DELETED".
| More betweens | https://api.github.com/repos/MonetDB/MonetDB/issues/2458/comments | 0 | 2020-11-30T10:44:56Z | 2024-06-28T13:42:49Z | https://github.com/MonetDB/MonetDB/issues/2458 | 753,366,472 | 2,458 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-08-16 08:07:22 +0200
From: @mlkersten
To: @mlkersten
Version: -- development
Last updated: 2009-12-04 02:41:15 +0100
## Comment 13634
Date: 2008-08-16 20:07:22 +0200
From: @mlkersten
observed 13-08-2008: printf accepts a long for an int formatter,
whereas it should complain (additionally the error could be clearer)
l := 12:lng;
f := 5:flt;
io.printf("%d\n", l);
io.printf("%d\n", f);
Ideally the format string should be checked against
the type and generalized to accept a variable argument list.
## Comment 13635
Date: 2009-12-04 14:41:15 +0100
From: @mlkersten
obsolete. The printf function code was forwarded from M4.
## Comment 13636
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2054978 at http://sourceforge.net/support/tracker.php?aid=2054978
| M5: io.printf extended checks | https://api.github.com/repos/MonetDB/MonetDB/issues/2457/comments | 0 | 2020-11-30T10:44:53Z | 2024-06-28T13:42:48Z | https://github.com/MonetDB/MonetDB/issues/2457 | 753,366,439 | 2,457 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-08-03 09:43:10 +0200
From: @mlkersten
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2009-06-04 10:13:50 +0200
## Comment 13630
Date: 2008-08-03 09:43:10 +0200
From: @mlkersten
With the major upcoming change of the BAT administration, it becomes mandatory to also design and implement a scheme that will simplify future changes in both low level storage, MAL function renames and SQL catalog changes.
Such a hook ensures that users don't stick too long on old versions.
## Comment 13631
Date: 2009-04-28 14:36:47 +0200
From: @grobian
isn't this currently policy enforced by Sjoerd?
## Comment 13632
Date: 2009-06-04 22:13:50 +0200
From: @grobian
is policy now.
## Comment 13633
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 2036498 at http://sourceforge.net/support/tracker.php?aid=2036498
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Automatic rollforward | https://api.github.com/repos/MonetDB/MonetDB/issues/2456/comments | 0 | 2020-11-30T10:44:50Z | 2024-06-28T13:42:48Z | https://github.com/MonetDB/MonetDB/issues/2456 | 753,366,398 | 2,456 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-06-09 03:54:17 +0200
From: @skinkie
To: clients devs <<bugs-clients>>
Version: -- development
Last updated: 2011-01-25 15:24:55 +0100
## Comment 13627
Date: 2008-06-09 15:54:17 +0200
From: @skinkie
It would be nice if the prepare statement could work like a printf.
## Comment 13628
Date: 2009-12-01 10:40:46 +0100
From: @grobian
patch welcome
## Comment 13629
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1988910 at http://sourceforge.net/support/tracker.php?aid=1988910
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-clients@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| MAPI: add mapi_query_va | https://api.github.com/repos/MonetDB/MonetDB/issues/2455/comments | 0 | 2020-11-30T10:44:47Z | 2024-06-28T13:42:47Z | https://github.com/MonetDB/MonetDB/issues/2455 | 753,366,361 | 2,455 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-06-08 08:32:08 +0200
From: @mlkersten
To: SQL devs <<bugs-sql>>
Version: -- development
CC: ajdamico, alexandr.kulik1988, hs.addington, @njnes, percy, yzan
Last updated: 2019-01-21 15:36:55 +0100
## Comment 13624
Date: 2008-06-08 08:32:08 +0200
From: @mlkersten
MonetDB should support the median and quantile functions
over groups. Perhaps even more of the standard statistical operators found in e.g. spreadsheets.
## Comment 13625
Date: 2008-06-15 12:22:51 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
MySQL and Oracle also provide stddev() and variance()
## Comment 13626
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1987928 at http://sourceforge.net/support/tracker.php?aid=1987928
## Comment 14933
Date: 2010-09-23 23:03:13 +0200
From: Alexandr Kulik <<alexandr.kulik1988>>
Created attachment 36
Patch for OCT-2010
Implementation of standard deviation for sample (stdev) and population (stdevp); variance for sample(vars) and population(varp). This function also available from sql as aggregation function.
> Attached file: [stdev+var](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_2454_stdev+var_36) (text/plain, 13050 bytes)
> Description: Patch for OCT-2010
## Comment 14934
Date: 2010-09-24 09:40:58 +0200
From: @njnes
The patch looks nice, but it misses the very important version's for group-ed aggregation. For example, there is an implementation in group.mx (GRPvar*), which seems close.
## Comment 14935
Date: 2010-09-24 09:45:05 +0200
From: Alexandr Kulik <<alexandr.kulik1988>>
(In reply to comment 4)
> The patch looks nice, but it misses the very important version's for group-ed
> aggregation. For example, there is an implementation in group.mx (GRPvar*),
> which seems close.
So should i create something similar to GRPvar_ for stdev and variance for sample?
## Comment 14936
Date: 2010-09-24 09:58:39 +0200
From: @grobian
Problem with the patch is that it changes the catalog. Doesn't that cause trouble Niels? Probably the functions should be added with a .sql file which declares them as external (=MAL) functions.
## Comment 14937
Date: 2010-09-24 10:13:30 +0200
From: Alexandr Kulik <<alexandr.kulik1988>>
(In reply to comment 6)
> Problem with the patch is that it changes the catalog. Doesn't that cause
> trouble Niels? Probably the functions should be added with a .sql file which
> declares them as external (=MAL) functions.
I try do do it from sql, but for a some reason i can't do it. SQL doesn't allow to create aggregation function.
## Comment 15004
Date: 2010-10-06 23:31:38 +0200
From: Alexandr Kulik <<alexandr.kulik1988>>
Created attachment 43
Standard deviation and variance function for MonetDB2010
http://bugs.monetdb.org/attachment.cgi?id=36 was updated now it support group by.
> Attached file: [stdev_var_oct2010_v2](https://github.com/MonetDB/monetdb-issues-attachments/blob/master/issue_2454_stdev_var_oct2010_v2_43) (text/plain, 108675 bytes)
> Description: Standard deviation and variance function for MonetDB2010
## Comment 15005
Date: 2010-10-06 23:33:15 +0200
From: Alexandr Kulik <<alexandr.kulik1988>>
I updated functions Standard deviation and variance, so now they support group by, could you please review it. http://bugs.monetdb.org/attachment.cgi?id=43
(In reply to comment 4)
> The patch looks nice, but it misses the very important version's for group-ed
> aggregation. For example, there is an implementation in group.mx (GRPvar*),
> which seems close.
## Comment 15125
Date: 2010-10-27 16:33:55 +0200
From: Alexandr Kulik <<alexandr.kulik1988>>
Any comments on latest patch?
## Comment 15429
Date: 2011-01-25 15:24:22 +0100
From: @grobian
Niels, ping?
## Comment 15505
Date: 2011-03-03 01:03:05 +0100
From: hsaddington <<hs.addington>>
Fabian:
We have been using the statistical functions that Alex wrote for us and they are working well for us. When the functions were used with "group by", the Oct2010 Release instantly crashed. But, the functions work just fine in Oct2010-SP1 release. Is it possible to have these functions included in the next release? Thanks.
## Comment 15506
Date: 2011-03-03 08:52:29 +0100
From: @grobian
Niels has to review it. It won't make it for the Mar2011 release, perhaps the Apr2011 release.
(Don't CC me, I'm on the alias)
## Comment 15678
Date: 2011-03-28 17:36:29 +0200
From: @sjoerdmullender
The Mar2011 version has been released.
## Comment 15712
Date: 2011-04-01 10:03:17 +0200
From: Alexandr Kulik <<alexandr.kulik1988>>
Did you already commit this patch into repo? Because i try to find functions by name and can't find them.
## Comment 16033
Date: 2011-07-29 11:00:02 +0200
From: @sjoerdmullender
Apr2011-SP2 has been released.
## Comment 16271
Date: 2011-09-16 15:10:37 +0200
From: @sjoerdmullender
The Aug2011 version has been released.
## Comment 17246
Date: 2012-05-09 17:38:34 +0200
From: @njnes
stddev and median have been added.
## Comment 17877
Date: 2012-11-05 12:32:22 +0100
From: Anthony Damico <<ajdamico>>
Hi Niels,
The stddev() function does not appear to be available in the Windows version of MonetDB?
When I run these commands, I get the following error when I hit stddev..
sql>select * from tables limit 3;
...
3 tuples (21.684ms)
sql>select sum(id) from tables;
...
1 tuple (22.006ms)
sql>select avg(id) from tables;
...
1 tuple (76.176ms)
sql>select median(id) from tables;
...
1 tuple (7.618ms)
sql>select stddev(id) from tables;
SELECT: no such unary operator 'stddev(int)'
Here's a copy of my server window, which does have a release date after comment 18 below (when this issue was marked as resolved) --
MonetDB 5 server v11.11.11 "Jul2012-SP2"
Serving database 'demo', using 2 threads
Compiled for x86_64-pc-winnt/64bit with 64bit OIDs dynamically linked
Found 3.905 GiB available main-memory.
Copyright (c) 1993-July 2008 CWI.
Copyright (c) August 2008-2012 MonetDB B.V., all rights reserved
Visit http://www.monetdb.org/ for further information
Listening for connection requests on mapi:monetdb://127.0.0.1:50000/
MonetDB/SQL module loaded
>
On this topic, is it possible to also add some sort of quantile() function so that users can access something like the 25th, 75th, 90th percentiles quickly? This seems like just a small variation on the median() function (which does work), since quantile( colname , 50 ) would be the same as median( colname )?
My apologies if I'm overlooking something obvious, I am pretty new to MonetDB.
Thanks for the amazing software!!
Anthony Damico
## Comment 17920
Date: 2012-11-08 12:42:22 +0100
From: Percy Wegmann <<percy>>
It appears that the problem is in sql/scripts/39_analytics.sql. It's missing bindings for stdev and stdevp.
For stdev, at least, it seems that one can mimic median with something like the following:
create aggregate stdev(val TINYINT) returns TINYINT
external name "aggr"."stddev";
create aggregate stdev(val SMALLINT) returns SMALLINT
external name "aggr"."stddev";
create aggregate stdev(val INTEGER) returns INTEGER
external name "aggr"."stddev";
create aggregate stdev(val BIGINT) returns BIGINT
external name "aggr"."stddev";
create aggregate stdev(val REAL) returns REAL
external name "aggr"."stddev";
create aggregate stdev(val DOUBLE) returns DOUBLE
external name "aggr"."stddev";
create aggregate stdev(val DATE) returns DATE
external name "aggr"."stddev";
create aggregate stdev(val TIME) returns TIME
external name "aggr"."stddev";
create aggregate stdev(val TIMESTAMP) returns TIMESTAMP
external name "aggr"."stddev";
This can of course be run through the mclient, so anyone needing to patch their environment prior to this bug being fixed might be able to just run these manually.
## Comment 20760
Date: 2015-04-04 11:09:49 +0200
From: yzan
I hate to reopen this bug but statistical functions stopped to work in MonetDB-11.19.9-x86_64-Darwin-9-bin.tar.bz2.
For example following queries:
CREATE TABLE t1 ( col1 INT, col2 DOUBLE );
INSERT INTO t1 VALUES ( 1, 3.0 ) , ( 2, 4.0 );
SELECT stddev_samp( col1 ) FROM t1;
SELECT stddev_samp( col2 ) FROM t1;
SELECT stddev_pop( col1 ) FROM t1;
SELECT stddev_pop( col2 ) FROM t1;
Return:
Error: SELECT: no such unary operator 'stddev_samp(int)'
SQLState: 22000
ErrorCode: 0
Respectively:
Error: SELECT: no such unary operator 'stddev_samp(double)'
SQLState: 22000
ErrorCode: 0
Similar error messages appear with other statistical messages mentioned at:
https://www.monetdb.org/Documentation/SQLreference/StatisticFunctions
The queries were tested on:
Version: MonetDB Database Server Toolkit v1.1 (Oct2014-SP2)
OS: OS X 10.9.5
The database was installed from archive: MonetDB-11.19.9-x86_64-Darwin-9-bin.tar.bz2
Interestingly, the queries worked with MonetDB-11.19.7-x86_64-Darwin-9-bin.tar.bz2.
## Comment 20761
Date: 2015-04-04 11:22:31 +0200
From: Anthony Damico <<ajdamico>>
this re-introduced bug is platform-specific. that code works fine for me on
MonetDB 5 server v11.19.9 "Oct2014-SP2"
Serving database 'demo', using 8 threads
Compiled for x86_64-pc-winnt/64bit with 64bit OIDs dynamically linked
Found 15.914 GiB available main-memory.
Copyright (c) 1993-July 2008 CWI.
Copyright (c) August 2008-2015 MonetDB B.V., all rights reserved
Visit http://www.monetdb.org/ for further information
Listening for connection requests on mapi:monetdb://127.0.0.1:50000/
MonetDB/SQL module loaded
## Comment 26820
Date: 2019-01-21 15:36:55 +0100
From: @sjoerdmullender
Closing since as far as I am aware, the statistical functions work on all supported platforms.
| SQL: add various statistical functions | https://api.github.com/repos/MonetDB/MonetDB/issues/2454/comments | 0 | 2020-11-30T10:44:44Z | 2024-06-27T11:46:02Z | https://github.com/MonetDB/MonetDB/issues/2454 | 753,366,320 | 2,454 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-05-28 02:55:31 +0200
From: @skinkie
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2009-12-01 11:33:35 +0100
## Comment 13620
Date: 2008-05-28 02:55:31 +0200
From: @skinkie
It would be nice to be able to format a time.
## Comment 13621
Date: 2009-12-01 11:14:52 +0100
From: @grobian
where?
## Comment 13622
Date: 2009-12-01 11:33:35 +0100
From: @grobian
timeout, need info
## Comment 13623
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1975744 at http://sourceforge.net/support/tracker.php?aid=1975744
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| TimestampTZ strftime | https://api.github.com/repos/MonetDB/MonetDB/issues/2453/comments | 0 | 2020-11-30T10:44:41Z | 2024-06-28T13:42:46Z | https://github.com/MonetDB/MonetDB/issues/2453 | 753,366,280 | 2,453 |
[
"MonetDB",
"MonetDB"
] | Date: 2008-02-13 03:18:00 +0100
From: @grobian
To: Merovingian devs <<bugs-merovingian>>
Version: -- development
Last updated: 2008-02-19 09:23:17 +0100
## Comment 13617
Date: 2008-02-13 15:18:00 +0100
From: @grobian
merovingian should listen to an environment variable (e.g. MONETDB5_CONF) which points to an monetdb5.conf file to use instead of the compile time specified sysconfdir/monetdb5.conf.
This would allow testing to test the basic functionalities of Merovingian, and to actually run the same merovingian multiple times (on other dbfarms).
## Comment 13618
Date: 2008-02-19 21:23:17 +0100
From: @grobian
Logged In: YES
user_id=963970
Originator: YES
This feature has just been implemented.
## Comment 13619
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1892752 at http://sourceforge.net/support/tracker.php?aid=1892752
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-merovingian@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Merovingian: listen to an environment variable for config | https://api.github.com/repos/MonetDB/MonetDB/issues/2452/comments | 0 | 2020-11-30T10:44:38Z | 2024-06-28T13:42:46Z | https://github.com/MonetDB/MonetDB/issues/2452 | 753,366,246 | 2,452 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-12-19 08:57:15 +0100
From: @drstmane
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2019-01-21 15:33:05 +0100
## Comment 13613
Date: 2007-12-19 08:57:15 +0100
From: @drstmane
(At least) for testing it would be handy if the set of Windows-related conditionals (currently NOT_WIN32 & NATIVE_WIN32) would be extended to also distinguish between 32-bit & 64-bit Windows, e.g.,
- NOT_WIN: set/true on all non-Windows platforms
- NOT_WIN32: set/true on all non-WIndows32 platforms (incl. Win64)
- NOT_WIN64: set/true on all non-Windows64 platforms (incl. Win32)
- NATIVE_WIN: set/true on all native Windows platforms
- NATIVE_WIN32: set/true on native 32-bit Windows platforms
- NATIVE_WIN64: set/true on native 64-bit Windows platforms
Possibly, other related conditionals and/or macros (defines) should be checked, too.
## Comment 13614
Date: 2009-06-04 22:14:44 +0200
From: @grobian
is this still an issue?
## Comment 13615
Date: 2009-12-01 10:45:14 +0100
From: @grobian
isn't this implemented currently?
## Comment 13616
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1853739 at http://sourceforge.net/support/tracker.php?aid=1853739
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 26819
Date: 2019-01-21 15:33:05 +0100
From: @sjoerdmullender
I haven't felt a need for this, so closing as won't fix.
| extending NOT_WIN32 & NATIVE_WIN32 for WIN64? | https://api.github.com/repos/MonetDB/MonetDB/issues/2451/comments | 0 | 2020-11-30T10:44:35Z | 2024-06-28T13:42:45Z | https://github.com/MonetDB/MonetDB/issues/2451 | 753,366,219 | 2,451 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-12-01 04:48:18 +0100
From: @romulogoncalves
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2016-04-11 11:46:01 +0200
## Comment 13610
Date: 2007-12-01 16:48:18 +0100
From: @romulogoncalves
MonetDB does not support constraints on temporary tables.
In the standard the temporary tables can contains constraints.
Regards,
Romulo
## Comment 13611
Date: 2007-12-01 18:15:20 +0100
From: @romulogoncalves
Logged In: YES
user_id=1498628
Originator: YES
Assign this to make Niels happy ;)
Regards,
Romulo
## Comment 13612
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1842353 at http://sourceforge.net/support/tracker.php?aid=1842353
## Comment 18038
Date: 2012-11-27 11:53:11 +0100
From: @romulogoncalves
Niels,
Is in your plans to support constraints in temporary tables?
Otherwise, I will mark this bug as resolved and wont fix.
| SQL: allow constraints on temporary tables | https://api.github.com/repos/MonetDB/MonetDB/issues/2450/comments | 0 | 2020-11-30T10:44:31Z | 2024-06-28T13:42:44Z | https://github.com/MonetDB/MonetDB/issues/2450 | 753,366,180 | 2,450 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-11-24 03:47:05 +0100
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2007-11-24 04:28:06 +0100
## Comment 13607
Date: 2007-11-24 15:47:05 +0100
From: @grobian
> - Need to specify a different port
Currently merovingian is "hardcoded" to use MEROV_PORT (= 50000), it is
trivial to make this a configuration file parameter, the entire code is
already prepared for this. Please file a feature request for this if
50000 is not sufficient for you at the moment.
## Comment 13608
Date: 2007-11-24 16:27:51 +0100
From: @grobian
Logged In: YES
user_id=963970
Originator: YES
implemented in CVS HEAD
## Comment 13609
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1837593 at http://sourceforge.net/support/tracker.php?aid=1837593
| merovingian: allow to set default port | https://api.github.com/repos/MonetDB/MonetDB/issues/2449/comments | 0 | 2020-11-30T10:44:28Z | 2024-06-28T13:42:44Z | https://github.com/MonetDB/MonetDB/issues/2449 | 753,366,138 | 2,449 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-11-24 03:46:11 +0100
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2009-08-25 02:37:46 +0200
## Comment 13595
Date: 2007-11-24 15:46:11 +0100
From: @grobian
From a mail to venks:
> Following are the basic requirements:
>
> - Need to specify the data/log/check-point directories pointing to
> a different directories
merovingian cannot change these variables for each database started, it
would need to know that information. However, it just uses the
monetdb5.conf file, which would allow you to do the following:
- make dbfarm and sql_logs point to some place in the filesystem where you
want them
- create the database using monetdb create <database>
- move the created directories in dbfarm and sql_logs to the filesystem
you want them
- create symlinks for the directories just moved
It's more a workaround, I admit, but it avoids making merovingian's
configuration very complex. If you really prefer these locations to be
configurable, please file a feature request, as the monetdb tool could
in theory do things like setting properties for databases, which
merovingian uses during starting of the database. I could imagine
something much like Solaris' zfs get/set commands and the properties
with inheritance and all for each (sub) filesystem.
In fact I very much like this idea, because it enables the advanced user
to do even more customisations while the defaults should be fine for the
average user.
The properties could just be stored in some .properties file in the dbfarm/dbname dir, and managed by the monetdb tool, honored on execution by merovingian.
## Comment 13596
Date: 2009-06-19 14:53:27 +0200
From: @grobian
Property support added, including the ability to set the logdir using monetdb for an existing database.
## Comment 13597
Date: 2009-07-04 04:20:29 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13598
Date: 2009-07-04 12:30:23 +0200
From: @grobian
robot overrule
## Comment 13599
Date: 2009-07-19 04:20:12 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13600
Date: 2009-07-19 20:26:55 +0200
From: @grobian
robot re-overrule
## Comment 13601
Date: 2009-08-03 04:20:16 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13602
Date: 2009-08-03 09:13:31 +0200
From: @grobian
robot re-re-overrule
## Comment 13603
Date: 2009-08-18 04:20:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
This Tracker item was closed automatically by the system. It was
previously set to a Pending status, and the original submitter
did not respond within 14 days (the time period specified by
the administrator of this Tracker).
## Comment 13604
Date: 2009-08-18 09:36:21 +0200
From: @grobian
boe
## Comment 13605
Date: 2009-08-25 14:37:45 +0200
From: @grobian
feature was delivered in Aug2008 release
## Comment 13606
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1837591 at http://sourceforge.net/support/tracker.php?aid=1837591
| merovingian: extensive properties support | https://api.github.com/repos/MonetDB/MonetDB/issues/2448/comments | 0 | 2020-11-30T10:44:25Z | 2024-06-28T07:23:02Z | https://github.com/MonetDB/MonetDB/issues/2448 | 753,366,103 | 2,448 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-11-19 11:44:59 +0100
From: mharrison <<matt_harrison>>
To: @njnes
Version: -- development
Last updated: 2009-04-28 02:40:10 +0200
## Comment 13588
Date: 2007-11-19 23:44:59 +0100
From: mharrison <<matt_harrison>>
After inserting a primary key that is autoincremented, if you insert without one it will fail the first time, but later succeed.
Here's the test case....
CREATE TABLE testusers (
id INT NOT NULL AUTO_INCREMENT,
name VARCHAR(40),
fullname VARCHAR(100),
PRIMARY KEY (id)
);
INSERT INTO testusers (id, name, fullname) VALUES (1, 'wendy', 'Wendy Wones');
-- This will fail the first time
INSERT INTO testusers (name, fullname) VALUES ('fred', 'Fred Flintstone');
-- Now it will succeed
INSERT INTO testusers (name, fullname) VALUES ('fred', 'Fred Flintstone');
## Comment 13589
Date: 2007-11-20 07:50:52 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: NO
error confirmed and test file added.
analysis pending.
## Comment 13590
Date: 2007-11-20 12:51:51 +0100
From: @romulogoncalves
Logged In: YES
user_id=1498628
Originator: NO
Well, after some debug I realized that this bug might not be a bug (maybe we will need a new error message or warning).
If you use auto-increment to set your pkey and then you also insert the values for the pkey by hand the auto-increment (in our case a sequence) is not aware of that insertion.
To be more specific I will give an example.
In the example used to open this bug, the first insertion uses id = 1. However, our sequence was not selected so its value remains 1. The next insertion is without the id value, in this case the sequence is called to give a value for the id. It will fail because the pkey with value 1 already exists, but it will increment the sequence to 2. The next insertion works because it does not break any constraint (there is no pkey with value 2).
Now imagine that the user do the following insertion:
INSERT INTO testusers (id, name, fullname) VALUES (3, 'wendy', 'WendyWones');
In this case only the third insertion without the value for id column will give a constraint error (pkey violated).
To conclude, the auto-increment is not aware of the values already used by the user. One solution is to improve the auto-increment, the other one is to not allow the user defined the value for the column that was set auto-increment.
Now is the SQL god father who has to give an opinion...
Regards,
Romulo
## Comment 13591
Date: 2007-11-20 14:38:06 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
As romulo put it, was/is sort of a feature that we do not update a sequence number when inserts supply the value (normally derived from the sequence number). Reason for this is that sequence numbers are sort of independed from the column value. They are incremented etc by functions. Our auto_increment is therefor simply implemented by a default value(function) applied on insert. So currently we you insert a value your self you should call, ALTER SEQUENCE "sequence_number" RESTART WITH next_value.
As auto_increment, isn't part of the sql specification, there is no clear semantics here. So I'l move this for now into a feature request.
## Comment 13592
Date: 2007-11-22 06:29:48 +0100
From: mharrison <<matt_harrison>>
Logged In: YES
user_id=1164236
Originator: YES
I didn't realize the this only fails when the first specified PK is one more than the number in the table. So this shouldn't happen often. I thought (without testing) that after specifying the PK, the next insert would always fail. This isn't the case. Feel free to close or do whatever you guys think is best. Thanks for the explanation Romulo
## Comment 13593
Date: 2009-04-28 14:40:08 +0200
From: @grobian
this looks like correct behaviour to me, the bug as invalid, and no feature to request :)
## Comment 13594
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1834820 at http://sourceforge.net/support/tracker.php?aid=1834820
| SQL:AUTO_INCREMENT failures... | https://api.github.com/repos/MonetDB/MonetDB/issues/2447/comments | 0 | 2020-11-30T10:44:22Z | 2024-06-28T07:23:02Z | https://github.com/MonetDB/MonetDB/issues/2447 | 753,366,054 | 2,447 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-11-05 10:30:59 +0100
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2007-11-24 03:40:00 +0100
## Comment 13585
Date: 2007-11-05 10:30:59 +0100
From: @grobian
start-stop-daemon cannot check the exit status of a process that doesn't detach itself when using its -b option, as it has to fork itself. This is not desirable, as this means it cannot be detected when merovingian could not start. For this reason it would be nice if merovingian had a way to start into the background, either via an option flag, configuration variable, or as default behaviour.
## Comment 13586
Date: 2007-11-24 15:40:00 +0100
From: @grobian
Logged In: YES
user_id=963970
Originator: YES
implemented in CVS HEAD.
## Comment 13587
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1825922 at http://sourceforge.net/support/tracker.php?aid=1825922
| merovingian: allow for option to detach (run in background) | https://api.github.com/repos/MonetDB/MonetDB/issues/2446/comments | 0 | 2020-11-30T10:44:18Z | 2024-06-28T07:23:02Z | https://github.com/MonetDB/MonetDB/issues/2446 | 753,366,018 | 2,446 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-09-17 11:59:18 +0200
From: @mlkersten
To: @sjoerdmullender
Version: -- development
Last updated: 2009-06-19 02:00:26 +0200
## Comment 13580
Date: 2007-09-17 23:59:18 +0200
From: @mlkersten
I would appreciate an installer for Windows Vista
## Comment 13581
Date: 2008-10-15 15:18:21 +0200
From: @drstmane
Don't our Windows installers (now) also work on/for Vista?
## Comment 13582
Date: 2008-10-15 17:11:36 +0200
From: @mlkersten
The Windows installer for ODBC breaks: MonetDB-ODBC-Installer-i386-20080812.msi .
".. A program required for this install to complete could not be run.
Contact your support personel or package vendor."
## Comment 13583
Date: 2009-06-19 14:00:26 +0200
From: @grobian
according to our download page we do Vista installers.
## Comment 13584
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1796693 at http://sourceforge.net/support/tracker.php?aid=1796693
| SQL: Vista Installer | https://api.github.com/repos/MonetDB/MonetDB/issues/2445/comments | 0 | 2020-11-30T10:44:15Z | 2024-06-28T07:23:02Z | https://github.com/MonetDB/MonetDB/issues/2445 | 753,365,979 | 2,445 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-08-14 10:00:40 +0200
From: Colin Foss <<cgfoss>>
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2009-12-04 11:49:31 +0100
## Comment 13577
Date: 2007-08-14 22:00:40 +0200
From: Colin Foss <<cgfoss>>
to_char(int, text) where text is a mask used to format the number conversion to character.
e.g. to_char(5, '999') would return ' 5',
and to_char(5, '009') would return '005'.
## Comment 13578
Date: 2009-12-04 11:49:31 +0100
From: @mlkersten
This is an instance of a user-defined function. We leave it to users to prepare cross-system (Postgresql,Mysql,,...) convenience libraries.
## Comment 13579
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1774294 at http://sourceforge.net/support/tracker.php?aid=1774294
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| SQL: formatting functions | https://api.github.com/repos/MonetDB/MonetDB/issues/2444/comments | 0 | 2020-11-30T10:44:12Z | 2024-06-28T07:23:01Z | https://github.com/MonetDB/MonetDB/issues/2444 | 753,365,937 | 2,444 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-08-07 07:24:45 +0200
From: Colin Foss <<cgfoss>>
To: @njnes
Version: -- development
Last updated: 2008-01-02 07:40:01 +0100
## Comment 13573
Date: 2007-08-07 19:24:45 +0200
From: Colin Foss <<cgfoss>>
Support cast(coalesce(aaa,bbb)) within where clause of PSM.
e.g.
AND cast(coalesce(et.time, '00:00') as varchar(13)) == 'something'
Currently an error is thrown:
ERROR = !CASE: currently not supported here
## Comment 13574
Date: 2008-01-02 18:38:15 +0100
From: @romulogoncalves
Logged In: YES
user_id=1498628
Originator: NO
This feature is already implemented (in the last check ins) and its basic functionality is covered by the test:
sql/src/test/coalesce.sql
Regards,
Romulo
Note: My user rights do not allow me to close this feature request, so I ask to someone else to do it.
## Comment 13575
Date: 2008-01-02 19:40:01 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: NO
closed upon request
## Comment 13576
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1769551 at http://sourceforge.net/support/tracker.php?aid=1769551
| support coalesce within where clause | https://api.github.com/repos/MonetDB/MonetDB/issues/2443/comments | 0 | 2020-11-30T10:44:09Z | 2024-06-28T07:23:01Z | https://github.com/MonetDB/MonetDB/issues/2443 | 753,365,897 | 2,443 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-08-02 09:36:44 +0200
From: @peterboncz
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2011-09-07 11:20:24 +0200
## Comment 13567
Date: 2007-08-02 21:36:44 +0200
From: @peterboncz
CData tags
Cdata tags disappear when I call back a document for example I store in the database :
pf:add-doc(“c://test.xmlâ€, “testâ€)
<new>
<hello><![CDATA[ myData]]></hello>
</new>
I get the document back as:
doc(“testâ€)
<new>
<hello>myData</hello>
</new>
(submitted on behalf of Loic.SAINT-GHISLAIN@mines.inpl-nancy.fr)
## Comment 13568
Date: 2007-08-02 21:49:13 +0200
From: @peterboncz
Logged In: YES
user_id=591107
Originator: YES
Our shredder just treats CDATA as a notation for characters. For serialization, any special chars (a reason for using CDATA) are serialized with proper escapes (..well, hopefully)
Are we supposed to conserve CDATA nodes?
In the XQery and xPath specs I could not find an answer to this. The XML Infoset spec even specifically states that this issue is out of its scope.
I found some remark on a Michael Rhys weblog about serialization standards, however, that seems to imply that CDATA section are to be treated as separate text nodes, and serialized again in CDATA form. Note that this implies that data(<a> <!CDATA [ spaces-around ]> </a>) produces double spacing around "spaces-around"..
I think this could be implemented , probably without too much effort (introducing a CDATA node that is treated identical to text node, expcept in serialization), but should we?
relaying this to you, Teggy, as the departure of Jens for pratical purposes has made you our stand conformance expert.
## Comment 13569
Date: 2007-08-03 15:13:36 +0200
From: @arjenpdevries
Logged In: YES
user_id=572978
Originator: NO
Notice that I raised this issue before, and then Jens commented it was expected behaviour:
http://sourceforge.net/mailarchive/message.php?msg_name=f7253dfd0701021516j10e31661m20f15764c023742d%40mail.gmail.com
Also, from my recent posts on monetdb-developers (JanR, I am trying your patch right now BTW):
Disclaimer: I am not very sure what should be correct.
http://www.w3.org/TR/2003/WD-xslt-xquery-serialization-20030502/
states
Note:
This is phrased to permit an implementor to provide an option that
attempts to preserve CDATA sections present in the source document.
I am a proponent of an attempt to preserve CDATA as it was inserted,
so this deviates from Jens' earlier reply, simply because now another
person ran in the same problem as I did before.
Cheers,
Arjen
## Comment 13570
Date: 2007-08-03 17:20:40 +0200
From: Jan Rittinger <<jan.rittinger>>
Logged In: YES
user_id=993208
Originator: NO
I had a look at the current version of the serialization spec (not the one from 2003). It comes with a serialization parameter cdata-section-elements that is bound to a list of qnames (http://www.w3.org/TR/xslt-xquery-serialization/XML_CDATA-SECTION-ELEMENTS). The list of qnames is then used to serialize all text nodes whose parents tagname matches one of the qnames as CDATA.
This means that CDATA nodes as Jens says are normal text nodes -- there is no distinction between them -- and only a special serialization parameter allows us to change the serialization format of some textnodes. (It depends on the list of qnames one uses for serialization and not on the input -- there is no *CDATA node*.)
My conclusion is that this bug report is no valid bug report but merely a feature request for more serialization parameters.
Jan
## Comment 13571
Date: 2007-11-10 23:20:00 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: NO
Let's move it to the feature request list.
## Comment 13572
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1766415 at http://sourceforge.net/support/tracker.php?aid=1766415
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 16189
Date: 2011-09-07 11:20:24 +0200
From: @sjoerdmullender
Development on MonetDB4 and MonetDB/XQuery has ceased. It is unlikely that
this bug will be fixed.
The last release of MonetDB/XQuery and MonetDB4 is Mar2011.
| XQ:CDATA segments are not (properly??) conserved | https://api.github.com/repos/MonetDB/MonetDB/issues/2442/comments | 0 | 2020-11-30T10:44:07Z | 2024-06-28T07:23:00Z | https://github.com/MonetDB/MonetDB/issues/2442 | 753,365,867 | 2,442 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-07-12 03:29:05 +0200
From: Colin Foss <<cgfoss>>
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2007-07-12 03:34:02 +0200
## Comment 13564
Date: 2007-07-12 15:29:05 +0200
From: Colin Foss <<cgfoss>>
Add support for data type "timestamp with time zone".
Allow MonetDB to understand timezone rules, GMT offsets, daylight saving time shifts.
## Comment 13565
Date: 2007-07-12 15:34:02 +0200
From: @grobian
Logged In: YES
user_id=963970
Originator: NO
This already exists, and all the tricky things of time zones are tested and implemented. See also the JDBC timezone/timestamp tests.
>start transaction;
&4 f
>create table z (i timestamp with time zone);
&3
>select * from z;
&1 0 0 1 0
% sys.z table_name
% i name
% timestamptz type
% 26 length
>insert into z values (now());
&2 1
>select * from z;
&1 1 1 1 1
% sys.z table_name
% i name
% timestamptz type
% 26 length
[ 2007-07-12 13:32:05.000000+00:00 ]
>rollback;
&4 t
>
## Comment 13566
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1752635 at http://sourceforge.net/support/tracker.php?aid=1752635
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| SQL: timestamp with time zone | https://api.github.com/repos/MonetDB/MonetDB/issues/2441/comments | 0 | 2020-11-30T10:44:03Z | 2024-06-28T07:23:00Z | https://github.com/MonetDB/MonetDB/issues/2441 | 753,365,830 | 2,441 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-07-12 03:17:19 +0200
From: Colin Foss <<cgfoss>>
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2009-12-01 11:13:49 +0100
## Comment 13561
Date: 2007-07-12 15:17:19 +0200
From: Colin Foss <<cgfoss>>
Add EPOCH to the list of valid options for EXTRACT ( datetime_field FROM scalar_exp ).
Current list is
datetime_field:
YEAR | MONTH | DAY | HOUR | MINUTE | SECOND
EPOCH would be defined as the number of seconds since 1970/01/01 00:00:00.000 GMT.
## Comment 13562
Date: 2009-12-01 11:13:49 +0100
From: @grobian
this is a duplicate of 2872897
## Comment 13563
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1752627 at http://sourceforge.net/support/tracker.php?aid=1752627
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| SQL: date/time extract epoch | https://api.github.com/repos/MonetDB/MonetDB/issues/2440/comments | 0 | 2020-11-30T10:44:00Z | 2024-06-28T07:22:59Z | https://github.com/MonetDB/MonetDB/issues/2440 | 753,365,795 | 2,440 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-07-05 02:15:30 +0200
From: @njnes
To: @sjoerdmullender
Version: -- development
Last updated: 2010-06-21 19:52:44 +0200
## Comment 13558
Date: 2007-07-05 14:15:30 +0200
From: @njnes
Currently msqldump uses mapi_slice_row for each data
row its outputing. Mapi_slice_row does a malloc/copy for
each data field. This afcourse is far from optimal.
A possible solution would be to let the sql server dump directly in a format which can be used for loading. Then msqldump should only pass the data directly from the server stream into the output stream.
This probably requires an new set output feature too sql.
## Comment 13559
Date: 2007-07-08 12:29:04 +0200
From: bendeguz <<bendeguz>>
Logged In: YES
user_id=1140413
Originator: NO
Perhaps the column (or column slice) at a time method would help here. Please consider the parallel dump/load operation too (eg.: with MAL) - multi-core processors are pretty common today.
## Comment 13560
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1748357 at http://sourceforge.net/support/tracker.php?aid=1748357
## Comment 14134
Date: 2010-06-21 19:52:36 +0200
From: @njnes
dumping maybe improved using copy into file.
| SQL: msqldump too slow, requires server side rendering | https://api.github.com/repos/MonetDB/MonetDB/issues/2439/comments | 0 | 2020-11-30T10:43:57Z | 2024-06-28T07:22:59Z | https://github.com/MonetDB/MonetDB/issues/2439 | 753,365,758 | 2,439 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-06-14 09:45:36 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2007-12-11 10:40:38 +0100
## Comment 13553
Date: 2007-06-14 21:45:36 +0200
From: @mlkersten
One of the active users reported unexpected performance problems on SQL queries involving IN predicates, e.g.
...
WHERE attr IN (1,2,3)
Analysis of the trace uncovered an area that deserves
attention. The semijoin produced by the SQL frontend
involved a large left BAT (13Mrows) and against a BAT
with the constants. The algorithm went to the default
kintersect.
A test-case to illustrate the inbalance in performance
is added under the name src/mal/Tests/venks.mal
Timing large operand is 1M, small 2 elements
type :bat[:oid,:sht]
kintersect(largeRev,smallRev) 51356
count 61
join(large,smallRev) 103289
count 61
kintersect(smallRev,largeRev) 33
count 3
join(small,largeRev) 37
count 3
## Comment 13554
Date: 2007-06-20 20:06:07 +0200
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
The respective test case fails due to missing stable output:
http://monetdb.cwi.nl/testing/projects/monetdb/Current/MonetDB5/.mTests103/GNU.64.64.d-Fedora6/src_mal/venks.out.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Current/MonetDB5/.mTests103/GNU.64.64.d-Fedora6/src_mal/venks.err.00.html
## Comment 13555
Date: 2007-11-10 23:33:13 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
files exist.. remainder is not a bug but a performance feature request.
## Comment 13556
Date: 2007-12-11 22:40:38 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
we now also handle this case efficiently. We use a direct hashing approach.
## Comment 13557
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1737438 at http://sourceforge.net/support/tracker.php?aid=1737438
| GDK: kintersect performance issue | https://api.github.com/repos/MonetDB/MonetDB/issues/2438/comments | 0 | 2020-11-30T10:43:53Z | 2024-06-28T13:42:43Z | https://github.com/MonetDB/MonetDB/issues/2438 | 753,365,719 | 2,438 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-06-13 08:37:50 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2009-12-01 11:31:14 +0100
## Comment 13549
Date: 2007-06-13 20:37:50 +0200
From: @mlkersten
Please provide cheap migration tools if the
SQL catalog structure changes. It may just be
too time consuming to rebuild a database
completely from scratch. In particular when it
can be avoided.
## Comment 13550
Date: 2009-04-28 14:41:07 +0200
From: @grobian
sounds similar to
https://sourceforge.net/tracker/?func=detail&atid=482471&aid=2036498&group_id=56
967
## Comment 13551
Date: 2009-12-01 11:31:14 +0100
From: @grobian
this is policy for M5/SQL
## Comment 13552
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1736650 at http://sourceforge.net/support/tracker.php?aid=1736650
| SQL: migration | https://api.github.com/repos/MonetDB/MonetDB/issues/2437/comments | 0 | 2020-11-30T10:43:50Z | 2024-06-28T13:42:42Z | https://github.com/MonetDB/MonetDB/issues/2437 | 753,365,688 | 2,437 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-06-11 11:03:06 +0200
From: bendeguz <<bendeguz>>
To: @mlkersten
Version: -- development
Last updated: 2012-05-01 10:51:28 +0200
## Comment 13541
Date: 2007-06-11 23:03:06 +0200
From: bendeguz <<bendeguz>>
Dear MonetDB developers,
Last week I have found an interesting paper about the MonetDB datacell module. I am new in the database topic but I have to said that the whole MonetDB concept is pretty amazing. I am trying to build a network monitoring application based on a data streaming, so this module would be a big learning opportunity to me. Any plan / eta for an initial cvs commit or an early preview release?
Regards,
B.
## Comment 13542
Date: 2007-06-13 22:08:23 +0200
From: bendeguz <<bendeguz>>
Logged In: YES
user_id=1140413
Originator: YES
These papers are triggered my interest: a workshop report from "Data always and everywhere: management of mobile, ubiquitous, pervasive and sensor data" and one of your lecture notes (Advanced Database Techniques).
I am trying to build a netflow (source ip:port, target ip:port, ..) monitoring and intrusion detection (with history support) application based on data streaming. The MonetDB/DataCell module looks the most promising candidate for this purpose, especially when I think about the zero degrees of freedom query execution. Thanks for your help.
## Comment 13543
Date: 2007-07-08 12:40:25 +0200
From: bendeguz <<bendeguz>>
Logged In: YES
user_id=1140413
Originator: YES
I didn't want to look implatient but any update with the Datacell module? Thanks. B.
## Comment 13544
Date: 2007-07-28 10:27:26 +0200
From: bendeguz <<bendeguz>>
Logged In: YES
user_id=1140413
Originator: YES
Unfortunately I haven't received any answer for long time so it seems I have to choose different streaming database :/
## Comment 13545
Date: 2007-07-28 12:40:51 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Originator: NO
This is probably mostly a question for Martin Kersten. I'm not sure he reads the feature requests very carefully, but anyway, he is currently on vacation.
I also think the mailing list monetdb-users@lists.sourceforge.net or monetdb-developers@lists.sourceforge.net is a better venue for this question.
## Comment 13546
Date: 2007-08-08 22:00:16 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Originator: NO
Dear B.
Yes, the DataCell facilities are not available for external use yet.
Holding your breath in this situation is not the best option.
Succes, Martin
## Comment 13547
Date: 2007-12-15 17:17:57 +0100
From: bendeguz <<bendeguz>>
Logged In: YES
user_id=1140413
Originator: YES
Dear Mr. Martin Kersten,
Thanks for your comment. I am still interested, so I am hoping that one day it will be released for external use too. I have noticed that the new MonetDB roadmap (Winter 2007) now contains the datacell I am looking forward to try it out.
Thanks for your help,
Bendeguz
## Comment 13548
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1735322 at http://sourceforge.net/support/tracker.php?aid=1735322
## Comment 17202
Date: 2012-05-01 10:51:28 +0200
From: @grobian
datacell is included in the sources for a while now
| MonetDB Datacell module | https://api.github.com/repos/MonetDB/MonetDB/issues/2436/comments | 0 | 2020-11-30T10:43:47Z | 2024-06-27T11:45:44Z | https://github.com/MonetDB/MonetDB/issues/2436 | 753,365,657 | 2,436 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-05-25 10:01:14 +0200
From: Colin Foss <<cgfoss>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2007-11-10 11:30:32 +0100
## Comment 13538
Date: 2007-05-25 22:01:14 +0200
From: Colin Foss <<cgfoss>>
To complete the loop
changes I'd like to see in mguardian script:
1) get the proper value for disk space free.
2) send stderr to log during mserver5 startup.
125c125
< full=`df $DBFARM/$1 |sed -e 1d -e "s/%.*//" -e "s/.* //"`
---
> full=`df $DBFARM/$1 |sed -e 1d -e "s/%.*//" -e "s/.* //" | tail -1`
195c195
< (cd $1; mserver5 --set daemon=yes --dbname=$1 $CONFIG $ARGS $STARTUP)&
---
> (cd $1; mserver5 --set daemon=yes --dbname=$1 $CONFIG $ARGS $STARTUP 2>&1 >> $DBFARM/$1/mserver5.log)&
## Comment 13539
Date: 2007-11-10 11:30:32 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: NO
the script is dropped in favor of the Merovingian
## Comment 13540
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1725813 at http://sourceforge.net/support/tracker.php?aid=1725813
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| mguardian script enhancements | https://api.github.com/repos/MonetDB/MonetDB/issues/2435/comments | 0 | 2020-11-30T10:43:44Z | 2024-06-28T13:42:42Z | https://github.com/MonetDB/MonetDB/issues/2435 | 753,365,614 | 2,435 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-05-22 09:24:08 +0200
From: Colin Foss <<cgfoss>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2007-11-10 11:31:01 +0100
## Comment 13535
Date: 2007-05-22 21:24:08 +0200
From: Colin Foss <<cgfoss>>
I would like to suggest changes to the mchkpnt script.
1) use monetdb.conf from dbfarm as default setting for the checkpoint directory.
2) use monetdb.conf from dbfarm as default setting for the database logs directory. look at sql_logdir or xquery_logdir
## Comment 13536
Date: 2007-11-10 11:31:00 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: NO
The role is taken over by Merovingian
## Comment 13537
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1723722 at http://sourceforge.net/support/tracker.php?aid=1723722
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| mchkpnt script enhancements | https://api.github.com/repos/MonetDB/MonetDB/issues/2434/comments | 0 | 2020-11-30T10:43:41Z | 2024-06-28T13:42:41Z | https://github.com/MonetDB/MonetDB/issues/2434 | 753,365,564 | 2,434 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-05-16 11:59:02 +0200
From: Colin Foss <<cgfoss>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2007-11-10 11:31:30 +0100
## Comment 13532
Date: 2007-05-16 23:59:02 +0200
From: Colin Foss <<cgfoss>>
1) I propose a new environment variable be used to indicate the location of the dbfarm. i.e. MONETDB5_DBFARM
2) Along with (1), I would like to see a change to the /var/lib/MonetDB/bin/monetdb script on line 45 to read:
DBFARM=${MONETDB5_DBFARM:-MONETDB5_PREFIX/var/MonetDB5/dbfarm}
which would take the new environment variable if defined or otherwise use the default.
3) on line 242 of the monetdb script, I would like to change the mguardian shutdown to match the startup.
i.e. $MONETDB5_PREFIX/bin/mguardian --stop --dbfarm=$DBFARM --dbname=$1
instead of $MONETDB5_PREFIX/bin/mguardian --stop $DBFARM/$1
which causes an error if you have the dbfarm located somewhere besides the default.
## Comment 13533
Date: 2007-11-10 11:31:30 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: NO
The script is taken over by Merovingian
## Comment 13534
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1720380 at http://sourceforge.net/support/tracker.php?aid=1720380
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| monetdb script enhancements | https://api.github.com/repos/MonetDB/MonetDB/issues/2433/comments | 0 | 2020-11-30T10:43:37Z | 2024-06-28T13:42:40Z | https://github.com/MonetDB/MonetDB/issues/2433 | 753,365,531 | 2,433 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-05-13 10:20:09 +0200
From: blck_hl <<blck_hl>>
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2009-12-01 11:29:42 +0100
## Comment 13527
Date: 2007-05-13 10:20:09 +0200
From: blck_hl <<blck_hl>>
For the binary distribution of monetdb/sql for Windows, please add the functionality to allow monetdb/sql to be used in Python as an embedded SQL server.
It seems that the only missing step is the inclusion of the monetdb python module. A more detailed description can be found here:
http://sourceforge.net/mailarchive/forum.php?thread_name=b78664310705120300y30ec9312y378d59ec1bd90c71%40mail.gmail.com&forum_name=monetdb-users
## Comment 13528
Date: 2007-05-16 11:03:51 +0200
From: @romulogoncalves
Logged In: YES
user_id=1498628
Originator: NO
I cannot open this link. I already tried several times and I get
500 Internal Server Error.
Send me another link or try to give here a more detailed description.
Regards,
Romulo
## Comment 13529
Date: 2007-05-18 01:14:02 +0200
From: blck_hl <<blck_hl>>
Logged In: YES
user_id=1791976
Originator: YES
Here is a shorter URL to the original discussion (in case you cannot open it due to it being too long):
http://tinyurl.com/364kfc
## Comment 13530
Date: 2009-12-01 11:29:41 +0100
From: @grobian
Since the old swig generatd python wrapper is no longer supported, this request has become impossible. We only have a native python connector now, so you can't have something external embedded in there.
## Comment 13531
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1718038 at http://sourceforge.net/support/tracker.php?aid=1718038
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Allow Embedded SQL Server to be used in Python on Windows | https://api.github.com/repos/MonetDB/MonetDB/issues/2432/comments | 0 | 2020-11-30T10:43:33Z | 2024-06-28T13:42:40Z | https://github.com/MonetDB/MonetDB/issues/2432 | 753,365,475 | 2,432 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-04-02 04:58:30 +0200
From: Colin Foss <<cgfoss>>
To: clients devs <<bugs-clients>>
Version: -- development
CC: @yzchang
Last updated: 2016-04-11 11:38:56 +0200
## Comment 13525
Date: 2007-04-02 16:58:30 +0200
From: Colin Foss <<cgfoss>>
It would be very useful to add a Test button to the MonetDB ODBC configuration panel.
The test button would establish a connection to the defined database. Similar in function to the Postgres ODBC driver panel.
The test button would be independent of any 3rd party applications that may or may not implement ODBC uniformly.
## Comment 13526
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1693030 at http://sourceforge.net/support/tracker.php?aid=1693030
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-clients@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 18146
Date: 2012-11-27 16:24:48 +0100
From: @yzchang
No test needed for feature request
| ODBC driver setup - test connection | https://api.github.com/repos/MonetDB/MonetDB/issues/2431/comments | 0 | 2020-11-30T10:43:30Z | 2024-06-28T13:42:39Z | https://github.com/MonetDB/MonetDB/issues/2431 | 753,365,430 | 2,431 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-02-05 11:03:43 +0100
From: Wouter Alink <<alink>>
To: clients devs <<bugs-clients>>
Version: -- development
CC: alvarobastolima, martin.van.dinther, @yzchang
Last updated: 2017-09-28 19:28:14 +0200
## Comment 13523
Date: 2007-02-05 11:03:43 +0100
From: Wouter Alink <<alink>>
it would be (very) nice to have some support for streaming of results in the JDBC interface. the stream functions such as 'getUnicodeStream()' are currently not implemented.
(this is just a reminder) i heard it was already on the wishlist... something with the implementation of references and id's.
## Comment 13524
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1652264 at http://sourceforge.net/support/tracker.php?aid=1652264
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-clients@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 18145
Date: 2012-11-27 16:24:04 +0100
From: @yzchang
No test needed for feature request
## Comment 18152
Date: 2012-11-27 16:36:33 +0100
From: @grobian
*** Bug #3147 has been marked as a duplicate of this bug. ***
## Comment 25676
Date: 2017-09-28 19:28:14 +0200
From: Martin van Dinther <<martin.van.dinther>>
Method getCharacterStream is implemented and supported in ResultSet and Clob classes.
see https://dev.monetdb.org/hg/monetdb-java/rev/1e49fc74dba4
and https://dev.monetdb.org/hg/monetdb-java/rev/352b54ff0436
Methods getAsciiStream and getUnicodeStream are not implemented. They are superseded by getCharacterStream. You should use that method instead.
Note: getUnicodeStream is also marked as deprecated in JDBC 4.1
Method getBinaryStream is implemented and supported in ResultSet and Blob classes.
Method setCharacterStream is implemented and supported in PreparedStatement.
Method setBinaryStream is not (yet) implemented in PreparedStatement.
Methods setCharacterStream and setBinaryStream are not (yet) implemented in Clob and Blob classes.
For latest list of not implemented/supported methods see:
https://dev.monetdb.org/hg/monetdb-java/file/tip/release.txt
| JDBC: support for streams | https://api.github.com/repos/MonetDB/MonetDB/issues/2430/comments | 0 | 2020-11-30T10:43:26Z | 2024-06-27T11:45:39Z | https://github.com/MonetDB/MonetDB/issues/2430 | 753,365,394 | 2,430 |
[
"MonetDB",
"MonetDB"
] | Date: 2007-01-21 01:26:30 +0100
From: Wouter Alink <<alink>>
To: XQuery devs <<bugs-xquery>>
Version: -- development
Last updated: 2011-04-28 16:38:54 +0200
## Comment 13520
Date: 2007-01-21 13:26:30 +0100
From: Wouter Alink <<alink>>
i don't know whether this feature is already implemented, i might have overlooked it, but i was wondering whether it would be possible to 'put' a node into the database as a new document and afterwards be able to update this new document. the only 'put()' that i could find creates read-only documents.
## Comment 13521
Date: 2009-12-02 14:31:06 +0100
From: @yzchang
No, such feature does not exist, and, IMHO, it is not possible with the standart built-in function fn:put.
Why is shredding an fn:put document with pf:add-doc as an updatable document not sufficient?
## Comment 13522
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1640744 at http://sourceforge.net/support/tracker.php?aid=1640744
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-xquery@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 15759
Date: 2011-04-28 16:38:54 +0200
From: @sjoerdmullender
Development on MonetDB4 and MonetDB/XQuery has ceased. It is unlikely that this bug will be fixed.
The last release of MonetDB/XQuery and MonetDB4 is Mar2011.
| XQ: put() with holes | https://api.github.com/repos/MonetDB/MonetDB/issues/2429/comments | 0 | 2020-11-30T10:43:24Z | 2024-06-28T13:42:38Z | https://github.com/MonetDB/MonetDB/issues/2429 | 753,365,348 | 2,429 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-12-21 02:04:01 +0100
From: Jan Rittinger <<jan.rittinger>>
To: XQuery devs <<bugs-xquery>>
Version: -- development
Last updated: 2011-04-28 16:38:59 +0200
## Comment 13518
Date: 2006-12-21 14:04:01 +0100
From: Jan Rittinger <<jan.rittinger>>
Until now I still wasn't aware that some time ago the Boehm garbage collection library was removed from the pf sources.
I'd like to have it back :) Perhaps a new version will work with MonetDB.
The reason for my request is that most queries that are bigger than 2 pages of XQuery run out of memory during algebraic optimization. (Every rewrite introduces new nodes, new properties, etc.)
An alternative I could live (at least for the time being) is to extend the pf_allocator struct to store also a marker. Each PFmalloc invocation could then have an additional optional marker argument and a corresponding call of PFfree(marker) would free all memory that was marked during malloc with this marker value.
I however prefer the first variant (adding a real garbage collection library) as this is certainly less error-prone.
## Comment 13519
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1620113 at http://sourceforge.net/support/tracker.php?aid=1620113
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-xquery@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 15773
Date: 2011-04-28 16:38:59 +0200
From: @sjoerdmullender
Development on MonetDB4 and MonetDB/XQuery has ceased. It is unlikely that this bug will be fixed.
The last release of MonetDB/XQuery and MonetDB4 is Mar2011.
| XQ: Garbage Collector | https://api.github.com/repos/MonetDB/MonetDB/issues/2428/comments | 0 | 2020-11-30T10:43:20Z | 2024-06-28T13:42:38Z | https://github.com/MonetDB/MonetDB/issues/2428 | 753,365,302 | 2,428 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-11-23 01:18:36 +0100
From: @romulogoncalves
To: @njnes
Version: -- development
Last updated: 2007-07-31 12:32:52 +0200
## Comment 13515
Date: 2006-11-23 01:18:36 +0100
From: @romulogoncalves
Dear developers,
I updates my M4, M5, and SQL. I run the Mtest for M5-SQL and the test benesamo (ADT2006) is red. After 30 minutes around the problem I found the reason.
If you run the test like it is in the cvs head the last select query gives a different output thant the stable test.
[ "Dr.", "Thomas", NULL, "Bond" ]
instead of
[ "Dr.", "George", "Baxter", "Phillips" ]
However, if you run the same test without the start transaction (in the top of the file) and the commit (after all the data insertions) the test will not fail.
To run the test I used a normal mclient and a M5.
It is the most interesting error that I found in M5 ;)
Regards,
Romulo
## Comment 13516
Date: 2007-07-31 00:32:51 +0200
From: @romulogoncalves
Logged In: YES
user_id=1498628
Originator: YES
I just did the experiment and it works fine so I assume that this bug can be closed.
Regards,
Romulo
## Comment 13517
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1601475 at http://sourceforge.net/support/tracker.php?aid=1601475
| SQL transactions | https://api.github.com/repos/MonetDB/MonetDB/issues/2427/comments | 0 | 2020-11-30T10:43:16Z | 2024-06-28T13:42:37Z | https://github.com/MonetDB/MonetDB/issues/2427 | 753,365,248 | 2,427 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-11-03 12:16:10 +0100
From: @grobian
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2006-11-05 10:03:53 +0100
## Comment 13508
Date: 2006-11-03 12:16:10 +0100
From: @grobian
While trying to make a daily build ebuild, I
encountered that the daily builds are:
1) neither archived
2) nor tagged with the date in the name they were generated
The first is a small problem for ebuilds (distfiles
mirrors might not have taken up the distfile), 2) is a
bigger problem because I need to be able to distinguish
the sources.
If possible, I would like to see the source tarbals to
be kept for a while, and to have the date in their
names. This allows to take a snapshot of "the day
before" if that one doesn't have a feature/bug
included. Additionally it doesn't screw up my digests.
## Comment 13509
Date: 2006-11-03 12:20:19 +0100
From: @grobian
Logged In: YES
user_id=963970
I suggest naming like this:
MonetDB-4.13.1-20061103.tar.gz
MonetDB-SQL-2.13.1-20061103.tar.gz
etc.
We never use hyphens in versions, don't we? Otherwise maybe
use an underscore.
## Comment 13510
Date: 2006-11-04 11:34:35 +0100
From: @drstmane
Logged In: YES
user_id=572415
(2) is done;
cf.
http://monetdb.cwi.nl/testing/projects/monetdb/Current/.DailyBuilds./SourceTarBalls/
http://monetdb.cwi.nl/testing/projects/monetdb/Current/.DailyBuilds./SourceRPMs/
http://monetdb.cwi.nl/testing/projects/monetdb/Current/.DailyBuilds./RPMs/Fedora4.x86_64.oid64/
for (1), I have to check disk space availability and find a
"convenient" (both in terms of disk space and usability)
archiving period/limit --- would a week be enough?
## Comment 13511
Date: 2006-11-05 19:51:17 +0100
From: @drstmane
Logged In: YES
user_id=572415
DailyBuilds are no kept for a week, i.e., per weekday, cf.
http://monetdb.cwi.nl/testing/projects/monetdb/Current/.DailyBuilds.Sunday./
## Comment 13512
Date: 2006-11-05 22:03:53 +0100
From: @grobian
Logged In: YES
user_id=963970
Isn't there somewhere a few gig left to just put an archive
online? The weekday thing is not really handy for me (us),
and a week also isn't that much of an improvement over a
single day from the perspective of what I want to use it for.
Thanks for all changes so far. I guess the archive thing is
low priority now, as I don't think I'm going to use daily
builds (any more).
## Comment 13513
Date: 2007-01-05 20:19:59 +0100
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
next step in archiving daily builds:
as of Jan 01 2007 (20070101), DailyBuilds are kept for a month
(more might follow, given we have/find enough storage space)
in the following structure, allowing simple ("canonical") references to
a particular, the latest, or all DailyBuilds:
(cf.
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/.DailyBuilds./
http://monetdb.cwi.nl/testing/projects/monetdb/Current/.DailyBuilds./
)
.DailyBuilds./
.DailyBuilds./0_latest/ -> 20070105/
.DailyBuilds./1____all/
.DailyBuilds./1____all/RPMs/
.DailyBuilds./1____all/RPMs/CentOS4.4.i686.oid32/
.DailyBuilds./1____all/RPMs/CentOS4.4.i686.oid32/MonetDB-client-4.14.0-1_20070104_CentOS4.4.oid32.i686.rpm -> ../../../20070104/RPMs/CentOS4.4.i686.oid32/MonetDB-client-4.14.0-1_20070104_CentOS4.4.oid32.i686.rpm
.DailyBuilds./1____all/RPMs/CentOS4.4.i686.oid32/MonetDB-client-4.14.0-1_20070105_CentOS4.4.oid32.i686.rpm -> ../../../20070105/RPMs/CentOS4.4.i686.oid32/MonetDB-client-4.14.0-1_20070105_CentOS4.4.oid32.i686.rpm
.DailyBuilds./1____all/RPMs/CentOS4.4.i686.oid32/MonetDB-devel-4.14.0-1_20070104_CentOS4.4.oid32.i686.rpm -> ../../../20070104/RPMs/CentOS4.4.i686.oid32/MonetDB-devel-4.14.0-1_20070104_CentOS4.4.oid32.i686.rpm
.DailyBuilds./1____all/RPMs/CentOS4.4.i686.oid32/MonetDB-devel-4.14.0-1_20070105_CentOS4.4.oid32.i686.rpm -> ../../../20070105/RPMs/CentOS4.4.i686.oid32/MonetDB-devel-4.14.0-1_20070105_CentOS4.4.oid32.i686.rpm
[...]
.DailyBuilds./1____all/RPMs/SuSE9.3.x86_64.oid64/
[...]
.DailyBuilds./1____all/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-0.14.0-1_20070104_SuSE9.3.oid64.x86_64.rpm -> ../../../20070104/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-0.14.0-1_20070104_SuSE9.3.oid64.x86_64.rpm
.DailyBuilds./1____all/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-0.14.0-1_20070105_SuSE9.3.oid64.x86_64.rpm -> ../../../20070105/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-0.14.0-1_20070105_SuSE9.3.oid64.x86_64.rpm
.DailyBuilds./1____all/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-devel-0.14.0-1_20070104_SuSE9.3.oid64.x86_64.rpm -> ../../../20070104/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-devel-0.14.0-1_20070104_SuSE9.3.oid64.x86_64.rpm
.DailyBuilds./1____all/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-devel-0.14.0-1_20070105_SuSE9.3.oid64.x86_64.rpm -> ../../../20070105/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-devel-0.14.0-1_20070105_SuSE9.3.oid64.x86_64.rpm
.DailyBuilds./1____all/SourceRPMs/
.DailyBuilds./1____all/SourceRPMs/MonetDB-4.14.0-1_20070104.src.rpm -> ../../20070104/SourceRPMs/MonetDB-4.14.0-1_20070104.src.rpm
.DailyBuilds./1____all/SourceRPMs/MonetDB-4.14.0-1_20070105.src.rpm -> ../../20070105/SourceRPMs/MonetDB-4.14.0-1_20070105.src.rpm
.DailyBuilds./1____all/SourceRPMs/MonetDB-SQL-2.14.0-1_20070104.src.rpm -> ../../20070104/SourceRPMs/MonetDB-SQL-2.14.0-1_20070104.src.rpm
.DailyBuilds./1____all/SourceRPMs/MonetDB-SQL-2.14.0-1_20070105.src.rpm -> ../../20070105/SourceRPMs/MonetDB-SQL-2.14.0-1_20070105.src.rpm
.DailyBuilds./1____all/SourceRPMs/MonetDB-XQuery-0.14.0-1_20070104.src.rpm -> ../../20070104/SourceRPMs/MonetDB-XQuery-0.14.0-1_20070104.src.rpm
.DailyBuilds./1____all/SourceRPMs/MonetDB-XQuery-0.14.0-1_20070105.src.rpm -> ../../20070105/SourceRPMs/MonetDB-XQuery-0.14.0-1_20070105.src.rpm
.DailyBuilds./1____all/SourceTarBalls/
.DailyBuilds./1____all/SourceTarBalls/MonetDB-4.14.0-20070104.tar.gz -> ../../20070104/SourceTarBalls/MonetDB-4.14.0-20070104.tar.gz
.DailyBuilds./1____all/SourceTarBalls/MonetDB-4.14.0-20070105.tar.gz -> ../../20070105/SourceTarBalls/MonetDB-4.14.0-20070105.tar.gz
.DailyBuilds./1____all/SourceTarBalls/MonetDB-SQL-2.14.0-20070104.tar.gz -> ../../20070104/SourceTarBalls/MonetDB-SQL-2.14.0-20070104.tar.gz
.DailyBuilds./1____all/SourceTarBalls/MonetDB-SQL-2.14.0-20070105.tar.gz -> ../../20070105/SourceTarBalls/MonetDB-SQL-2.14.0-20070105.tar.gz
.DailyBuilds./1____all/SourceTarBalls/MonetDB-XQuery-0.14.0-20070104.tar.gz -> ../../20070104/SourceTarBalls/MonetDB-XQuery-0.14.0-20070104.tar.gz
.DailyBuilds./1____all/SourceTarBalls/MonetDB-XQuery-0.14.0-20070105.tar.gz -> ../../20070105/SourceTarBalls/MonetDB-XQuery-0.14.0-20070105.tar.gz
.DailyBuilds./20070104/
[...]
.DailyBuilds./20070105/
.DailyBuilds./20070105/RPMs/
.DailyBuilds./20070105/RPMs/CentOS4.4.i686.oid32/
.DailyBuilds./20070105/RPMs/CentOS4.4.i686.oid32/MonetDB-client-4.14.0-1_20070105_CentOS4.4.oid32.i686.rpm
.DailyBuilds./20070105/RPMs/CentOS4.4.i686.oid32/MonetDB-client-4.14.0-1_CentOS4.4.oid32.i686.rpm -> MonetDB-client-4.14.0-1_20070105_CentOS4.4.oid32.i686.rpm
.DailyBuilds./20070105/RPMs/CentOS4.4.i686.oid32/MonetDB-devel-4.14.0-1_20070105_CentOS4.4.oid32.i686.rpm
.DailyBuilds./20070105/RPMs/CentOS4.4.i686.oid32/MonetDB-devel-4.14.0-1_CentOS4.4.oid32.i686.rpm -> MonetDB-devel-4.14.0-1_20070105_CentOS4.4.oid32.i686.rpm
[...]
.DailyBuilds./20070105/RPMs/SuSE9.3.x86_64.oid64/
[...]
.DailyBuilds./20070105/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-0.14.0-1_20070105_SuSE9.3.oid64.x86_64.rpm
.DailyBuilds./20070105/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-0.14.0-1_SuSE9.3.oid64.x86_64.rpm -> MonetDB-XQuery-0.14.0-1_20070105_SuSE9.3.oid64.x86_64.rpm
.DailyBuilds./20070105/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-devel-0.14.0-1_20070105_SuSE9.3.oid64.x86_64.rpm
.DailyBuilds./20070105/RPMs/SuSE9.3.x86_64.oid64/MonetDB-XQuery-devel-0.14.0-1_SuSE9.3.oid64.x86_64.rpm -> MonetDB-XQuery-devel-0.14.0-1_20070105_SuSE9.3.oid64.x86_64.rpm
.DailyBuilds./20070105/SourceRPMs/
.DailyBuilds./20070105/SourceRPMs/MonetDB-4.14.0-1_20070105.src.rpm
.DailyBuilds./20070105/SourceRPMs/MonetDB-4.14.0-1.src.rpm -> MonetDB-4.14.0-1_20070105.src.rpm
.DailyBuilds./20070105/SourceRPMs/MonetDB-SQL-2.14.0-1_20070105.src.rpm
.DailyBuilds./20070105/SourceRPMs/MonetDB-SQL-2.14.0-1.src.rpm -> MonetDB-SQL-2.14.0-1_20070105.src.rpm
.DailyBuilds./20070105/SourceRPMs/MonetDB-XQuery-0.14.0-1_20070105.src.rpm
.DailyBuilds./20070105/SourceRPMs/MonetDB-XQuery-0.14.0-1.src.rpm -> MonetDB-XQuery-0.14.0-1_20070105.src.rpm
.DailyBuilds./20070105/SourceTarBalls/
.DailyBuilds./20070105/SourceTarBalls/MonetDB-4.14.0-20070105.tar.gz
.DailyBuilds./20070105/SourceTarBalls/MonetDB-4.14.0.tar.gz -> MonetDB-4.14.0-20070105.tar.gz
.DailyBuilds./20070105/SourceTarBalls/MonetDB-SQL-2.14.0-20070105.tar.gz
.DailyBuilds./20070105/SourceTarBalls/MonetDB-SQL-2.14.0.tar.gz -> MonetDB-SQL-2.14.0-20070105.tar.gz
.DailyBuilds./20070105/SourceTarBalls/MonetDB-XQuery-0.14.0-20070105.tar.gz
.DailyBuilds./20070105/SourceTarBalls/MonetDB-XQuery-0.14.0.tar.gz
## Comment 13514
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1589844 at http://sourceforge.net/support/tracker.php?aid=1589844
| Daily build sources don't carry a date in the name | https://api.github.com/repos/MonetDB/MonetDB/issues/2426/comments | 0 | 2020-11-30T10:43:13Z | 2024-06-28T13:42:36Z | https://github.com/MonetDB/MonetDB/issues/2426 | 753,365,207 | 2,426 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-09-22 03:31:43 +0200
From: @drstmane
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2009-06-05 02:05:44 +0200
## Comment 13503
Date: 2006-09-22 15:31:43 +0200
From: @drstmane
In case testing detects that any tests (incl.
compilation) (of the "public" SF-hosted code) ran
"worse" than the with the previous test run (usually
the previous day/night), the TestTools should (in
addition to sending test result mails) also
automatically file a bug report with SF.
This bug report should list all tests that "ran worse
than before", provide pointers to the entry page of
packages (MonetDB, sql, pathfinder, etc.) with such
tests, and point out that changes in buildtools and/or
MonetDB might cause failures in sql, pathfinder, etc.
This bug report should either be assigned to one of the
developers that did check in since the last test run or
even be replicated and assigned to each developer that
did check in since the last test run.
In either case, all developers that did check in since
the last test run should be listed to allow
communication among the developers.
Possibly, even a list of all checkins since the last
test run could be added.
Implementing this maily requires some HTML hacking to
fill in the SF bug submission form non-interactively.
All other ingrediens should already be available in the
TestTools code --- at least in some basic version.
## Comment 13504
Date: 2006-09-22 15:39:01 +0200
From: @grobian
Logged In: YES
user_id=963970
this would result in endless bugs for:
a) known bugs
b) dangling bugs (those who fail today, and don't tomorrow)
Apart from some session related (log-in stuff) difficulties
of TestTools filing bugs, I'd like it when it would just be
able to effectively notify a (group of) user(s) by sending a
targeted mail. Targetted as in: not +43K mails, that make
it hard to retrieve the actual information.
## Comment 13505
Date: 2009-06-04 22:17:44 +0200
From: @grobian
With SF's login mechanism this simply looks like being almost impossible to me.
## Comment 13506
Date: 2009-06-05 14:05:44 +0200
From: @drstmane
I agree that my ambitions with this one were/are too eager --- in fact, the "error extraction email" that we have for quite some time, now, mostly serve the purpose of my original intention of more focused and informative (compilation-) error reporting.
Closing this one.
## Comment 13507
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1563499 at http://sourceforge.net/support/tracker.php?aid=1563499
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| TestTools (CWI internal): SF bug report from test results | https://api.github.com/repos/MonetDB/MonetDB/issues/2425/comments | 0 | 2020-11-30T10:43:10Z | 2024-06-28T13:43:33Z | https://github.com/MonetDB/MonetDB/issues/2425 | 753,365,160 | 2,425 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-08-24 06:00:04 +0200
From: Kevin Benson <<stormshadowtx>>
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2007-11-10 05:39:40 +0100
## Comment 13499
Date: 2006-08-24 18:00:04 +0200
From: Kevin Benson <<stormshadowtx>>
I currently use XMLDB:API to connect to eXist and have
been looking at MonetDB as also using XMLDB for
querying. I don't use XUpdate just the storeResource
areas from XMLDB. It would be nice if I could store
xml via this api (similiar I guess to shred_doc in
your mil). currently from looking at your
Resource.java code you don't support storage or
collections. Be nice if you did any chance your
working on this?
Thanks you for any replies,
Kevin Benson
## Comment 13500
Date: 2006-08-31 13:48:43 +0200
From: @grobian
Logged In: YES
user_id=963970
Hi Kevin,
XML:DB is as good as dead. A group seems to be working with
Sun on the XQJ API, that is similar to JDBC. Hence, I don't
like to make any real development efforts on XML:DB. Very
recently we had an addition in the development branch of
XQuery a feature added to shred documents from stream.
Since this will also be used in XQJ, I am willing to look at
how to implement this in JDBC and how to hook it up in
XML:DB. Can you give me any hints on which methods of which
classes I have to touch/implement to use storeResource?
## Comment 13501
Date: 2007-11-10 17:39:40 +0100
From: @grobian
Logged In: YES
user_id=963970
Originator: NO
CLOSED/NEEDINFO
Please reopen if there is a continued interest.
## Comment 13502
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1546069 at http://sourceforge.net/support/tracker.php?aid=1546069
| More of XMLDB available | https://api.github.com/repos/MonetDB/MonetDB/issues/2424/comments | 0 | 2020-11-30T10:43:07Z | 2024-06-28T13:43:33Z | https://github.com/MonetDB/MonetDB/issues/2424 | 753,365,126 | 2,424 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-08-18 04:17:06 +0200
From: @peterboncz
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2006-11-22 09:50:20 +0100
## Comment 13488
Date: 2006-08-18 16:17:06 +0200
From: @peterboncz
Due to the change in XQuery parameter passing I was
exposed to the current state of the mapi protocol,
which left me puzzled about its inefficiency.
This inefficiency does not show when querying localhost
or even client-server on a LAN. But XQuery is an
internet technology, and the internet is not a LAN.
Maybe one needs to have gone through the experience of
deploying MonetDB for a multi-national company in a WAN
architecture, as I have (servers in Norway, users all
over the world -- complaining about the resonse times
of their GUIs). Basically, in WAN, TCP/IP roundtrips
cost 0.3-0.4 seconds, which leads to the desire to
*minimize * their number.
Currently, MAPI performs a network handshake for each
query-language option (something starting woth 'X')
passed from client to server.
It would a step forwards to buffer all options in a
single network handshake.
And it would be best to be able to combine the options
with the first block of the query, in a single network
handshake. In that case 95% of the queries will send
one network message containing options and query, and
receive one message (-stream) with the answer, paying
single round-trip costs only.
Note that what I write here is based on superfically
observing the structure of stream-flushes and
stream-reads in pathfinder/runtime/pathfinder.mx
(xquery_server_engine). Maybe on the network level,
things get buffered after all by the strea, library??
some MAPI expert please react..
## Comment 13489
Date: 2006-08-18 16:27:20 +0200
From: @grobian
Logged In: YES
user_id=963970
as far as I can see the options are sent once and in one go,
and subsequent writes are done on a buffered stream, hence
get only written upon an explicit flush or when the buffer
is full.
Does this answer your concerns? If so, close the bug.
## Comment 13490
Date: 2006-08-18 16:47:28 +0200
From: @peterboncz
Logged In: YES
user_id=591107
Thanks, Fabian. It may be my lack of understanding, but I
see one output-flush and input-read for each option. Can
you explain how this leads to just one network message?
And still, it would be better to not flush/handshake between
the parameters and the query block. Just read a block on the
server end, and peel off the mapi messages received, asking
for more when the semantics derived from the protocol tell
you so.
So, I'll leave it open, then.
## Comment 13491
Date: 2006-08-18 17:47:57 +0200
From: @grobian
Logged In: YES
user_id=963970
Just to be sure we talk about the same things here. A
strace for me does the following:
% echo 1 + 1 | strace $INSTALL_DIR/bin/MapiClient -lxq
server challenge:
read(3, "12", 2) = 2
read(3, "::mserver:5\n", 12) = 12
client response:
write(3, "monetdb:monetdb:xquery:blocked\n", 31) = 31
bytesex handshake (I'd like to get rid of it, but it can be
useful too):
write(3, "\322\4", 2) = 2
read(3, "\322\4", 2) = 2
block with client query:
write(3, "\17\0", 2) = 2
write(3, "S1 + 1\n", 7) = 7
block with server answer:
read(3, "\215\0", 2) = 2
read(3, "<?xml version=\"1.0\" encoding=\"ut"..., 70) = 70
with an output mode:
% echo 1 + 1 | strace $INSTALL_DIR/bin/MapiClient -lxq -oseq
(same stuff):
read(3, "12", 2) = 2
read(3, "::mserver:5\n", 12) = 12
write(3, "monetdb:monetdb:xquery:blocked\n", 31) = 31
write(3, "\322\4", 2) = 2
read(3, "\322\4", 2) = 2
mode block:
write(3, "\31\0", 2) = 2
write(3, "Xoutput seq\n", 12) = 12
(write and read query/answer)
write(3, "\17\0", 2) = 2
write(3, "S1 + 1\n", 7) = 7
read(3, "\271\0", 2) = 2
read(3, "&1 0 1 1 1\n xquery.result\t tab"..., 92) = 92
You can argue that this can be more optimal in some places,
but I don't see so much inefficiency here. How do you run
MapiClient?
## Comment 13492
Date: 2006-08-18 18:52:34 +0200
From: @peterboncz
Logged In: YES
user_id=591107
Hm, it becomes worse..
(1) first the client handshakes to see what version of the
server its meeting.
(2) then, a second handshake to see byte order (?) or
something ('bytesex' handshake).
(3) then, a handshake for the output mode
(4) final handshake for query and response.
what I was originally objecting against, was that if you do
MapiClient -T -oseq bla.xq
so with *two* options (-T profiling and -o output mode), we
get two separate handshakes.
So you made me see that in total in that case there are 5
handshakes. With the 300msec roundtrip to that server in
Norway we are at already lost 1.5 seconds then for the dummy
query (of which 1.2 sec arguably for nothing - as an
optimized protocol would just need one roundtrip).
If you don't want to fully redesign, one could at least combine:
- mserver version and byte exchange
- all subsequent option specifications
but it is surprising that round-trip latency was apparently
not considered during the recent redesign of the mapi
protocol (what was, then?).
Peter
## Comment 13493
Date: 2006-08-18 18:57:33 +0200
From: @grobian
Logged In: YES
user_id=963970
It was, and it was way worse. A challenge/response is
unavoidable. Our startup costs, are quite low compared to
others. One should just use persistent connections,
something which _also_ works for XQuery for a very very long
time. Since Niels optimised for persistent connections, I
would suggest you to consider persistent connections on your
trip to Norway. Many application servers have this built in
by default.
## Comment 13494
Date: 2006-08-18 18:58:16 +0200
From: @grobian
Logged In: YES
user_id=963970
ohw, by the way, I agree that the separate sending of the
options could/should be done at once, if possible.
## Comment 13495
Date: 2006-08-18 21:04:54 +0200
From: @mlkersten
Logged In: YES
user_id=490798
A significant amount of energy was spent in reducing the
overhead of client interactions and streamlining the MAPI
protocol. The thread reporting on the result and some open
issues are archived under
http://sourceforge.net/mailarchive/message.php?msg_id=14228009
Before spending more time in this area (aside from minor
cleanups) a clear application and benchmark
requirement should be on the table. Also taking into account
that modern client-server application implementation
techniques have been deployed.
I suggest to close this bug report. It is at best a feature
request to further improve the protocol.
## Comment 13496
Date: 2006-08-21 11:15:36 +0200
From: @peterboncz
Logged In: YES
user_id=591107
Hi,
It is indeed a feature request.
As Fabian writes, for long running applications that keep
their connection open, this is not much of an issue as
parameters will be passed infrequently, mostly at session
startup. It is also true to say that in most IT
infrastructures, low-level database connections are not used
over WANs, rather work behind firewalls in the back end.
Over the WAN, HTTP is the protocol of choice.
But it remains that MAPI causes more handshakes than
necessary, and the impact can be 300msec per handshake
simply due to speed-of-light.
A handshake is only neccessary if the other sides needs a
response from the other side that potentially changes its
further way of operating. From the client's perspective,
knowing the server's version and its byte order passes this
test, as this may change the sub-protocol it uses in further
communication.
Passing options does not pass this test for me, unless the
server needs to give back answers to those option requests
that steer e.g. the format of a subsequent query.
Peter
## Comment 13497
Date: 2006-11-22 21:50:20 +0100
From: @grobian
Logged In: YES
user_id=963970
Originator: NO
I removed the block-mode protocol specific hand-shake in the next major version of MonetDB. I guess that removes all handshakes that are are too much.
## Comment 13498
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1542671 at http://sourceforge.net/support/tracker.php?aid=1542671
| mapi inefficient parameter passing | https://api.github.com/repos/MonetDB/MonetDB/issues/2423/comments | 0 | 2020-11-30T10:43:03Z | 2024-06-28T13:43:32Z | https://github.com/MonetDB/MonetDB/issues/2423 | 753,365,080 | 2,423 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-07-23 11:34:21 +0200
From: @mlkersten
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2010-08-20 14:04:00 +0200
## Comment 13486
Date: 2006-07-23 11:34:21 +0200
From: @mlkersten
Factor out the error handling to prepare for
internationalization.
## Comment 13487
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1527238 at http://sourceforge.net/support/tracker.php?aid=1527238
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
Bug has invalid status, setting status to "NEW".
Previous status was "DELETED".
## Comment 14321
Date: 2010-07-09 15:08:30 +0200
From: @sjoerdmullender
Bug was DELETED in SF.
| Internationalization | https://api.github.com/repos/MonetDB/MonetDB/issues/2422/comments | 0 | 2020-11-30T10:42:59Z | 2024-06-28T13:43:31Z | https://github.com/MonetDB/MonetDB/issues/2422 | 753,365,034 | 2,422 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-07-23 11:19:35 +0200
From: @mlkersten
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2016-04-11 11:43:14 +0200
## Comment 13483
Date: 2006-07-23 11:19:35 +0200
From: @mlkersten
Factor out the error handling to prepare for
internationalization.
## Comment 13484
Date: 2009-12-01 11:41:24 +0100
From: @grobian
no, just use gettext which is designed exactly for this purpose.
## Comment 13485
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1527235 at http://sourceforge.net/support/tracker.php?aid=1527235
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Internationalization | https://api.github.com/repos/MonetDB/MonetDB/issues/2421/comments | 0 | 2020-11-30T10:42:56Z | 2024-06-28T13:43:31Z | https://github.com/MonetDB/MonetDB/issues/2421 | 753,365,004 | 2,421 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-06-18 08:13:57 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2007-06-13 08:47:17 +0200
## Comment 13480
Date: 2006-06-18 20:13:57 +0200
From: @mlkersten
I propose to selectively add session variables
depending on the backend. For M5 the following seem
problematic.
debug is dangerous as it could crash the system,
e.g. try
set debug=10; select count(*) from tables;
explain='plan' is already supported by the explain
statement modifier.
after you have set it there is no way back, e.g.
set explain='plan'; set explain=''; doesn't work
set explain='profile' I would turn it into a boolean
e.g. set profile=true or false;
The same holds for set explain='trace' ->
set trace= true or false
And of course also for 'explain' itself.
Setting the flags is handy for running SQL scripts
## Comment 13481
Date: 2007-06-13 20:47:17 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
Most of what is needed is already provided.
Close this until more specifics are specified.
## Comment 13482
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1508214 at http://sourceforge.net/support/tracker.php?aid=1508214
| SQL predefined variables | https://api.github.com/repos/MonetDB/MonetDB/issues/2420/comments | 0 | 2020-11-30T10:42:52Z | 2024-06-28T13:43:30Z | https://github.com/MonetDB/MonetDB/issues/2420 | 753,364,956 | 2,420 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-05-04 09:05:52 +0200
From: @grobian
To: @sjoerdmullender
Version: -- development
Last updated: 2006-05-05 12:22:57 +0200
## Comment 13476
Date: 2006-05-04 21:05:52 +0200
From: @grobian
Invoking MapiClient without password should not
terminate it:
% MapiClient -P
MapiClient: option requires an argument -- P
Instead it should ask for the password on the console.
In general it is not safe to enter a password on the
command line, so it would be nice if MapiClient would
do that. Now to get this behaviour, the -u flag has to
be used to give a username first.
## Comment 13477
Date: 2006-05-04 21:08:53 +0200
From: @grobian
Logged In: YES
user_id=963970
same holds for the -u flag. `MapiClient -u` should ask for
the username on the console. This is the same behaviour as
clients from other popular open source DBMSs
## Comment 13478
Date: 2006-05-05 12:22:57 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Seems like a feature request to me.
But this is in fact very simple to do, so I'll do it.
## Comment 13479
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1482065 at http://sourceforge.net/support/tracker.php?aid=1482065
| MapiClient -P flag should also work without argument | https://api.github.com/repos/MonetDB/MonetDB/issues/2419/comments | 0 | 2020-11-30T10:42:49Z | 2024-06-28T13:43:29Z | https://github.com/MonetDB/MonetDB/issues/2419 | 753,364,910 | 2,419 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-02-09 03:53:25 +0100
From: Matthew Cornell <<matthewcornell>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2006-11-22 10:15:43 +0100
## Comment 13469
Date: 2006-02-09 15:53:25 +0100
From: Matthew Cornell <<matthewcornell>>
Hi Folks,
Our support group wants to upgrade some of our Linux
boxes - which are currently running RedHat 9 - to
CentOS 4.2 (which I've never heard of before). Before I
give the OK, I'd greatly appreciate hearing about
MonetDB compatibility. We're currently running MonetDB
V4.6.2.
Thanks!
matt
Matthew Cornell, MS
Senior Research Software Engineer
Department of Computer Science
Computer Science Building
140 Governors Drive
University of Massachusetts
Amherst, MA 01003-9264
## Comment 13470
Date: 2006-02-09 22:36:15 +0100
From: @njnes
Logged In: YES
user_id=43556
CentOS is not a platform which we have access on, ie we cannot
test it. As it is a repackaged Redhat Enterprice (which we
do have) it is very likely that everything will work as
expected.
## Comment 13471
Date: 2006-02-10 12:51:37 +0100
From: @grobian
Logged In: YES
user_id=963970
from www.centos.org:
"CentOS conforms fully with the upstream vendors
redistribution policy and aims to be 100% binary compatible.
(CentOS mainly changes packages to remove upstream vendor
branding and artwork.)"
Combining it with:
"CentOS is an Enterprise-class Linux Distribution derived
from sources freely provided to the public by a prominent
North American Enterprise Linux vendor."
It makes me think like Niels that centos will not be any
problem for MonetDB. We'd love to hear back whether it
indeed works out of the box.
## Comment 13472
Date: 2006-02-10 15:55:05 +0100
From: Matthew Cornell <<matthewcornell>>
Logged In: YES
user_id=1311640
Thanks for the info, guys. I'll have them go ahead and
install it, and I'll report back (probably mid-next week) on
the results of the compilation.
## Comment 13473
Date: 2006-02-15 18:35:59 +0100
From: Matthew Cornell <<matthewcornell>>
Logged In: YES
user_id=1311640
A quick update: Our IT folks installed the OS, and my
MonetDB V4.6.2 binaries started just fine. I haven't the
time to try compiling, so no news there...
## Comment 13474
Date: 2006-11-22 22:15:43 +0100
From: @grobian
Logged In: YES
user_id=963970
Originator: NO
We currently have access to a CentOS 4.4 (Linux 2.6.9) box, but it's not used as it crashes during/compilation testing if I recall correctly.
Given the nature of this feature request, I think this is sufficient to close it.
## Comment 13475
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1428368 at http://sourceforge.net/support/tracker.php?aid=1428368
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| CentOS 4.2 support? | https://api.github.com/repos/MonetDB/MonetDB/issues/2418/comments | 0 | 2020-11-30T10:42:46Z | 2024-06-28T13:43:29Z | https://github.com/MonetDB/MonetDB/issues/2418 | 753,364,876 | 2,418 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-01-31 11:42:42 +0100
From: leemarks <<leemarks>>
To: @njnes
Version: -- development
Last updated: 2007-11-10 10:57:59 +0100
## Comment 13465
Date: 2006-01-31 11:42:42 +0100
From: leemarks <<leemarks>>
Currently MonetDB temporary tables can only live within
a single transaction. It would be useful if this could
be extended to also support connection-based life spans.
One approach is the way it's done in Postgres where the
desired behaviour can be selected with an 'on commit'
clause on the create table statement.
Specifically :-
CREATE TEMP TABLE tbl ( ... ) ON COMMIT
{ PRESERVE ROWS | DELETE ROWS | DROP }
where
PRESERVE ROWS gives a connection based life span
DELETE ROWS truncates the table at the end of each trans
DROP drops the table (the current monetdb behaviour)
## Comment 13466
Date: 2007-11-10 22:57:27 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
this feature has been implemented some time ago.
## Comment 13467
Date: 2007-11-10 22:57:59 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
this feature has been implemented some time ago.
## Comment 13468
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1420335 at http://sourceforge.net/support/tracker.php?aid=1420335
| Connection-scoped temporary tables | https://api.github.com/repos/MonetDB/MonetDB/issues/2417/comments | 0 | 2020-11-30T10:42:42Z | 2024-06-28T13:43:28Z | https://github.com/MonetDB/MonetDB/issues/2417 | 753,364,844 | 2,417 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-01-16 11:11:43 +0100
From: @MarcinZukowski
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2019-04-30 12:36:00 +0200
## Comment 13459
Date: 2006-01-16 11:11:43 +0100
From: @MarcinZukowski
It would be nice to be able to specify that all
MonetDB-created files and directories have a given
mask, e.g. 664/775.
This is useful e.g. when multiple people from the same
group want to work on the same database with MonetDB.
## Comment 13460
Date: 2006-01-16 11:18:21 +0100
From: @grobian
Logged In: YES
user_id=963970
How about setting the umask before starting the Mserver process?
I think allowing this is potentially dangerous, so it should
not be facilitated in any way...
## Comment 13461
Date: 2006-01-16 11:22:58 +0100
From: @MarcinZukowski
Logged In: YES
user_id=607094
umask doesn't work.
There are places in GDK when 755 is hardcoded.
## Comment 13462
Date: 2006-01-17 14:25:35 +0100
From: @MarcinZukowski
Logged In: YES
user_id=607094
One more comment.
Maybe I do something wrong, but umask also doesn't influence
file creation.
umask 0002
Mserver --dbfarm=$PWD --dbname=tmp
ls -l tmp/bat/BACKUP
I get
-rw-r--r-- 1 marcin ins 54 Jan 17 14:23 BBP.dir
I think possibility to specify umask for files/dir in
MonetDB could also be used to protect one's data, e.g. by
running Mserver --umask=077
## Comment 13463
Date: 2006-01-17 14:37:09 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
You cannot get more permissions through the use of umask.
Umask only removes permissions.
## Comment 13464
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1407204 at http://sourceforge.net/support/tracker.php?aid=1407204
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
## Comment 26817
Date: 2019-01-21 15:09:43 +0100
From: MonetDB Mercurial Repository <<hg>>
Changeset [8b97f691d862](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=8b97f691d862) made by Sjoerd Mullender <sjoerd@acm.org> in the MonetDB repo, refers to this bug.
For complete details, see [https//devmonetdborg/hg/MonetDB?cmd=changeset;node=8b97f691d862](https://dev.monetdb.org/hg/MonetDB?cmd=changeset;node=https//devmonetdborg/hg/MonetDB?cmd=changeset;node=8b97f691d862)
Changeset description:
Trust user's umask setting to protect created files/directories.
Note that monetdbd set umask to 077 (i.e. only access by owner).
This fixes bug #2416.
## Comment 26818
Date: 2019-01-21 15:10:38 +0100
From: @sjoerdmullender
monetdbd set umask to 077, but otherwise the user's umask is used to protect files and directories.
| GDK: file/dir creation mask | https://api.github.com/repos/MonetDB/MonetDB/issues/2416/comments | 0 | 2020-11-30T10:42:39Z | 2024-06-27T11:45:25Z | https://github.com/MonetDB/MonetDB/issues/2416 | 753,364,807 | 2,416 |
[
"MonetDB",
"MonetDB"
] | Date: 2006-01-05 03:03:49 +0100
From: @sjoerdmullender
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2016-04-11 11:46:30 +0200
## Comment 13452
Date: 2006-01-05 15:03:49 +0100
From: @sjoerdmullender
In interval types, the most significant part has a
leading precision, and if seconds are part of the
interval, it also has a seconds precision (number of
digits in the fraction). This means that an interval
of just second can have both:
INTERVAL SECOND(p,q)
with p the interval leading precision and q the
interval seconds precision.
MonetDB/SQL only allows a single value here (presumably
the leading precision, but I'm not sure about that).
## Comment 13453
Date: 2006-01-05 15:14:20 +0100
From: @grobian
Logged In: YES
user_id=963970
This bug seems to be related to http://sourceforge.net/support/tracker.php?aid=1357338. The SECONDS
in an interval should be of double type to be able to
represent fractions of seconds.
## Comment 13454
Date: 2006-01-15 10:57:52 +0100
From: @grobian
Logged In: YES
user_id=963970
This is more like a feature request. It will at least not
be included in the next release.
## Comment 13455
Date: 2007-01-17 19:45:14 +0100
From: @romulogoncalves
Logged In: YES
user_id=1498628
Originator: NO
Niels,
WHy this bug is not closed?
Regards,
Romulo
## Comment 13456
Date: 2009-02-17 11:58:14 +0100
From: @drstmane
Related
[ 1357338 ] SQL: EXTRACT returns wrong answer for SECOND
https://sourceforge.net/tracker/index.php?func=detail&aid=1357338&group_id=56967&atid=482468
has been closed with "won't fix".
What about this one?
Also "won't fix"?
Or should we rather turn both into a feature request?
## Comment 13457
Date: 2009-11-22 23:18:43 +0100
From: @njnes
moving to feature requests. This requires that we move to 'decimal' types for intervals.
## Comment 13458
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1397758 at http://sourceforge.net/support/tracker.php?aid=1397758
| implement interval seconds precision for INTERVAL SECOND | https://api.github.com/repos/MonetDB/MonetDB/issues/2415/comments | 0 | 2020-11-30T10:42:36Z | 2024-06-28T13:43:27Z | https://github.com/MonetDB/MonetDB/issues/2415 | 753,364,772 | 2,415 |
[
"MonetDB",
"MonetDB"
] | Date: 2005-11-09 12:57:59 +0100
From: @grobian
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2006-01-16 11:19:05 +0100
## Comment 13449
Date: 2005-11-09 12:57:59 +0100
From: @grobian
Currently, MapiClient can be instructed to use
auto_commit mode using \A and \a escape sequences.
This does not work when reading from a file, and
implies an uncompatibility with other SQL
implementations, as the SQL standard does not define
auto_commit mode. However, the SQL standard does
define a START TRANSACTION and COMMIT/ROLLBACK block,
which can be implicitly taken as auto_commit switch.
Using the JdbcClient utility this behaviour of
switching auto_commit upon such transaction block has
been experimented with, and proven not to limit the
user. Many other vendor's client implementation use
auto_commit as default too.
The SQL front-end should start in auto_commit mode, and
disable auto_commit mode when encountering START
TRANSACTION. Upon encountering a COMMIT or ROLLBACK
(not ROLLBACK TO SAVEPOINT) auto_commit mode is enabled
again.
A summary of the proposed changes:
- default to auto_commit mode
- disable auto_commit mode when in a START
TRANSACTION/COMMIT block
## Comment 13450
Date: 2006-01-16 11:19:05 +0100
From: @grobian
Logged In: YES
user_id=963970
this has been implemented for the next release.
## Comment 13451
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1352124 at http://sourceforge.net/support/tracker.php?aid=1352124
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-sql@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| SQL: (implicit) control over auto_commit mode needed | https://api.github.com/repos/MonetDB/MonetDB/issues/2414/comments | 0 | 2020-11-30T10:42:33Z | 2024-06-28T13:43:27Z | https://github.com/MonetDB/MonetDB/issues/2414 | 753,364,731 | 2,414 |
[
"MonetDB",
"MonetDB"
] | Date: 2005-08-25 10:50:51 +0200
From: @drstmane
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2009-04-29 05:17:09 +0200
## Comment 13445
Date: 2005-08-25 10:50:51 +0200
From: @drstmane
Since the BuildTools (autogen, Mx, mel, burg) have been
removed from the MonetDB and Pathfinder trees, the
respective generated files get no longer packages in
the MonetDB and pathfinder packages, and their
documentation (if available at all) gets no longer
generated.
Hence, we need to enable packaging (e.g., "make dist" &
"make rpm") and documentation generation (e.g., "make
docs") with the buildtools to at least generate to
priviously available documentation, again.
Note, the documentation (e.g., for the mel tool) is
also required for building the MonetDB web site.
This must be done before the next release of MonetDB[/*].
## Comment 13446
Date: 2009-04-28 14:44:19 +0200
From: @grobian
hmmm, "the next release" is a while ago
## Comment 13447
Date: 2009-04-29 05:17:09 +0200
From: @drstmane
Since we package and distribute only pre-bootstrapped sources, there is no need to package buildtools.
"Users" don't need them.
"Developers" use CVS and know how to use buildtools.
## Comment 13448
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1271216 at http://sourceforge.net/support/tracker.php?aid=1271216
| Documentation & Packaging of BuildTools | https://api.github.com/repos/MonetDB/MonetDB/issues/2413/comments | 0 | 2020-11-30T10:42:30Z | 2024-06-28T13:43:26Z | https://github.com/MonetDB/MonetDB/issues/2413 | 753,364,700 | 2,413 |
[
"MonetDB",
"MonetDB"
] | Date: 2005-07-06 11:18:58 +0200
From: Steffen Goeldner <<sgoeldner>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2006-11-22 10:10:36 +0100
## Comment 13442
Date: 2005-07-06 11:18:58 +0200
From: Steffen Goeldner <<sgoeldner>>
To be able to use (server-side) prepare/execute via Mapi,
I'd like to add mapi_get_tableid() to Mapi.mx:
+mapi_export int mapi_get_tableid(MapiHdl hdl);
+int
+mapi_get_tableid(MapiHdl hdl)
+{
+ struct MapiResultSet *result;
+
+ mapi_hdl_check0(hdl, "mapi_get_tableid");
+ if ((result = hdl->result) != 0)
+ return result->tableid;
+ mapi_setError(hdl->mid, "No query
result", "mapi_get_tableid", MERROR);
+ return 0;
+}
## Comment 13443
Date: 2006-11-22 22:10:36 +0100
From: @grobian
Logged In: YES
user_id=963970
Originator: NO
This patch was added to cvs already ages ago. Thanks.
## Comment 13444
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1233276 at http://sourceforge.net/support/tracker.php?aid=1233276
The original assignee of this bug does not have
an account here. Reassigning to the default assignee
for the component, bugs-common@monetdb.org.
Previous assignee was nobody@users.sourceforge.net.
| Mapi: add mapi_get_tableid | https://api.github.com/repos/MonetDB/MonetDB/issues/2412/comments | 0 | 2020-11-30T10:42:28Z | 2024-06-28T13:43:26Z | https://github.com/MonetDB/MonetDB/issues/2412 | 753,364,667 | 2,412 |
[
"MonetDB",
"MonetDB"
] | Date: 2005-06-30 04:15:42 +0200
From: thijs <<thijswes>>
To: @sjoerdmullender
Version: -- development
Last updated: 2006-12-12 10:52:49 +0100
## Comment 13438
Date: 2005-06-30 16:15:42 +0200
From: thijs <<thijswes>>
It would be nice if the shredder would support external
entities.
Currently, shredding the attached document (book.xml)
produces the error:
ERROR: XML input not well-formed.
This error is the case both when the chapter files
(chap*.xml) are empty and when they contain something
like.
<chapter>
bla bla bla
</chapter>
## Comment 13439
Date: 2006-11-22 22:18:46 +0100
From: @grobian
Logged In: YES
user_id=963970
Originator: NO
I'm not familiar with this stuff, but has this been fixed in recent versions by chance?
## Comment 13440
Date: 2006-12-12 10:52:49 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
Originator: NO
First of all, the attached document is indeed invalid. It declares a DOCTYPE but fails to declare the elements used. A better document would be
<?xml version="1.0"?>
<!DOCTYPE book [
<!ENTITY chap1 SYSTEM "chap1.xml">
<!ENTITY chap2 SYSTEM "chap2.xml">
<!ENTITY chap3 SYSTEM "chap3.xml">
<!ELEMENT book (chapter)*>
<!ELEMENT chapter (PCDATA)>
]>
<book>
&chap1;
&chap2;
&chap3;
</book>
I'm not totally sure if this document is valid. I think it is, but I'd have to study the spec more carefully than I care to do right now. It still fails though.
Having said all this, this is outside our control since it is an external library that handles the parsing of XML documents (xmllib2). So, the feature request should be addressed to the maintainers of that library and not to us.
## Comment 13441
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1230379 at http://sourceforge.net/support/tracker.php?aid=1230379
| XQ: shred external entities | https://api.github.com/repos/MonetDB/MonetDB/issues/2411/comments | 0 | 2020-11-30T10:42:24Z | 2024-06-28T13:43:25Z | https://github.com/MonetDB/MonetDB/issues/2411 | 753,364,638 | 2,411 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.