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: 2004-12-08 10:35:14 +0100
From: Steffen Goeldner <<sgoeldner>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2007-11-30 04:20:04 +0100
## Comment 2224
Date: 2004-12-08 10:35:14 +0100
From: Steffen Goeldner <<sgoeldner>>
OS: Windows 2000
Compiler: MS VC++ Toolkit 2003
Parser tools: UnxUtils
Running MSQLserver --debug=32
and a perl MapiLib test script,
the server aborts:
Assertion failed: s->nr < sizeof(s->buf),
file .\..\..\..\src\common\stream.mx, line 1792
## Comment 2225
Date: 2004-12-13 15:25:01 +0100
From: Steffen Goeldner <<sgoeldner>>
Logged In: YES
user_id=1174087
As of 2004-12-13 (after cvs update), I can no longer
reproduce this bug.
## Comment 2226
Date: 2005-11-09 14:37:00 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: RELEASED
Don't know how to test on Windows.
REASSIGN to Sjoerd.
## Comment 2227
Date: 2005-11-09 17:06:41 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: NO TEST
Testing perl scripts is not (yet) supported.
## Comment 2228
Date: 2006-11-10 04:20:01 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 2229
Date: 2006-11-29 10:04:31 +0100
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
over-ruling "sf-robot":
re-set to "Pending" since adding a proper test has neither been done nor finally been discarded, yet.
User (submitter) & developer (assignee),
please check, again.
## Comment 2230
Date: 2007-11-30 04:20:04 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
Originator: NO
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 2231
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1081218 at http://sourceforge.net/support/tracker.php?aid=1081218
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.
| Assertion failed: s->nr < sizeof(s->buf) | https://api.github.com/repos/MonetDB/MonetDB/issues/410/comments | 0 | 2020-11-30T08:33:20Z | 2024-06-27T10:59:54Z | https://github.com/MonetDB/MonetDB/issues/410 | 753,271,762 | 410 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-07 11:21:25 +0100
From: @menzowindhouwer
To: Jan Rittinger <<jan.rittinger>>
Version: -- development
Last updated: 2005-01-06 01:34:47 +0100
## Comment 2218
Date: 2004-12-07 11:21:25 +0100
From: @menzowindhouwer
My first attempts at using pathfinder fail. Notice that
all the following
xqueries work without problem with the latest Saxon ...
so I assume
that at least the queries are syntactically correct.
The following query:
declare namespace
tds="http://www.let.uu.nl/uilots/tds/ns/TDS/";
<results>
{
for $l in doc("TDS.xml")//tds:language
order by $l/@code return
<tds:language code="{ $l/@code }"/>
}
</results>
results in a parse error:
[mwindhouwer@FGW2004E054 pf]$ pf -M < simple.xq
Tue Dec 7 11:16:41 2004: syntax error, unexpected
$undefined, expecting $end on line 1 (next token is `;')
Tue Dec 7 11:16:41 2004: parse error: XQuery parsing
failed
Tue Dec 7 11:16:41 2004: halted in parser.y (PFparse),
line 2036
So I simplified this query to contain no namespaces:
<results>
{
for $l in doc("TDS.xml")//*:language
order by $l/@code return
<language code="{ $l/@code }"/>
}
</results>
results in a segmentation fault:
[mwindhouwer@FGW2004E054 pf]$ pf -M < test.xq
Segmentation fault
Maybe I'm using an unsupported feature (order by ?),
but well it
shouldn't crash ...
## Comment 2219
Date: 2004-12-07 11:44:34 +0100
From: Jan Rittinger <<jan.rittinger>>
Logged In: YES
user_id=993208
1. The namespace declaration is a unsupported feature
2. The segmentation fault has the same reason like http://sourceforge.net/support/tracker.php?aid=1076344.
A short solution is to disable recognize_join() in
PFprintMILtemp().
## Comment 2220
Date: 2004-12-11 21:18:34 +0100
From: @drstmane
Logged In: YES
user_id=572415
My latest patch of compiler/mil/milprint_summer.c
avoids the segfault with "pf -M";
but I'm not sure, whether my "fix" is correct,
hence some "expert" should have a look at it!
Could not add a proper test script to CVS, since there is no
sample document.
(For the time being,) "order by" is indeed not supported by
Pathfinder;
it simply issues a warning and ignores it.
## Comment 2221
Date: 2005-01-06 01:34:47 +0100
From: Jan Rittinger <<jan.rittinger>>
Logged In: YES
user_id=993208
see http://sourceforge.net/support/tracker.php?aid=1076344 - fixed
namespace declarations are still not supported
## Comment 2222
Date: 2005-11-09 14:34:14 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: NO TEST ADDED
1. The namespace declaration is a unsupported feature
2. The segmentation fault has the same reason like http://sourceforge.net/support/tracker.php?aid=1076344, which is ALREADY IN TESTWEB
## Comment 2223
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1080515 at http://sourceforge.net/support/tracker.php?aid=1080515
| unparsable xquery and a crash | https://api.github.com/repos/MonetDB/MonetDB/issues/409/comments | 0 | 2020-11-30T08:33:17Z | 2024-06-27T10:59:53Z | https://github.com/MonetDB/MonetDB/issues/409 | 753,271,726 | 409 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-07 10:43:26 +0100
From: @grobian
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2005-11-10 09:48:18 +0100
## Comment 2208
Date: 2004-12-07 10:43:26 +0100
From: @grobian
A few examples from the big SQL bible:
monetdb-> select date '1997-07-31' + interval '1' month;
+----------------------+
| sql_add_single_value |
+======================+
| 1997-08-31 |
+----------------------+
1 row
monetdb-> select date '1997-10-31' + interval '1'
month;
+----------------------+
| sql_add_single_value |
+======================+
| 1997-12-01 |
+----------------------+
1 row
monetdb-> select timestamp '1997-07-15 19:00:00' +
interval '9' hour;
Error: Types sec_interval(0,0) (lng) and timestamp(0,0)
(timestamp) are not equal
monetdb-> select time '19:00:00' + interval '9' hour;
Error: Types sec_interval(0,0) (lng) and time(0,0)
(daytime) are not equal
monetdb-> select date '1997-07-31' + interval '1' hour;
+----------------------+
| sql_add_single_value |
+======================+
| 1997-07-31 |
+----------------------+
1 row
Additionally, some arithmetic (as prescribed in Psalm
8) does not work:
monetdb-> select interval '0' year + interval '0' month;
Error: Binary operator:
sql_add(month_interval,month_interval) unknown
monetdb-> select interval '00:00' hour to minute +
interval '00:00' minute to second;
Error: Binary operator:
sql_add(sec_interval,sec_interval) unknown
The following expressions should generate an error
because interval fields must follow "the natural rules
for intervals" (p. 198):
monetdb-> select interval '2:74' hour to minute;
+--------------+
| single_value |
+==============+
| 11640 |
+--------------+
1 row
however, when trying to write this in a legitimate manner:
monetdb-> select interval '2' hour + interval '74' minute;
Error: Binary operator:
sql_add(sec_interval,sec_interval) unknown
## Comment 2209
Date: 2004-12-20 14:22:57 +0100
From: @njnes
Logged In: YES
user_id=43556
added missing functions
## Comment 2210
Date: 2005-11-09 14:18:17 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/interval_on_time_stamp.SF-1080488.*
## Comment 2211
Date: 2005-11-09 14:36:36 +0100
From: @grobian
Logged In: YES
user_id=963970
Jennie,
if I interpret your approved output correctly, the test
doesn't fail on
select interval '2:74' hour to minute;
which according to the specs should.
If this is really what is happening, then the bug should be
reopened and the test output adjusted (auto_commit problem,
but you can get around by reshuffling the statements ;) )
## Comment 2212
Date: 2005-11-09 14:41:53 +0100
From: @yzchang
Logged In: YES
user_id=341633
Fabian,
"select interval '2:74' hour to minute" didn't cause any
failure.
Would you please reopen the bug and adjust the output? Since
I can't change the bug status and don't what the right
output should be according to the specs.
Thanks!
## Comment 2213
Date: 2005-11-09 14:47:44 +0100
From: @grobian
Logged In: YES
user_id=963970
reopening bug.
the statement
select interval '2:74' hour to minute;
should result in an error.
## Comment 2214
Date: 2005-11-09 23:23:09 +0100
From: @njnes
Logged In: YES
user_id=43556
added overflow checking of interval
## Comment 2215
Date: 2005-11-10 09:32:52 +0100
From: @grobian
Logged In: YES
user_id=963970
you introduced a new bug when fixing this:
select interval '2' hour + interval '74' minute;
*IS* valid SQL99.
sorry... (approved output hence is invalid too)
## Comment 2216
Date: 2005-11-10 09:48:18 +0100
From: @njnes
Logged In: YES
user_id=43556
fixed now we can handle both
select interval '2' hour + interval '74' minute; -- which is
correct
select interval '2:74' hour to minute; -- which will now
correctly return an error
## Comment 2217
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1080488 at http://sourceforge.net/support/tracker.php?aid=1080488
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: INTERVAL does not work on TIME and TIMESTAMP data types | https://api.github.com/repos/MonetDB/MonetDB/issues/408/comments | 0 | 2020-11-30T08:33:14Z | 2024-06-27T10:59:52Z | https://github.com/MonetDB/MonetDB/issues/408 | 753,271,697 | 408 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-06 06:36:38 +0100
From: @peterboncz
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-12-07 05:15:52 +0100
## Comment 2201
Date: 2004-12-06 18:36:38 +0100
From: @peterboncz
I get an empty result for q4.
this can't be right, can it.
let $auction := doc('auctions10MB.xml')
for $b in $auction/site/open_auctions/open_auction
where some $pr1 in $b/bidder/personref
[@person="person20"],
$pr2 in $b/bidder/personref
[@person="person51"] satisfies
$pr1 << $pr2
return <history> { $b/reserve/text() } </history>
## Comment 2202
Date: 2004-12-06 19:32:58 +0100
From: @drstmane
Logged In: YES
user_id=572415
- q4 (and all other XMark queries except q18 & q19) did work
fine to day during testing (i.e., before Peter's today's
changes to Pathfinder); cf.
http://monetdb.cwi.nl/testing/projects/monetdb/Current/pathfinder/.mTests103/index.html
- I just ran "Mtest_pathfinder -r" (i.e. Mtest with
Pathfinder) with Pathfinder compiled with the code base as
of yesterday, and I got
## Comment 2203
Date: 2004-12-06 19:41:50 +0100
From: @drstmane
Logged In: YES
user_id=572415
[ sorry, accidential premature submit; I'll try it again... ]
- q4 (and all other XMark queries except q18 & q19) did work
fine to day during testing (i.e., before Peter's today's
changes to Pathfinder); cf.
http://monetdb.cwi.nl/testing/projects/monetdb/Current/pathfinder
/.mTests103/index.html
- I just ran "Mtest_pathfinder -r" (i.e. Mtest with
Pathfinder) with Pathfinder compiled with the code base as
of yesterday, and I got
" 4 out of 55 tests produced SIGNIFICANTLY different
output"
^
(these 4 tests are the usage tests of compiler and loader
due to the echoed version number, and XMark's q18 & q19)
- I then updated to the most recent version of Pathfinder
(i.e., added Peter's changes of today), recompiled and run
Mtest:
"24 out of 55 tests produced SIGNIFICANTLY different
output"
^^
All XMark queries now seem to produce different output --- a
(quick) closer look seems to reveal, that the output is
basically the same, only the format has changed. If that was
desired/intended, the new output should be approved.
Also q4 does still produce (basically) the same result as
before:
"<history>389.61</history>" (with the Galax XMark document).
## Comment 2204
Date: 2004-12-07 14:55:49 +0100
From: @peterboncz
Logged In: YES
user_id=591107
It is strange that my Mserver (64-bits opteron, 32 bits
oids) does produce the correct result in the test web, but
in my separate XMark database, it does not (on none of
the sizes 116KB, 1MB, 10MB, 100MB, 1GB).
to be continued.....
My query is:
let $auction := doc("auctions10MB.xml")
for $b in $auction/site/open_auctions/open_auction
where some $pr1 in $b/bidder/personref
[@person="person20"],
$pr2 in $b/bidder/personref
[@person="person51"] satisfies
$pr1 << $pr2
return <history> { $b/reserve/text() } </history>
and I just do:
pf -M 04.xq | Mserver
which has always worked for all my other xmark queries
## Comment 2205
Date: 2004-12-07 17:15:52 +0100
From: @drstmane
Logged In: YES
user_id=572415
Peter,
if you're using Jan's XMark documents, an empty result for
Q4 seems to be correct for all sizes but the Galax document:
the simple grep below reveals that in none but the Galax
document there is an open_auctions with bidders "person20"
and "person51".
I supposed, the person id's that do yield a non-empty result
depend on the very document --- afterall, there is some
randomness involved when generating the document...
Stefan
========
egrep 'open_auction|personref person="person(20|51)"'
~rittinge/xml/*xml
~rittinge/xml/auctions/auctions{116K,1{{,0{,0}}M,G}}B.xml |
grep -1 personref
--------
/ufs/rittinge/xml/auction_galax.xml:<open_auction
id="open_auction1">
/ufs/rittinge/xml/auction_galax.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auction_galax.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auction_galax.xml:</open_auction>
--
/ufs/rittinge/xml/auction_galax.xml:<open_auction
id="open_auction3">
/ufs/rittinge/xml/auction_galax.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auction_galax.xml:</open_auction>
--
/ufs/rittinge/xml/auctions116KB.xml:<open_auction
id="open_auction1">
/ufs/rittinge/xml/auctions116KB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions116KB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions116KB.xml:<open_auction
id="open_auction3">
/ufs/rittinge/xml/auctions116KB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions116KB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions116KB.xml:<open_auction
id="open_auction1">
/ufs/rittinge/xml/auctions/auctions116KB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions116KB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions116KB.xml:<open_auction
id="open_auction3">
/ufs/rittinge/xml/auctions/auctions116KB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions116KB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions1MB.xml:<open_auction
id="open_auction52">
/ufs/rittinge/xml/auctions/auctions1MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions1MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions1MB.xml:<open_auction
id="open_auction59">
/ufs/rittinge/xml/auctions/auctions1MB.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auctions/auctions1MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions1MB.xml:<open_auction
id="open_auction92">
/ufs/rittinge/xml/auctions/auctions1MB.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auctions/auctions1MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions1MB.xml:<open_auction
id="open_auction109">
/ufs/rittinge/xml/auctions/auctions1MB.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auctions/auctions1MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions1MB.xml:<open_auction
id="open_auction117">
/ufs/rittinge/xml/auctions/auctions1MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions1MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions10MB.xml:<open_auction
id="open_auction219">
/ufs/rittinge/xml/auctions/auctions10MB.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auctions/auctions10MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions10MB.xml:<open_auction
id="open_auction637">
/ufs/rittinge/xml/auctions/auctions10MB.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auctions/auctions10MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions10MB.xml:<open_auction
id="open_auction653">
/ufs/rittinge/xml/auctions/auctions10MB.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auctions/auctions10MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions10MB.xml:<open_auction
id="open_auction1093">
/ufs/rittinge/xml/auctions/auctions10MB.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auctions/auctions10MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions100MB.xml:<open_auction
id="open_auction945">
/ufs/rittinge/xml/auctions/auctions100MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions100MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions100MB.xml:<open_auction
id="open_auction2196">
/ufs/rittinge/xml/auctions/auctions100MB.xml:<personref
person="person51"/>
/ufs/rittinge/xml/auctions/auctions100MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions100MB.xml:<open_auction
id="open_auction3333">
/ufs/rittinge/xml/auctions/auctions100MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions100MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions100MB.xml:<open_auction
id="open_auction3957">
/ufs/rittinge/xml/auctions/auctions100MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions100MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions100MB.xml:<open_auction
id="open_auction5227">
/ufs/rittinge/xml/auctions/auctions100MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions100MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions100MB.xml:<open_auction
id="open_auction8085">
/ufs/rittinge/xml/auctions/auctions100MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions100MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions100MB.xml:<open_auction
id="open_auction8983">
/ufs/rittinge/xml/auctions/auctions100MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions100MB.xml:</open_auction>
--
/ufs/rittinge/xml/auctions/auctions100MB.xml:<open_auction
id="open_auction11069">
/ufs/rittinge/xml/auctions/auctions100MB.xml:<personref
person="person20"/>
/ufs/rittinge/xml/auctions/auctions100MB.xml:</open_auction>
========
## Comment 2206
Date: 2005-11-09 13:43:09 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: ALREADY IN TESTWEB
And this bug is invalid
## Comment 2207
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1080087 at http://sourceforge.net/support/tracker.php?aid=1080087
| xmark q4 does not work | https://api.github.com/repos/MonetDB/MonetDB/issues/407/comments | 0 | 2020-11-30T08:33:11Z | 2024-06-28T13:35:46Z | https://github.com/MonetDB/MonetDB/issues/407 | 753,271,667 | 407 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-06 03:54:15 +0100
From: @sjoerdmullender
To: @njnes
Version: -- development
Last updated: 2004-12-08 09:44:40 +0100
## Comment 2197
Date: 2004-12-06 15:54:15 +0100
From: @sjoerdmullender
This SQL statement causes the server to abort:
select 5 not between 4 and 6;
## Comment 2198
Date: 2004-12-08 09:44:40 +0100
From: @njnes
Logged In: YES
user_id=43556
between in selection wasn't handled properly
## Comment 2199
Date: 2005-11-09 13:41:53 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/sql_server_abort.SF-1080025.*
## Comment 2200
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1080025 at http://sourceforge.net/support/tracker.php?aid=1080025
| sql server abort | https://api.github.com/repos/MonetDB/MonetDB/issues/406/comments | 0 | 2020-11-30T08:33:08Z | 2024-06-27T10:59:50Z | https://github.com/MonetDB/MonetDB/issues/406 | 753,271,633 | 406 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-06 03:50:16 +0100
From: @sjoerdmullender
To: @njnes
Version: -- development
Last updated: 2004-12-08 09:42:22 +0100
## Comment 2193
Date: 2004-12-06 15:50:16 +0100
From: @sjoerdmullender
This SQL statement causes the server to crash:
select NULLIF(NULLIF('first','second'),'first');
## Comment 2194
Date: 2004-12-08 09:42:22 +0100
From: @njnes
Logged In: YES
user_id=43556
added missing code
## Comment 2195
Date: 2005-11-09 13:37:31 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/sql_server_crash.SF-1080024.*
## Comment 2196
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1080024 at http://sourceforge.net/support/tracker.php?aid=1080024
| sql server crash | https://api.github.com/repos/MonetDB/MonetDB/issues/405/comments | 0 | 2020-11-30T08:33:06Z | 2024-06-27T10:59:49Z | https://github.com/MonetDB/MonetDB/issues/405 | 753,271,604 | 405 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-03 06:08:09 +0100
From: @arjenpdevries
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2005-01-14 01:50:44 +0100
## Comment 2184
Date: 2004-12-03 18:08:09 +0100
From: @arjenpdevries
Load data:
----------
XMLFILE=bond-repository.xml
pf-shred -v -c $XMLFILE
Mserver --dbname=bond-demo << __EOF__
module(pathfinder);
var wdir := monet_environment.find("monet_cwd") + "/";
import_doc(wdir + "$XMLFILE", "$XMLFILE");
__EOF__
Mserver, module(streams,mapi,pathfinder), fork(listen());
time pf -M << __EOF__ | MapiClient
<result title="Document $XMLFILE category Corel1/Croatia:">
{ for \$e in
doc("$XMLFILE")//Category[@name="Corel1/Croatia"]
return \$e }
</result>
__EOF__
Maybe the XQuery is also not correct;
it gives some warnings about string / string* conversions.
But, it crashes the MServer as well...
Cheers,
Arjen
## Comment 2185
Date: 2004-12-03 18:12:02 +0100
From: @arjenpdevries
Logged In: YES
user_id=572978
File cannot be uploaded (too large), but it can be found here:
http://www.cwi.nl/~arjen/tmp/bond-repository.xml.bz2
Cheers!
BTW: the query also takes too long :-)
## Comment 2186
Date: 2004-12-11 21:14:12 +0100
From: @drstmane
Logged In: YES
user_id=572415
My latest patch of compiler/mil/milprint_summer.c
avoids this segfault;
but I'm not sure, whether my "fix" is correct,
hence some "expert" should have a look at it!
Also added a test scipt for this bug to the CVS repository.
The warnings about string / string* conversions do still
occur, but the result seems to be correct.
## Comment 2187
Date: 2004-12-17 14:38:35 +0100
From: Jan Rittinger <<jan.rittinger>>
Logged In: YES
user_id=993208
"Fri Dec 17 13:17:17 2004: cast from 'string' to 'string*'
ignored" is only a debugging message, which says that it omits
the cast from string* to string - which of course wouldn't
make sense :)
What does "BTW: the query also takes too long :-)" mean - it
takes 1.1 sec on my machine.
By the way, PLEASE look if the output is correct. Found a bug
in this query: attribute output is wrong.
## Comment 2188
Date: 2004-12-18 15:56:00 +0100
From: @arjenpdevries
Logged In: YES
user_id=572978
JanR, did you mean "the query is wrong" or "the answer is
wrong" with
"Found a bug in this query"??
PS: Re: "it takes too long".
The query processing time is far too long in practice (btw,
3secs on my machine).
We aimed to use this query in a demo of BOND (an algorithm
for top-k content-based image querying). The images would be
ranked by content similarity in some 50-200x faster than
simply listing the images with a specific label... strange,
not?! So, we are now falling back on a simple SQL solution
for the latter.
## Comment 2189
Date: 2004-12-20 11:51:30 +0100
From: @drstmane
Logged In: YES
user_id=572415
Arjen,
JanR actually meant that the answer generated by Pathfinder
was wrong: the "title" attribute of the "result" node was
incorrectly empty. I hadn't seen this when I added the (i.e.
your) test script for this bug to CVS (cf.
http://monetdb.cwi.nl/testing/projects/monetdb/Current/pathfinder/.mTests103/GNU_32-d_Fedora2/tests_BugsViaSourgeforce/ID.1078462_run.out.00.html
). JanR fixed the bug in the code, and I just fixed the
stable test output.
... and we're also working on the "takes too long" issue, cf.
http://www.inf.uni-konstanz.de/dbis/pathfinder/wiki/index.php/Performance_Results
...
Stefan
PS: I don't close this bug, yet, as JanR's (temporary) fixes
cause some problems on other places...
## Comment 2190
Date: 2005-01-14 13:50:44 +0100
From: @drstmane
Logged In: YES
user_id=572415
Closing, now, since JanR seesm to have "finalized" his
"temporary" fixes --- at least all PF tests (including the
one for this bug report) do work fine, now...
## Comment 2191
Date: 2005-11-09 13:11:54 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: ALREADY IN TESTWEB
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1078462_run.stable.out
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1078462_run.stable.err
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1078462_import.milS
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1078462_import.stable.err
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1078462_doc.xml
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1078462_import.stable.out
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1078462_run.xq
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1078462_import.modules
## Comment 2192
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1078462 at http://sourceforge.net/support/tracker.php?aid=1078462
| XQuery query generating segmentation fault in server | https://api.github.com/repos/MonetDB/MonetDB/issues/404/comments | 0 | 2020-11-30T08:33:03Z | 2024-06-27T10:59:48Z | https://github.com/MonetDB/MonetDB/issues/404 | 753,271,567 | 404 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-02 08:23:44 +0100
From: @drstmane
To: @sjoerdmullender
Version: -- development
Last updated: 2004-12-08 04:47:39 +0100
## Comment 2181
Date: 2004-12-02 08:23:44 +0100
From: @drstmane
Since yesterdays changes, compiling MonetDB, both the
CVS head (v4.5) and the MonetDB_4-4-0 branch (v4.4.3)
fails on Windows with:
cl -GF -W3 -wd4273 -wd4102 -MD -nologo -Zi -G6
-I. -I.\..\.. "-IC:\Pthreads\include" -DHAVE_CONFIG_H
-I.\..\..\..\src\common -I.\..\..\..\src\common
-DLIBSTREAM "-Fostream.obj" -c "stream.c"
stream.c
..\..\..\src\common\stream.mx(1039) : error C2065:
'EWOULDBLOCK' : undeclared identifier
..\..\..\src\common\stream.mx(1102) : error C2065:
'SHUT_RDWR' : undeclared identifier
NMAKE : fatal error U1077: 'cl' : return code '0x2'
## Comment 2182
Date: 2005-11-09 13:08:09 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: NO TEST ADDED / COMPILATION
## Comment 2183
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1077361 at http://sourceforge.net/support/tracker.php?aid=1077361
| compiling MonetDB fails on Windows due to undefined marcos | https://api.github.com/repos/MonetDB/MonetDB/issues/403/comments | 0 | 2020-11-30T08:33:00Z | 2024-06-27T10:59:47Z | https://github.com/MonetDB/MonetDB/issues/403 | 753,271,533 | 403 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-01 10:22:52 +0100
From: @grobian
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2005-11-09 11:43:53 +0100
## Comment 2174
Date: 2004-12-01 10:22:52 +0100
From: @grobian
monetdb-> select 'test' like 'test';
Error: ERROR: interpret: no matching MIL operator to
'like(str, str)'.
MAYBE YOU MEAN:
like(BAT[any::1,str], str) : BAT[any::1,str]
obviously, in a where clause it does work:
monetdb-> select * from equality_test where str1 like
'test';
+------+------+
| str1 | str2 |
+======+======+
| test | test |
+------+------+
1 row
## Comment 2175
Date: 2004-12-01 22:26:45 +0100
From: @njnes
Logged In: YES
user_id=43556
the like mil function does not exist!
## Comment 2176
Date: 2004-12-22 00:48:38 +0100
From: @njnes
Logged In: YES
user_id=43556
missing like function is now part of the algebra module
## Comment 2177
Date: 2005-11-09 13:06:18 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: TEST ADDED / FAILURE
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/like_doesnot_like_strings.SF-1076605.*
This bug was closed. But if I now run the query mentioned in
the bug report,
I get errors. I will re-open this bug. Please correct it, if
I'm wrong.
## Comment 2178
Date: 2005-11-09 13:26:38 +0100
From: @njnes
Logged In: YES
user_id=43556
wrong like function is called
## Comment 2179
Date: 2005-11-09 23:43:53 +0100
From: @njnes
Logged In: YES
user_id=43556
fixed in MonetDB like procs
## Comment 2180
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1076605 at http://sourceforge.net/support/tracker.php?aid=1076605
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: LIKE doesn't like strings in select? | https://api.github.com/repos/MonetDB/MonetDB/issues/402/comments | 0 | 2020-11-30T08:32:57Z | 2024-06-27T10:59:46Z | https://github.com/MonetDB/MonetDB/issues/402 | 753,271,496 | 402 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-12-01 10:21:22 +0100
From: @drstmane
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Duplicates: #399
Last updated: 2010-05-04 11:13:54 +0200
## Comment 2170
Date: 2004-12-01 10:21:22 +0100
From: @drstmane
changes that were made between Sunday, 24.11.2004, and
Thursday, 28.11.2004, seem to cause various arithmetic
/ casting / nil-related problems with the PGI compiler
on the Opteron cluster running Debian-3.0:
Old file: Mtest10..2004.11.24_06-14-04..out
New file: Mtest10..2004.11.29_04-43-38..out
===========================================
Tests that are worse than before:
---------------------------------
D3.0 PGI64d src/gdk/casts
o o X o
D3.0 PGI64d src/modules/contrib/color
o o X o
D3.0 PGI64d src/modules/plain/algebra_PROCs
o o X o
D3.0 PGI64d src/modules/plain/arith
o o X o
D3.0 PGI64d src/modules/plain/mmath2
o o X o
D3.0 PGI64d src/modules/plain/str
o o X o
D3.0 PGI64d src/modules/plain/xtables
o o X o
D3.0 PGI64d tests/BugReports/no.148
o o X o
D3.0 PGI64d tests/suite_00/test_05
o o X o
D3.0 PGI64d tests/suite_01/test_70
o o X o
Tests that still fail:
----------------------
D3.0 PGI64d src/modules/plain/mmath
X o
D3.0 PGI64d src/modules/plain/mmath_nan
X o
See also
http://monetdb.cwi.nl/testing/intern/monetdb/Current/MonetDB/.mTests103/PGI_64-d_Debian3.0/src_gdk/casts.out.00.html
## Comment 2171
Date: 2004-12-01 13:25:04 +0100
From: @drstmane
Logged In: YES
user_id=572415
bug report somehow appeared twice ...
## Comment 2172
Date: 2005-11-09 12:36:39 +0100
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-11-09, Jennie: ALREADY IN TESTWEB
This bug is about failures of existing test scripts, so they
are already in testweb.
## Comment 2173
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1076604 at http://sourceforge.net/support/tracker.php?aid=1076604
Bug has invalid status, setting status to "NEW".
Previous status was "DELETED".
## Comment 13978
Date: 2010-05-04 11:13:54 +0200
From: @grobian
*** This bug has been marked as a duplicate of bug #399 ***
| arithmetic problems with PGI compiler on Opteron/Debia-3.0 | https://api.github.com/repos/MonetDB/MonetDB/issues/401/comments | 0 | 2020-11-30T08:32:54Z | 2024-06-28T13:35:45Z | https://github.com/MonetDB/MonetDB/issues/401 | 753,271,466 | 401 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-30 11:32:42 +0100
From: Jan Rittinger <<jan.rittinger>>
To: Jan Rittinger <<jan.rittinger>>
Version: -- development
Last updated: 2005-01-06 01:32:40 +0100
## Comment 2165
Date: 2004-11-30 23:32:42 +0100
From: Jan Rittinger <<jan.rittinger>>
for $a in (1,2) return 3 gives a segmentation fault
## Comment 2166
Date: 2004-12-11 21:12:53 +0100
From: @drstmane
Logged In: YES
user_id=572415
My latest patch of compiler/mil/milprint_summer.c
avoids this segfault;
but I'm not sure, whether my "fix" is correct,
hence some "expert" should have a look at it!
Also added a test scipt for this bug to the CVS repository.
## Comment 2167
Date: 2005-01-06 01:32:40 +0100
From: Jan Rittinger <<jan.rittinger>>
Logged In: YES
user_id=993208
fix was ok - perhaps a little bit complicated :)
## Comment 2168
Date: 2005-11-09 13:59:09 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: NO TEST ADDED / ALREADY IN TESTWEB
## Comment 2169
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1076344 at http://sourceforge.net/support/tracker.php?aid=1076344
| segmentation fault in join recognition | https://api.github.com/repos/MonetDB/MonetDB/issues/400/comments | 0 | 2020-11-30T08:32:51Z | 2024-06-27T10:59:44Z | https://github.com/MonetDB/MonetDB/issues/400 | 753,271,432 | 400 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-30 10:44:16 +0100
From: @drstmane
To: @sjoerdmullender
Version: -- development
Last updated: 2010-05-04 11:13:54 +0200
## Comment 2161
Date: 2004-11-30 22:44:16 +0100
From: @drstmane
changes that were made between Sunday, 24.11.2004, and
Thursday, 28.11.2004, seem to cause various arithmetic
/ casting / nil-related problems with the PGI compiler
on the Opteron cluster running Debian-3.0:
Old file: Mtest10..2004.11.24_06-14-04..out
New file: Mtest10..2004.11.29_04-43-38..out
===========================================
Tests that are worse than before:
---------------------------------
D3.0 PGI64d src/gdk/casts
o o X o
D3.0 PGI64d src/modules/contrib/color
o o X o
D3.0 PGI64d src/modules/plain/algebra_PROCs
o o X o
D3.0 PGI64d src/modules/plain/arith
o o X o
D3.0 PGI64d src/modules/plain/mmath2
o o X o
D3.0 PGI64d src/modules/plain/str
o o X o
D3.0 PGI64d src/modules/plain/xtables
o o X o
D3.0 PGI64d tests/BugReports/no.148
o o X o
D3.0 PGI64d tests/suite_00/test_05
o o X o
D3.0 PGI64d tests/suite_01/test_70
o o X o
Tests that still fail:
----------------------
D3.0 PGI64d src/modules/plain/mmath
X o
D3.0 PGI64d src/modules/plain/mmath_nan
X o
See also
http://monetdb.cwi.nl/testing/intern/monetdb/Current/MonetDB/.mTests103/PGI_64-d_Debian3.0/src_gdk/casts.out.00.html
## Comment 2162
Date: 2005-01-13 19:07:01 +0100
From: @drstmane
Logged In: YES
user_id=572415
Sjoerd's recent changes for NAN and INFINITY checks seem to
fix this bug.
Thanks, Sjoerd!
## Comment 2163
Date: 2005-11-09 13:56:22 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: NO TEST ADDED / COMPILATION
## Comment 2164
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1076316 at http://sourceforge.net/support/tracker.php?aid=1076316
## Comment 13979
Date: 2010-05-04 11:13:54 +0200
From: @grobian
*** Bug #401 has been marked as a duplicate of this bug. ***
| arithmetic problems with PGI compiler on Opteron/Debia-3.0 | https://api.github.com/repos/MonetDB/MonetDB/issues/399/comments | 0 | 2020-11-30T08:32:49Z | 2024-06-27T10:59:43Z | https://github.com/MonetDB/MonetDB/issues/399 | 753,271,391 | 399 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-30 06:46:50 +0100
From: @peterboncz
To: Jan Rittinger <<jan.rittinger>>
Version: -- development
Last updated: 2004-12-22 08:49:11 +0100
## Comment 2157
Date: 2004-11-30 18:46:50 +0100
From: @peterboncz
conditions in xpath steps are evaluated from the root
instead of local from the context nodes.
//A[.//B]
e.g. all A-s that have a B-child, does not work correctly
(it returns all A-s that have a B ancestor or descendent)
## Comment 2158
Date: 2004-12-22 20:49:11 +0100
From: Jan Rittinger <<jan.rittinger>>
Logged In: YES
user_id=993208
self axis is now implemented and root(), position(), as well as
last() work
## Comment 2159
Date: 2005-11-09 13:55:07 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: TEST ADDED / SUCCESS
bug is also implicitly tested in XPathMark/Tests/Q15.xq
## Comment 2160
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1076161 at http://sourceforge.net/support/tracker.php?aid=1076161
| xpath steps | https://api.github.com/repos/MonetDB/MonetDB/issues/398/comments | 0 | 2020-11-30T08:32:46Z | 2024-06-27T10:59:42Z | https://github.com/MonetDB/MonetDB/issues/398 | 753,271,365 | 398 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-30 06:43:25 +0100
From: @peterboncz
To: Jan Flokstra <<jflokstra>>
Version: -- development
Last updated: 2004-12-10 02:33:43 +0100
## Comment 2151
Date: 2004-11-30 18:43:25 +0100
From: @peterboncz
when run on the 1GB xmark document, this crashes.
note that the 1GB document can only be processed on
64-bits systems with a lot of memory.
maybe it also crashes on smaller sizes, but you have to
modify the value of the document name (auction_xml)
in the attached script, probably.
## Comment 2152
Date: 2004-12-07 10:26:28 +0100
From: Maurice van Keulen <<mvankeulen>>
Logged In: YES
user_id=654674
Assigned to Jan Flokstra (MvK)
## Comment 2153
Date: 2004-12-08 10:36:14 +0100
From: @peterboncz
Logged In: YES
user_id=591107
XQUERY
element v { doc("auctions1GB.xml") }
## Comment 2154
Date: 2004-12-10 14:33:43 +0100
From: Jan Flokstra <<jflokstra>>
Logged In: YES
user_id=1054297
Solved the bug!
Problem was incorrect implementation of data sharing of
fragments
## Comment 2155
Date: 2005-11-09 13:11:39 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: NO TEST ADDED
no test added because document is too large
## Comment 2156
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1076157 at http://sourceforge.net/support/tracker.php?aid=1076157
| crash in print_result | https://api.github.com/repos/MonetDB/MonetDB/issues/397/comments | 0 | 2020-11-30T08:32:43Z | 2024-06-28T13:35:44Z | https://github.com/MonetDB/MonetDB/issues/397 | 753,271,333 | 397 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-23 09:35:51 +0100
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-11-26 06:52:36 +0100
## Comment 2145
Date: 2004-11-23 21:35:51 +0100
From: @grobian
Not all queries executed seem to be recorded in the
history table; queries with where clauses are not stored.
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.4.3
Driver: MonetDB Native Driver 1.0 (Dawn SR-1)
Type \q to quit, \h for a list of available commands
auto commit mode: on
monetdb-> set explain='performance';
monetdb-> delete from history;
9 affected rows
monetdb-> \d
TABLE sys.bond_demo
TABLE sys.entry
TABLE sys.link
monetdb-> select * from bond_demo limit 10;
+-----------------------------------+
| image |
+===================================+
| content/Corel1/RuralUK/131000.jpg |
| content/Corel1/RuralUK/131001.jpg |
| content/Corel1/RuralUK/131002.jpg |
| content/Corel1/RuralUK/131003.jpg |
| content/Corel1/RuralUK/131004.jpg |
| content/Corel1/RuralUK/131005.jpg |
| content/Corel1/RuralUK/131006.jpg |
| content/Corel1/RuralUK/131007.jpg |
| content/Corel1/RuralUK/131008.jpg |
| content/Corel1/RuralUK/131009.jpg |
+-----------------------------------+
10 rows
At this stage already a few queries are missing: the
JDBC queries are not shown:
monetdb-> select * from history;
+-------------------------+-----------------------------------+----------+----------+----------+----------+---------+
| start | query
| parse | optimize | exec | total |
user |
+=========================+===================================+==========+==========+==========+==========+=========+
| 2004-11-23 20:32:25.0 | delete from history;
| 37 | 1510 | 1876 | 3426 |
monetdb |
| 2004-11-23 20:32:54.0 | set reply_size = 125;
| 54 | 1199 | 997 | 2253 |
monetdb |
| 2004-11-23 20:33:03.0 | set reply_size = 250;
| 56 | 10 | 357 | 424 |
monetdb |
| 2004-11-23 20:33:03.0 | select * from bond_demo
limit 10; | 60 | 699 | 1094 | 1856 |
monetdb |
+-------------------------+-----------------------------------+----------+----------+----------+----------+---------+
4 rows
monetdb-> select * from bond_demo where image like
'content/Corel1/RuralUK/131%.jpg';
+-----------------------------------+
| image |
+===================================+
| content/Corel1/RuralUK/131000.jpg |
| content/Corel1/RuralUK/131001.jpg |
...
| content/Corel1/RuralUK/131098.jpg |
| content/Corel1/RuralUK/131099.jpg |
+-----------------------------------+
100 rows
monetdb-> select * from history;
+-------------------------+-----------------------------------+----------+----------+----------+----------+---------+
| start | query
| parse | optimize | exec | total |
user |
+=========================+===================================+==========+==========+==========+==========+=========+
| 2004-11-23 20:32:25.0 | delete from history;
| 37 | 1510 | 1876 | 3426 |
monetdb |
| 2004-11-23 20:32:54.0 | set reply_size = 125;
| 54 | 1199 | 997 | 2253 |
monetdb |
| 2004-11-23 20:33:03.0 | set reply_size = 250;
| 56 | 10 | 357 | 424 |
monetdb |
| 2004-11-23 20:33:03.0 | select * from bond_demo
limit 10; | 60 | 699 | 1094 | 1856 |
monetdb |
| 2004-11-23 20:33:23.0 | select * from history;
| 53 | 1188 | 2159 | 3402 |
monetdb |
+-------------------------+-----------------------------------+----------+----------+----------+----------+---------+
5 rows
## Comment 2146
Date: 2004-11-26 11:10:16 +0100
From: @grobian
Logged In: YES
user_id=963970
problem was that the queries were attempted to be stored in
the history table without quoting them first, so ' chars in
the queries would brake it.
## Comment 2147
Date: 2005-11-09 13:03:58 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: RE-ASSIGNED TO fabian
Fabian, the script below would probably check the correct
behaviour for this bug. Could you please confirm this and
check it in? Thanks!
DELETE FROM history;
SELECT 1;
SELECT 'test';
SELECT * FROM history;
## Comment 2148
Date: 2005-11-09 13:20:40 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: REASSIGNED vzzzbx:
confirming, please check in yourself
(am not participating that much today)
## Comment 2149
Date: 2005-11-09 15:52:04 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: TEST ADDED / SUCCEED
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/history_table_escape.SF-1072089.s*
(Wouter was even less participating than me, so he won)
## Comment 2150
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1072089 at http://sourceforge.net/support/tracker.php?aid=1072089
| SQL: history table not updated consistently | https://api.github.com/repos/MonetDB/MonetDB/issues/396/comments | 0 | 2020-11-30T08:32:41Z | 2024-06-27T10:59:40Z | https://github.com/MonetDB/MonetDB/issues/396 | 753,271,306 | 396 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-23 06:22:13 +0100
From: @njnes
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2004-11-30 06:34:32 +0100
## Comment 2141
Date: 2004-11-23 18:22:13 +0100
From: @njnes
The test script src/gdk/qsort fails, indicating a
problem in the recent order/sorting changes. SQL tpch q
18 fails when testing with --debug=8 (or 10).
## Comment 2142
Date: 2004-11-30 18:34:06 +0100
From: @peterboncz
Logged In: YES
user_id=591107
fixed a recently introduced bug in sort_rev
bats that were sorted, were not reversely sorted when asked
because the already-sorted check was wrong.
this fixes both TPC-H q18 as well as some qsort tests in the
MonetDB test web
## Comment 2143
Date: 2005-11-09 12:50:07 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: NO TEST ADDED / ALREADY IN TESTWEB
## Comment 2144
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1071920 at http://sourceforge.net/support/tracker.php?aid=1071920
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.
| order/sort problems | https://api.github.com/repos/MonetDB/MonetDB/issues/395/comments | 0 | 2020-11-30T08:32:38Z | 2024-06-28T13:36:32Z | https://github.com/MonetDB/MonetDB/issues/395 | 753,271,276 | 395 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-21 08:38:55 +0100
From: Koen van de Sande <<shadowmorpher>>
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2007-04-17 03:46:28 +0200
## Comment 2131
Date: 2004-11-21 20:38:55 +0100
From: Koen van de Sande <<shadowmorpher>>
We have been trying to open a bi-directional connection
between two MonetDB instances under Windows, however
the commands from the TCP/IP module cause frequent hangs.
For example, the TCPIP2.* test included with the source
will hang/block frequently on the "open" command (but
it works sometimes after you reboot the computer).
However, the "export" command *always* causes blocking,
we've never been able to export a variable succesfully.
Has the TCP/IP module been tested under Windows, and is
it even supported there?
Will this script work correctly under Linux?
And note that this problem is not related to the open
printf bug, because if we remove the printf's then we
still hang...
SCRIPT
=====
module("tcpip", "alarm");
module("ascii_io");
proc get_mapi_port() : void {
if (monet_environment.exist("mapi_port")){
return int(monet_environment.find("mapi_port"));
} else {
return 55555;
}
}
VAR tcpip_port := get_mapi_port();
fork(listen(tcpip_port));
sleep(2);
printf("open...");
VAR c := open("localhost:"+str(tcpip_port));
printf(" it worked!\n");
var test1 := bat(oid, oid);
test1.import("k:/tcpip2.init.bat");
test1.kunique().count().print();
var test2 := bat(oid, oid);
test2.import("k:/tcpip2.init.bat");
printf("Loaded files, attempting RPC...");
we can get here if open didn't hang, but doing
anything here will fail
RPC:join will block, export will block and even
close(c); will block!!!
var result_bat := c.rpc("join(test1, %s)", test2);
this will send local_bat to the other side, join
it there with 'remote_bat', and send back the result.
result_bat.kunique().count().print();
close(c);
quit();
printf("Export...");
c.export(test1, "test1");
printf("ghe\n");
var test2 := import("test1", true);
test2.kunique().count().print();
close(c);
quit();
## Comment 2132
Date: 2004-11-23 11:43:02 +0100
From: @grobian
Logged In: YES
user_id=963970
MonetDB>module(tcpip);
MonetDB>var c := open("localhost:" + str(8080));
MonetDB>close(c);
MonetDB>!ERROR: socket_gets:
conn_waiter: lost connection, exiting.
MonetDB>var c := open("localhost:" + str(54123));
!ERROR: TCPopen(localhost:54123): could not connect with.
!OS: Connection refused
!ERROR: TCPopen: operation failed.
MonetDB>var c := open("localhost:" + str(45123));
MonetDB>conn_waiter: lost connection, exiting.
MonetDB>close(c);
!ERROR: TCPclose: operation failed.
MonetDB>fork(listen(55555));
MonetDB>var c := open("localhost:" + str(55555));
MonetDB>close(c);
MonetDB>!ERROR: socket_gets:
conn_waiter: lost connection, exiting.
conn_waiter: lost connection, exiting.
MonetDB>
It doesn't feel very groovy to me... :(
## Comment 2133
Date: 2004-11-23 17:36:01 +0100
From: Koen van de Sande <<shadowmorpher>>
Logged In: YES
user_id=270334
Hmmm... lets rewrite the initial report. After a Windows
system reboot, then the following script works (almost always):
module(tcpip);
fork(listen(25000));
var c:=open("localhost:25000");
var test:=bat(oid,oid);
test.insert(0@0,24@0);
c.export(test,"test");
var test2:=import("test");
print(test2);
close(c);
terminate(25000);
Now, if we close MonetDB by issueing a quit() command, and
then restart the SQL server, then either it will work again,
or not (most of the time, it won't). Changing the port
number maybe will allow for another server run to 'work'.
We're not sure how long the server will stay working/not
hanging running this script if we do not restart the server.
That will need another reboot...
## Comment 2134
Date: 2004-11-23 20:15:55 +0100
From: Koen van de Sande <<shadowmorpher>>
Logged In: YES
user_id=270334
I've activated TCP/IP debuging by setting gdk_debug to
65536... and the following script *will* cause MonetDB to
hang on my Windows 2000 PC if you run it multiple times in a
row (after 2-3 times, typically).
SCRIPT
======
module(tcpip);
fork(listen(23000));
var c:=open("localhost:23000");
var test:=bat(oid,oid);
test.insert(0@0,24@0);
c.export(test,"test");
var test2:=import("test");
close(c);
terminate(23000);
MONETDB WITH GDK_DEBUG=65536
============================
Monet Database Server V4.4.2
Copyright (c) 1993-2004, CWI. All rights reserved.
Compiled for i686-pc-win32/32bit; dynamically linked.
Visit http://monetdb.cwi.nl/ for further information.
MonetDB>module(tcpip);
module(tcpip);
MonetDB>fork(listen(23000));
fork(listen(23000));
MonetDB>var c:=open("localhost:23000");
var c:=open("localhost:23000");
Allocate connection record 0 (23000, -1)
Allocate connection record 1 (6404, 23000)
MonetDB>var test:=bat(oid,oid);
var test:=bat(oid,oid);
MonetDB>test.insert(0@0,24@0);
test.insert(0@0,24@0);
MonetDB>c.export(test,"test");
c.export(test,"test");
send:bat:test:1*8:oid@0+k:oid@4+k:0:0:0@0:0@0:0@0
:53
MonetDB>var test2:=import("test");
var test2:=import("test");
bat:test:1*8:oid@0+k:oid@4+k:0:0:0@0:0@0:0@0
MonetDB>close(c);
close(c);
Delete connection record 0 (23000 -1)
conn_waiter: lost connection, exiting.
Delete connection record 1 (6404 23000)
!ERROR: socket_gets:
conn_waiter: lost connection, exiting.
Delete connection record 0 (0 0)
MonetDB>terminate(23000);
terminate(23000);
!WARNING: TCPlisten(23000): stopped.
send:stop
:5
MonetDB>module(tcpip);
module(tcpip);
MonetDB>fork(listen(23000));
fork(listen(23000));
MonetDB>var c:=open("localhost:23000");
var c:=open("localhost:23000");
Allocate connection record 0 (23000, -1)
Allocate connection record 1 (6916, 23000)
MonetDB>var test:=bat(oid,oid);
var test:=bat(oid,oid);
MonetDB>test.insert(0@0,24@0);
test.insert(0@0,24@0);
MonetDB>c.export(test,"test");
c.export(test,"test");
send:bat:test:1*8:oid@0+k:oid@4+k:0:0:0@0:0@0:0@0
:53
MonetDB>var test2:=import("test");
var test2:=import("test");
bat:test:1*8:oid@0+k:oid@4+k:0:0:0@0:0@0:0@0
MonetDB>close(c);
close(c);
Delete connection record 0 (23000 -1)
(... MonetDB now hangs...)
## Comment 2135
Date: 2005-02-10 14:46:33 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
Fixed by identifying that both sides of a connection are the
same process and then not starting a second reader thread on
the same connection.
## Comment 2136
Date: 2005-11-09 12:48:15 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: TEST ADDED / PARTIALLY SUCCESSFUL
Test: tcpip_hangs_on_close.SF-1070569.stable.out
I don't know whether my monetdb-build returns the intended
output. (no output). I deleted some "connection lost"
messages from the stable output.
(and i just saw that Niels beat me with a similar bug...
Niels could you verify the correct output of this bug?)
(note: the test-scripts in src/modules/plain/Tests that test
the tcpip-module should be extended in that they test ALL
functions the module offers)
## Comment 2137
Date: 2005-11-14 15:20:52 +0100
From: @drstmane
Logged In: YES
user_id=572415
re-opened test, since test and/or stable output need to be
checked and possibly fixed;
see e.g.,
http://monetdb.cwi.nl/testing/projects/monetdb/Current/MonetDB/.mTests103/ntv.64.64.d-Fedora4/tests_BugDay_2005-11-09_4.9.3/tcpip_hangs_on_close.SF-1070569.out.00.html
## Comment 2138
Date: 2006-01-22 12:07:41 +0100
From: @drstmane
Logged In: YES
user_id=572415
Peter,
could you please check, whether this output is ok,
http://monetdb.cwi.nl/testing/projects/monetdb/Current/MonetDB/.mTests103/GNU.64.64.d-Fedora4/tests_BugDay_2005-11-09_4.9.3/tcpip_hangs_on_close.SF-1070569.out.00.html
and if not, possibly have a look at the bug?
Thanks!
Stefan
ps: This bug might be related
https://sourceforge.net/tracker/index.php?func=detail&aid=1042370&group_id=56967&atid=482468
## Comment 2139
Date: 2007-04-17 15:46:28 +0200
From: @peterboncz
Logged In: YES
user_id=591107
Originator: NO
output seems correct, closing this
## Comment 2140
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1070569 at http://sourceforge.net/support/tracker.php?aid=1070569
| TCP/IP module causes hangs under Windows | https://api.github.com/repos/MonetDB/MonetDB/issues/394/comments | 0 | 2020-11-30T08:32:34Z | 2024-06-27T10:59:38Z | https://github.com/MonetDB/MonetDB/issues/394 | 753,271,246 | 394 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-15 05:12:36 +0100
From: @sjoerdmullender
To: @sjoerdmullender
Version: -- development
Last updated: 2007-11-30 04:20:03 +0100
## Comment 2122
Date: 2004-11-15 17:12:36 +0100
From: @sjoerdmullender
The query SQLColumns uses to query the server contains
this:
cast(c."type" as smallint) as data_type,
cast(c."type" as varchar) as type_name,
This is obviously incorrect. The c."type" (from
sys.columns) cannot be cast both to smallint and
varchar (it is a varchar column).
The bug also occurs in 2.4.2.
## Comment 2123
Date: 2004-11-15 17:22:54 +0100
From: @grobian
Logged In: YES
user_id=963970
I assume the problem here is not the cast itself, but the
column usage.
JDBC:
String query =
"SELECT '" + cat + "' AS \"TABLE_CAT\",
\"schemas\".\"name\" AS \"TABLE_SCHEM\", " +
"\"tables\".\"name\" AS \"TABLE_NAME\",
\"columns\".\"name\" AS \"COLUMN_NAME\", " +
"cast(CASE \"columns\".\"type\" " +
"WHEN 'table' THEN " + Types.ARRAY + " " +
"WHEN 'boolean' THEN " + Types.BOOLEAN + " " +
"WHEN 'bool' THEN " + Types.BOOLEAN + " " +
"WHEN 'ubyte' THEN " + Types.CHAR + " " +
"WHEN 'char' THEN " + Types.CHAR + " " +
"WHEN 'character' THEN " + Types.CHAR + " " +
"WHEN 'varchar' THEN " + Types.VARCHAR + " " +
"WHEN 'text' THEN " + Types.LONGVARCHAR + " " +
"WHEN 'tinytext' THEN " + Types.LONGVARCHAR
+ " " +
"WHEN 'string' THEN " + Types.LONGVARCHAR +
" " +
"WHEN 'tinyint' THEN " + Types.TINYINT + " " +
"WHEN 'smallint' THEN " + Types.SMALLINT + " " +
"WHEN 'mediumint' THEN " + Types.INTEGER + " " +
"WHEN 'oid' THEN " + Types.OTHER + " " +
"WHEN 'int' THEN " + Types.INTEGER + " " +
"WHEN 'integer' THEN " + Types.INTEGER + " " +
"WHEN 'bigint' THEN " + Types.BIGINT + " " +
"WHEN 'number' THEN " + Types.INTEGER + " " +
"WHEN 'decimal' THEN " + Types.DECIMAL + " " +
"WHEN 'numeric' THEN " + Types.NUMERIC + " " +
"WHEN 'float' THEN " + Types.FLOAT + " " +
"WHEN 'double' THEN " + Types.DOUBLE + " " +
"WHEN 'real' THEN " + Types.DOUBLE + " " +
"WHEN 'month_interval' THEN " +
Types.INTEGER + " " +
"WHEN 'sec_interval' THEN " + Types.BIGINT +
" " +
"WHEN 'date' THEN " + Types.DATE + " " +
"WHEN 'time' THEN " + Types.TIME + " " +
"WHEN 'datetime' THEN " + Types.TIMESTAMP +
" " +
"WHEN 'timestamp' THEN " + Types.TIMESTAMP +
" " +
"WHEN 'blob' THEN " + Types.BLOB + " " +
"ELSE " + Types.OTHER + " " +
"END AS smallint) AS \"DATA_TYPE\", " +
"\"columns\".\"type\" AS \"TYPE_NAME\", " +
"\"columns\".\"type_digits\" AS \"COLUMN_SIZE\", " +
"\"columns\".\"type_scale\" AS
\"DECIMAL_DIGITS\", 0 AS \"BUFFER_LENGTH\", " +
"10 AS \"NUM_PREC_RADIX\", " +
"cast(CASE \"null\" " +
"WHEN true THEN " +
ResultSetMetaData.columnNullable + " " +
"WHEN false THEN " +
ResultSetMetaData.columnNoNulls + " " +
"END AS int) AS \"NULLABLE\", cast(null AS
varchar) AS \"REMARKS\", " +
"\"columns\".\"default\" AS \"COLUMN_DEF\", 0 AS
\"SQL_DATA_TYPE\", " +
"0 AS \"SQL_DATETIME_SUB\", 0 AS
\"CHAR_OCTET_LENGTH\", " +
"\"columns\".\"number\" + 1 AS
\"ORDINAL_POSITION\", " +
"cast(null AS varchar) AS \"SCOPE_CATALOG\", " +
"cast(null AS varchar) AS \"SCOPE_SCHEMA\", " +
"cast(null AS varchar) AS \"SCOPE_TABLE\", " +
"cast(" +
((MonetDriver)driver).getJavaType("other") + " AS smallint)
AS \"SOURCE_DATA_TYPE\", " +
"CASE \"null\" " +
"WHEN true THEN CAST ('YES' AS varchar) " +
"WHEN false THEN CAST ('NO' AS varchar) " +
"END AS \"IS_NULLABLE\" " +
"FROM \"sys\".\"columns\" AS \"columns\", " +
"\"sys\".\"tables\" AS \"tables\", " +
"\"sys\".\"schemas\" AS \"schemas\" " +
"WHERE \"columns\".\"table_id\" =
\"tables\".\"id\" " +
"AND \"tables\".\"schema_id\" =
\"schemas\".\"id\" ";
It's only such a small query...
the above stated problem is solved with the CASE statement
to get the right DATA_TYPE, while the real value is used as
TYPE_NAME
## Comment 2124
Date: 2004-11-15 21:33:54 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
The value in columns.type is not correct for the type_name
column of the result. The former is the server's internal
name, whereas the latter is the SQL name. The former could
be something like "month_interval" or "int" when the SQL
name should be "interval month" and "integer" respectively.
Anyway, the bug is fixed in ODBC.
## Comment 2125
Date: 2005-11-09 12:14:41 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: RE-ASSIGNED TO sjoerd
(i don't understand this bug, but appearantly something went
wrong in the ODBC driver)
## Comment 2126
Date: 2005-11-09 12:40:31 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: NO TEST / POSTPONED
We need tests for *all* ODBC functions.
(The report was a reminder that the implementation of
SQLColumns was not correct.)
## Comment 2127
Date: 2006-11-10 04:20:01 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 2128
Date: 2006-11-29 10:04:24 +0100
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
over-ruling "sf-robot":
re-set to "Pending" since adding a proper test has neither been done nor finally been discarded, yet.
User (submitter) & developer (assignee),
please check, again.
## Comment 2129
Date: 2007-11-30 04:20:03 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
Originator: NO
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 2130
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1066751 at http://sourceforge.net/support/tracker.php?aid=1066751
| SQLColumns uses incorrect query | https://api.github.com/repos/MonetDB/MonetDB/issues/393/comments | 0 | 2020-11-30T08:32:31Z | 2024-06-27T10:59:37Z | https://github.com/MonetDB/MonetDB/issues/393 | 753,271,202 | 393 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-12 09:57:07 +0100
From: Jan Rittinger <<jan.rittinger>>
To: @njnes
Version: -- development
Last updated: 2004-11-13 04:36:55 +0100
## Comment 2118
Date: 2004-11-12 21:57:07 +0100
From: Jan Rittinger <<jan.rittinger>>
I have no clue where something wrong happens but
adding the two semicolons in
/Linux/src/monet/monet_parse.y did fix it
## Comment 2119
Date: 2004-11-13 16:36:55 +0100
From: @njnes
Logged In: YES
user_id=43556
added the semicolons.
## Comment 2120
Date: 2005-11-09 12:05:23 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: NO TEST ADDED / bug
cannot be reproduced
## Comment 2121
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1065430 at http://sourceforge.net/support/tracker.php?aid=1065430
| monet_parse.y contains to less semicolons | https://api.github.com/repos/MonetDB/MonetDB/issues/392/comments | 0 | 2020-11-30T08:32:28Z | 2024-06-27T10:59:36Z | https://github.com/MonetDB/MonetDB/issues/392 | 753,271,173 | 392 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-12 07:54:13 +0100
From: Koen van de Sande <<shadowmorpher>>
To: @njnes
Version: -- development
Last updated: 2004-11-16 02:05:40 +0100
## Comment 2113
Date: 2004-11-12 19:54:13 +0100
From: Koen van de Sande <<shadowmorpher>>
Under Windows, I cannot connect to the server in MIL
mode using either Mknife (MILdemo) or MapiClient.bat.
Mknife reports "connection failed", MapiClient says it
can't connect to port 50000.
I've started the server both through the start menu
entry and through Mserver.bat and MSQLserver.bat, but
this makes no difference.
In other 4.4 versions I could get a MIL> prompt if I
started MapiClient.bat (IIRC), right?
## Comment 2114
Date: 2004-11-13 09:05:06 +0100
From: @grobian
Logged In: YES
user_id=963970
This is indeed since 4.4.2.
If you want to start a mapi listener, I think you have to
change the start menu item a bit. I think the item now is
something like Mserver.bat
%MONETDB%\lib\MonetDB\sql_server.mil. To start mapi, it
should be Mserver.bat %MONETDB%\lib\MonetDB\mapi.mil.
I have no windows installation to test this against, so this
remark should be considered as a possible solution...
## Comment 2115
Date: 2004-11-13 16:33:58 +0100
From: @njnes
Logged In: YES
user_id=43556
Are you able to connect in 'sql' mode? Starting with the
2.4.2 sql release it
is for security/stability reasons no longer possible to
connect in both
sql and mil mode. If you know what you are doing ie mix
backends, you
could re-enable this by adding the line
mapi_listen(int(monet_environment.find("mapi_port")), 5,
"").fork();
in your Mserver startup scripts.
## Comment 2116
Date: 2005-11-09 12:03:17 +0100
From: Wouter Alink <<alink>>
Logged In: YES
user_id=621590
BugDay_2005-11-09, vzzzbx: NO TEST ADDED /
Documentation issue (for Windows)
## Comment 2117
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1065373 at http://sourceforge.net/support/tracker.php?aid=1065373
| [Win/4.4.2] Cannot connect to server in MIL mode | https://api.github.com/repos/MonetDB/MonetDB/issues/391/comments | 0 | 2020-11-30T08:32:26Z | 2024-06-28T13:36:31Z | https://github.com/MonetDB/MonetDB/issues/391 | 753,271,145 | 391 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-11-09 02:26:19 +0100
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-11-11 12:38:54 +0100
## Comment 2109
Date: 2004-11-09 14:26:19 +0100
From: @grobian
Sietse Dijkstra stumbled upon the following bug in
2.4.2, which is also in 2.5.0:
CREATE TABLE tempEntries (
url varchar(255),
psize int,
);
CREATE TABLE entries (
id int,
url varchar(255) not null unique,
doc date,
dom date,
psize int,
ranking int,
primary key (id)
);
CREATE TABLE links (
entryid int,
url varchar(255) not null,
primary key (entryid, url)
);
% MapiClient -hpictor
sql>select locallinks.url, allentries.psize from
more> links locallinks inner join (
more> select url, psize from entries union select
url, psize from
more>tempEntries
more> ) as allentries on locallinks.url=allentries.url;
MAPI = monetdb@pictor:45123
ACTION= read_line
QUERY = select locallinks.url, allentries.psize from
ERROR = Connection terminated
The problematic query:
select locallinks.url, allentries.psize from
links locallinks inner join (
select url, psize from entries union select url,
psize from tempEntries
) as allentries on locallinks.url=allentries.url;
## Comment 2110
Date: 2004-11-11 12:38:54 +0100
From: @njnes
Logged In: YES
user_id=43556
alias tables names where not allways properly used
## Comment 2111
Date: 2005-11-09 15:42:36 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/alias_union_join.SF-1063089.s*
## Comment 2112
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1063089 at http://sourceforge.net/support/tracker.php?aid=1063089
| SQL: meltdown on subquery with UNION and JOIN | https://api.github.com/repos/MonetDB/MonetDB/issues/390/comments | 0 | 2020-11-30T08:32:23Z | 2024-06-27T10:59:34Z | https://github.com/MonetDB/MonetDB/issues/390 | 753,271,120 | 390 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-27 09:43:45 +0200
From: @njnes
To: @sjoerdmullender
Version: -- development
Last updated: 2005-05-29 10:17:16 +0200
## Comment 2102
Date: 2004-10-27 21:43:45 +0200
From: @njnes
When adding some debug statements (printf's) to the
decimal.mx code, the Mtest run of the decimal test
hangs with a python running
at 100% cpu usage.
## Comment 2103
Date: 2004-10-27 22:45:03 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Where were the printf's inserted (to make trying to
reproduce the problem a no-brainer).
## Comment 2104
Date: 2004-10-27 22:55:30 +0200
From: @drstmane
Logged In: YES
user_id=572415
sounds like a similar problem as Marcin had recently --- pipes blocking
each other probably due to overflowing buffers --- but that one solved it
self resp. disappeared "misteriously" ...
## Comment 2105
Date: 2004-10-27 22:57:08 +0200
From: @njnes
Logged In: YES
user_id=43556
printfs where added in the decimal_normalize code (4 printfs)
## Comment 2106
Date: 2005-05-29 10:17:16 +0200
From: @mlkersten
Logged In: YES
user_id=490798
This bug has not been seen since, which implies we call it
extinguished in the process of improving the code base.
## Comment 2107
Date: 2005-11-09 13:37:08 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: NO TEST /
COMPILATION+OUTOFDATE+TESTINGTHETESTER
## Comment 2108
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1055593 at http://sourceforge.net/support/tracker.php?aid=1055593
| Mtest hang | https://api.github.com/repos/MonetDB/MonetDB/issues/389/comments | 0 | 2020-11-30T08:32:20Z | 2024-06-27T10:59:33Z | https://github.com/MonetDB/MonetDB/issues/389 | 753,271,084 | 389 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-26 12:35:03 +0200
From: Jens Teubner <<teubner>>
To: @njnes
Version: -- development
Last updated: 2004-10-27 12:25:31 +0200
## Comment 2098
Date: 2004-10-26 12:35:03 +0200
From: Jens Teubner <<teubner>>
The following example demonstrates my problem:
var a := new (void, BAT).seqbase (0@0);
var b := new (void, BAT);
var c := new (void, int).seqbase (0@0);
c.insert (nil, 42);
c.insert (nil, 47);
b.insert (nil, c);
a.append (b);
`a' is supposed to collect BATs. When adding BATs to `a', I want to do that
in bulk. I thus create another BAT collection `b', in which I insert (for the
example) a single BAT `c'. This new collection `b', I add to `a' using
append(). The head of `b' is void(nil). I'd thus expect the result to have a
void head, with a single BUN that has `c' in its tail:
void | BAT
-----+--------
0@0 | <c>
An `a.print()', however, just produces error messages:
mil>a.print ();
MAPI = anonymous@localhost:50000
QUERY = a.print ();
ERROR = !WARNING: BBPname: range error 980714
-----------------
h tmp_32 name
void BAT type
-----------------
MAPI = anonymous@localhost:50000
QUERY = a.print ();
ERROR = !WARNING: BBPname: range error 980714
[ 0@0, nil ]
Using MonetDB 4.4.0 here (CVS version from a few days ago), on Gentoo
Linux i386.
## Comment 2099
Date: 2004-10-27 00:25:31 +0200
From: @njnes
Logged In: YES
user_id=43556
One problem with this bug is which command gets called?
This was solve by shuffling the value function before the
bat, bat function.
The append value command also had a bug, it inserted
the bat pointer instead of the bad id.
## Comment 2100
Date: 2005-11-09 15:34:49 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.9.3/Tests/append_bat_bat.SF-1054499.*
## Comment 2101
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1054499 at http://sourceforge.net/support/tracker.php?aid=1054499
| append (BAT[void, BAT], BAT[void, BAT]) buggy | https://api.github.com/repos/MonetDB/MonetDB/issues/388/comments | 0 | 2020-11-30T08:32:18Z | 2024-06-27T10:59:32Z | https://github.com/MonetDB/MonetDB/issues/388 | 753,271,069 | 388 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-26 11:45:11 +0200
From: @sjoerdmullender
To: @sjoerdmullender
Version: -- development
Last updated: 2004-10-27 05:15:54 +0200
## Comment 2094
Date: 2004-10-26 11:45:11 +0200
From: @sjoerdmullender
Agustin Shapira wrote:
I am trying to get our Proximity code to run with the
new MIL parser from Monet 4.4.0, and I am encountering
a few problems. In particular, the new version forces
you to specify the return types of a PROC, along with
the types of its arguments. I used to have a PROC
defined as
PROC get_nst_children (root, result) { ... }
which now has to be defined as
PROC get_nst_children (bat root, bat[void,str] result)
: void { ... }
The funny thing is that I am getting the following
error when I call it:
!ERROR: interpret: no matching MIL operator to
'[get_nst_children](BAT[oid,str], BAT[void,str])'.
!MAYBE YOU MEAN:
! get_nst_children(BAT[any,any], BAT[void,str]) :
void
How can this be? It's telling me that it needs a first
parameter of type BAT[any,any], and that a BAT[oid,str]
is not good enough? It does work with a BAT[oid,oid],
by the way! Does this make any sense?
I tried this on 4.3.18a, and found the same problem,
which leads me to believe that I am doing something wrong.
In any case, I am attaching a sample script, which
should be run against an empty database. It creates an
NST (our multi-column tables), defines a few PROCs, and
then calls them, generating the error. Can you see
anything strange in the code?
Back to me:
I attached Agustin's script after adding a few missing
()s. It fails in the same way with both the HEAD and
the 4.4 branch.
## Comment 2095
Date: 2004-10-27 17:15:54 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
The recursive call of [get_nst_children](...) is invalid.
It is not possible to multiplex over a function that returns
void.
The error message is misleading to say the least, however it
is correct that you get an error.
## Comment 2096
Date: 2005-11-09 13:34:01 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: NO TEST / INVALID
(bug was resolved as being invalid, hence no test)
## Comment 2097
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1054476 at http://sourceforge.net/support/tracker.php?aid=1054476
| MIL parser doesn't recognize matching operators | https://api.github.com/repos/MonetDB/MonetDB/issues/387/comments | 0 | 2020-11-30T08:32:15Z | 2024-06-28T13:36:30Z | https://github.com/MonetDB/MonetDB/issues/387 | 753,271,047 | 387 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-20 10:07:13 +0200
From: D.R. Boxhoorn <<drb>>
To: @njnes
Version: -- development
Last updated: 2004-10-27 09:02:51 +0200
## Comment 2083
Date: 2004-10-20 10:07:13 +0200
From: D.R. Boxhoorn <<drb>>
A large number of inserts into a simple table causes a
segmentation fault of the `Mserver' after some time.
Commits are performed every 100,000 inserts.
`sql_server' cannot be restarted afterwards
This behaviour can be reproduced with at least one dataset.
The sql causing the problem looks like
create table demo (a float, b float, c float, d float);
commit;
100,000 time the next line
insert into demo values (1.0, 2.0, 3.0, 4.0);
commit;
repeat the previous lines a few hundred (ten?) times
until the segmentation fault occurs
ERRORS
% Mserver
Monet Database Server V4.4.0
Copyright (c) 1993-2004, CWI. All rights reserved.
Compiled for i686-pc-linux-gnu/32bit; dynamically linked.
Visit http://monetdb.cwi.nl/ for further information.
MonetDB>module(sql_server);
MonetDB>!ERROR: GDKload: read error: name=13/751,
ext=buns, n=800000
!ERROR: BATcopy: BAT required.
!ERROR: BATname: BAT required.
!ERROR: BATrename: BAT argument missing
!ERROR: BATSIGcrash: Mserver internal error
(Segmentation fault), please restart.
!ERROR: (One potential cause could be that your disk
might be full...)
% Mserver
Monet Database Server V4.4.0
Copyright (c) 1993-2004, CWI. All rights reserved.
Compiled for i686-pc-linux-gnu/32bit; dynamically linked.
Visit http://monetdb.cwi.nl/ for further information.
MonetDB>module(sql_server);
!ERROR: Incompatible database version 000000, this
server supports version 010100
The error on the client-side reads
[ ... previous debug output not copied here ... ]
4 querytype
affected table_name
rows name
int type
0 length
1 tuplecount
5457532 id
4 querytype
affected table_name
rows name
int type
0 length
1 tuplecount
5457533 id
[ 1 ]
MAPI = monetdb@localhost:45123
ACTION= read_line
QUERY = insert into demo values
(8.67017259259,-67.9165138889,18.5,17.5);
ERROR = Connection terminated
VERSIONS
MonetDB 4.4.0
SQL 2.4.0
% cat /etc/redhat-release
Red Hat Enterprise Linux WS release 3 (Taroon Update 3)
% uname -rvmpi
2.4.21-20.ELsmp 1 SMP Wed Aug 18 20:46:40 EDT 2004
i686 i686 i386
% gcc --version
gcc (GCC) 3.2.3 20030502 (Red Hat Linux 3.2.3-42)
% lex --version
lex version 2.5.4
% bison --version
bison (GNU Bison) 1.875c
% autoconf --version
autoconf (GNU Autoconf) 2.57
% automake --version
automake (GNU automake) 1.6.3
% libtool --version
ltmain.sh (GNU libtool) 1.4.3 (1.922.2.111 2002/10/23
02:54:36)
## Comment 2084
Date: 2004-10-20 17:12:48 +0200
From: D.R. Boxhoorn <<drb>>
Logged In: YES
user_id=66299
Using the above example a segmentation fault occurs after
about 9,000,000 inserts. The following backtrace shows where
it occurs.
(gdb) bt
0 0x002019a1 in __read_nocancel () from /lib/tls/libc.so.6
1 0x0019e9c8 in _IO_file_read_internal () from
/lib/tls/libc.so.6
2 0x0019dcde in _IO_new_file_underflow () from
/lib/tls/libc.so.6
3 0x0019fdad in __underflow () from /lib/tls/libc.so.6
4 0x0019ee78 in _IO_file_xsgetn_internal () from
/lib/tls/libc.so.6
5 0x001a04e8 in _IO_sgetn_internal () from /lib/tls/libc.so.6
6 0x001935dd in fread () from /lib/tls/libc.so.6
7 0x00b3ffc9 in file_read (s=0x9043fc8, buf=0x9087830,
elmsize=1, cnt=1)
at stream.mx:535
8 0x00b3fc1c in stream_read (s=0x9043fc8, buf=0x9087830,
elmsize=1, cnt=1)
at stream.mx:374
9 0x00c860b9 in readClient (c=0xce6060) at monet_client.mx:528
10 0x00c867b8 in serveClient (stk=0, lt=0x91dd580,
res=0xbfffa270)
at monet_client.mx:726
11 0x00c6b713 in interpret (stk=0, lt=0x91dd580,
res=0xbfffa270)
at monet_interpreter.mx:676
12 0x00c82aed in handleRequest (t=0x7bdde0, q=0xa473130,
res=0xbfffa270)
at monet_queue.mx:499
13 0x00c82f09 in doRequest (t=0x7bdde0, preference=0x0) at
monet_queue.mx:523
14 0x00cc8c70 in monetInterpreter (status=0xbfffa2d8) at
monet_process.mx:109
15 0x08049da4 in main (argc=1, av=0xbfffc394) at Mserver.mx:426
## Comment 2085
Date: 2004-10-20 19:27:20 +0200
From: @grobian
Logged In: YES
user_id=963970
I'm trying to reproduce this problem using JDBC, however,
I'm still at 1460000 :/
It is really slow due to the commit after each insert.
Would you mind giving a little detail about your platform,
like cpu architecture and speed?
I'll let it run, with this speed I'll be at 9000000 in a day
or so. Is the commit really required after each statement?
Instead of an implicit commit I use the auto-commit feature
of the server.
## Comment 2086
Date: 2004-10-20 22:53:39 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Bugs of this kind are not acceptable and require high
priority action. Segmation faults are prirority >8
Fabian thx for already starting the analysis
## Comment 2087
Date: 2004-10-20 23:16:28 +0200
From: @drstmane
Logged In: YES
user_id=572415
Fabian,
if I understand the original bug report correctly, commits
are/were done only once every 100,000 inserts, not after
each insert.
This should speed-up your analysis significantly...
Stefan
## Comment 2088
Date: 2004-10-20 23:41:43 +0200
From: D.R. Boxhoorn <<drb>>
Logged In: YES
user_id=66299
Fabian,
Yes, you should commit after a reasonable amount of inserts.
With a commit every 100,000th insert it takes less than an
hour on my mediocre pc.
PS. My apologies for reading the "acceptable use policy" of
the bugtracker on the MonetDB website _after_ reporting this
bug.
## Comment 2089
Date: 2004-10-21 09:44:54 +0200
From: @grobian
Logged In: YES
user_id=963970
Not that there was any doubt, but indeed, I found my
computer in a quiet state this morning:
8050000
Exception in thread "main" java.sql.SQLException: End of
stream reached (Mserver still alive?)
So this bug is confirmed and reproducable.
Regarding the commit intervals I assume this problem is
related to the amount of data in the bats, not the commit in
any way. Will try to use one transaction to see what happens.
## Comment 2090
Date: 2004-10-21 11:26:21 +0200
From: D.R. Boxhoorn <<drb>>
Logged In: YES
user_id=66299
Using the above example a segmentation fault occurs after
about 9,000,000 inserts. The following backtrace shows where
it occurs.
(gdb) bt
0 0x002019a1 in __read_nocancel () from /lib/tls/libc.so.6
1 0x0019e9c8 in _IO_file_read_internal () from
/lib/tls/libc.so.6
2 0x0019dcde in _IO_new_file_underflow () from
/lib/tls/libc.so.6
3 0x0019fdad in __underflow () from /lib/tls/libc.so.6
4 0x0019ee78 in _IO_file_xsgetn_internal () from
/lib/tls/libc.so.6
5 0x001a04e8 in _IO_sgetn_internal () from /lib/tls/libc.so.6
6 0x001935dd in fread () from /lib/tls/libc.so.6
7 0x00b3ffc9 in file_read (s=0x9043fc8, buf=0x9087830,
elmsize=1, cnt=1)
at stream.mx:535
8 0x00b3fc1c in stream_read (s=0x9043fc8, buf=0x9087830,
elmsize=1, cnt=1)
at stream.mx:374
9 0x00c860b9 in readClient (c=0xce6060) at monet_client.mx:528
10 0x00c867b8 in serveClient (stk=0, lt=0x91dd580,
res=0xbfffa270)
at monet_client.mx:726
11 0x00c6b713 in interpret (stk=0, lt=0x91dd580,
res=0xbfffa270)
at monet_interpreter.mx:676
12 0x00c82aed in handleRequest (t=0x7bdde0, q=0xa473130,
res=0xbfffa270)
at monet_queue.mx:499
13 0x00c82f09 in doRequest (t=0x7bdde0, preference=0x0) at
monet_queue.mx:523
14 0x00cc8c70 in monetInterpreter (status=0xbfffa2d8) at
monet_process.mx:109
15 0x08049da4 in main (argc=1, av=0xbfffc394) at Mserver.mx:426
## Comment 2091
Date: 2004-10-27 09:02:51 +0200
From: @njnes
Logged In: YES
user_id=43556
This bug is fixed in the current cvs versions both 4.5 and
4.4. It was
propably caused by a bat leak which slipped into the
released SQL
version. Therefor we wil bring out a bug fix release soon.
## Comment 2092
Date: 2005-11-09 13:32:47 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: NO TEST / TOO LARGE
(perhaps a program can test it efficiently?)
## Comment 2093
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1050539 at http://sourceforge.net/support/tracker.php?aid=1050539
| Mserver internal error after large number of INSERTs | https://api.github.com/repos/MonetDB/MonetDB/issues/386/comments | 0 | 2020-11-30T08:32:12Z | 2024-06-27T10:59:31Z | https://github.com/MonetDB/MonetDB/issues/386 | 753,271,016 | 386 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-19 02:29:09 +0200
From: Jan Rittinger <<jan.rittinger>>
To: @sjoerdmullender
Version: -- development
Last updated: 2004-10-19 04:24:25 +0200
## Comment 2074
Date: 2004-10-19 14:29:09 +0200
From: Jan Rittinger <<jan.rittinger>>
rittinge@triton04:/net/privstud/dbis/rittinge/work/MonetDB>
./bootstrap
automake 1.6.3 is 1.5 or newer. Good.
autoconf 2.53 is 2.53 or newer. Good.
libtool 1.4.2 is 1.4 or newer. Good.
Python 2.2.1 is 2.0.0 or newer. Good.
Traceback (most recent call last):
File "src/utils/autogen/autogen.py", line 156, in ?
(InstallList, DocList, OutList) = main(topdir,
topdir, automake, [])
File "src/utils/autogen/autogen.py", line 132, in main
codegen(p.curvar, cwd, topdir, incdirsmap)
File "src/utils/autogen/codegen.py", line 700, in codegen
do_deps(targets,deps,includes,incmap,cwd,incdirsmap)
File "src/utils/autogen/codegen.py", line 415, in do_deps
buildincs = shelve.open( buildincsfile, "c")
File "/usr/lib/python2.2/shelve.py", line 158, in open
return DbfilenameShelf(filename, flag)
File "/usr/lib/python2.2/shelve.py", line 148, in
__init__
Shelf.__init__(self, anydbm.open(filename, flag))
File "/usr/lib/python2.2/anydbm.py", line 86, in open
return mod.open(file, flag, mode)
dbm.error: (37, 'No locks available')
cat: acout.in: No such file or directory
Remember to add `AC_PROG_LIBTOOL' to `configure.in'.
You should add the contents of
`/usr/share/aclocal/libtool.m4' to `aclocal.m4'.
Putting files in AC_CONFIG_AUX_DIR, `conf'.
automake/aclocal 1.6.3 is older than 1.7.
Patching aclocal.m4 for Intel compiler on Linux (icc/ecc).
patching file aclocal.m4
Hunk 1 succeeded at 3270 (offset 728 lines).
patching file aclocal.m4
Hunk 1 succeeded at 3377 (offset 2193 lines).
Hunk 2 succeeded at 4644 (offset 2200 lines).
Hunk 3 succeeded at 4664 (offset 2200 lines).
autoheader: `conf/config.h.in' is created
configure.in:283: no files mentioned in `0'
## Comment 2075
Date: 2004-10-19 14:33:13 +0200
From: @drstmane
Logged In: YES
user_id=572415
IMHO bootstrap does works fine on NFS mounted directories
(at least, I can run bootstrap in my home on machines that
mount my home via nfs...) --- apparently, the nfs mount
needs to support file locking, though ...
## Comment 2076
Date: 2004-10-19 14:36:02 +0200
From: @grobian
Logged In: YES
user_id=963970
Ah... yeah I recall that... I used not to have nfslockd
running when I got the same errors. I usually use NFS
shares as well for bootstrapping, compiling, etc.
## Comment 2077
Date: 2004-10-19 15:16:46 +0200
From: Jan Rittinger <<jan.rittinger>>
Logged In: YES
user_id=993208
locking is (and was) activated.
@Stefan - you can test it yourself on
titan03.inf.uni-konstanz.de
## Comment 2078
Date: 2004-10-19 15:35:39 +0200
From: @drstmane
Logged In: YES
user_id=572415
Jan,
indeed, in Konstanz on titan03, bootstrap does not work in
my home, which is nfs mounted from titan01 (while it works
fine in /local_tmp/ ...).
Here at CWI, I have no problems with bootstrapping in my
linux-nfs-served home on both linux- & solaris- nfs clients...
I guess, it's a local problem in Konstanz --- and the
(pyhton) error message suggests that it is lock related ...
Unfortunately, I cannot see/say more for now ... sorry ...
## Comment 2079
Date: 2004-10-19 16:24:24 +0200
From: @njnes
Logged In: YES
user_id=43556
we rely on pythons anydb module which requires locks. It
seems your
nfs server doesn't support locks.
## Comment 2080
Date: 2004-10-19 17:41:34 +0200
From: Jens Teubner <<teubner>>
Logged In: YES
user_id=731390
Well, we do have locking enabled here. (Everybody here is reading email over
NFS. Locking does work here.)
But the man pages reveal the problem. From flock(2): ``flock(2) does not lock
files over NFS. Use fcntl(2) instead: that does work over NFS...'' Python's
anydbm module uses flock() to lock its files.
But the anydbm developers already seem to have fixed the problem. On our new
SuSE 9.2 machines, ./bootstrap runs (almost) fine. Here the Python library uses
fcntl() which succeeds.
So Jan, maybe you just pick one of the SuSE 9.2 machines.
And the Monet people are now at least aware of problems that may occur when
working with directories shared over NFS. SuSE 9.0 btw. also has the problem
fixed in anydbm, but has a similar problem in autoconf: http://lists.gnu.org/
archive/html/autoconf/2004-06/msg00087.html (I could reproduce this.).
## Comment 2081
Date: 2005-11-09 13:31:29 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: NO TEST / COMPILATION
## Comment 2082
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1049990 at http://sourceforge.net/support/tracker.php?aid=1049990
| bootstrap doesn't work in nfs mounted directories | https://api.github.com/repos/MonetDB/MonetDB/issues/385/comments | 0 | 2020-11-30T08:32:09Z | 2024-06-28T07:36:36Z | https://github.com/MonetDB/MonetDB/issues/385 | 753,270,999 | 385 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-19 12:51:02 +0200
From: Alex van Ballegooij <<bally>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2004-10-27 12:13:12 +0200
## Comment 2067
Date: 2004-10-19 12:51:02 +0200
From: Alex van Ballegooij <<bally>>
The TCP/IP module works fine on 32bit linux,
but fails in the 64bit version when returning a bat.
(returning something "simple" such as an int works fine...)
---
server side:
module(tcpip,radix);
proc f():bat[void,dbl] = {
var a := uniform(100000).[dbl]().copy();
printf("done...\n");
RETURN a;
}
fork(listen());
client side:
VAR id := open(localhost);
VAR result := id.rpc("f();");
result (given on "server" side, the "client" just ahngs
while waiting for a result that never arrives):
Mserver:
/data/home/alexb/local/Monet/Monet/Development/MonetDB/src/gdk/gdk_bbp.mx:1261:
BBPdescriptor: Assertion
`BBP[(((int)(i))>0?((int)(i)):-((int)(i)))].refs' failed.
./start_slave.sh: line 5: 22190 Aborted
Mserver --config monet.conf --dbname=`hostname`
Database/general.mil
/data/home/alexb/jdim_experiments/slave.mil
## Comment 2068
Date: 2004-10-19 13:36:35 +0200
From: Alex van Ballegooij <<bally>>
Logged In: YES
user_id=594789
The same tests seem to work fine with 4.4.0 (release version).
It turns out we tested on 32bit with 4.4.0,
so the test may have failed for reasons other
than 32/64bit differences (which was the assumption
at time of posting).
## Comment 2069
Date: 2004-10-19 14:30:40 +0200
From: @drstmane
Logged In: YES
user_id=572415
Just for info:
In case tests were run with up-to-date (i.e., 18/10/2004)
versions of 4.4.0 & 4.4.1, respectively, the minor (textual)
differences between modules tcpip itself in both branches
(see below) suggest, that the problem is located somewhere
else --- proprobabaly in code that has changed only in the
main trunk since 4.4.0 was released on 30.09.2004.
Cetero Censeo:
It would be a nice "exercise" to provide a "real" test
script for this bug report, and check it in both the 4.4.0
branch and the main trunk ... ;-))
========
14:24:00 manegold@draco:~/Monet/4.4/MonetDB $ cvs diff -0
-rMonetDB_4-4-0 -rHEAD src/modules/plain/tcpip.mx
Index: src/modules/plain/tcpip.mx
===================================================================
RCS file: /cvsroot/monetdb/MonetDB/src/modules/plain/tcpip.mx,v
retrieving revision 1.49.2.3
retrieving revision 1.53
diff -u -0 -r1.49.2.3 -r1.53
--- src/modules/plain/tcpip.mx 10 Oct 2004 20:10:58 -0000
1.49.2.3
+++ src/modules/plain/tcpip.mx 15 Oct 2004 12:50:39 -0000
1.53
@@ -46 +46 @@
-are substitued.
+are substituted.
@@ -285 +285 @@
-(without {\tt <user>_} perappended) of a type.
+(without {\tt <user>_} prepended) of a type.
@@ -512 +512 @@
-The connection afterwards is bidirectional. This asymmetry
on how
+Afterwards the connection is bidirectional. This asymmetry
on how
@@ -782 +782 @@
-The errormessage is stored in the 'errbuf' field of the
cache item.
+The error message is stored in the 'errbuf' field of the
cache item.
@@ -1561 +1561 @@
-availabe as a complex variant of the 'mil' command (for
remote MIL
+available as a complex variant of the 'mil' command (for
remote MIL
14:24:18 manegold@draco:~/Monet/4.4/MonetDB $
========
## Comment 2070
Date: 2004-10-19 14:34:59 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
I get the same error on 32-bit Linux using the development
version (with a couple of changes to the script which I have
attached).
## Comment 2071
Date: 2004-10-27 00:13:12 +0200
From: @njnes
Logged In: YES
user_id=43556
the tcpip module was never updated for the new bat reference
model. Once
updated to the new model the bug was gone.
## Comment 2072
Date: 2005-11-09 15:02:52 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.9.3/Tests/tcpip_rpc_crash_bat.SF-1049942.*
## Comment 2073
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1049942 at http://sourceforge.net/support/tracker.php?aid=1049942
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.
| TCP/IP module RPC crashes server when returning a bat | https://api.github.com/repos/MonetDB/MonetDB/issues/384/comments | 0 | 2020-11-30T08:32:06Z | 2024-06-27T10:59:28Z | https://github.com/MonetDB/MonetDB/issues/384 | 753,270,966 | 384 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-17 04:01:08 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-11-08 09:24:06 +0100
## Comment 2063
Date: 2004-10-17 16:01:08 +0200
From: @grobian
monetdb-> insert into test values (1234567890);
1 affected row
monetdb-> insert into test values (9876543210);
1 affected row
monetdb-> select * from test;
+----------+
| val |
+==========+
| 1 |
| 1 |
| 2 |
| <NULL> |
| 1234567890 |
| 0 |
+----------+
6 rows
I'd expect and error, telling 9876543210 does not fit
in the underlying storage type, just like I cannot
store a too large string.
## Comment 2064
Date: 2004-11-08 09:24:06 +0100
From: @njnes
Logged In: YES
user_id=43556
Added more dynamic check functions, which give an error on
overflow
## Comment 2065
Date: 2005-11-09 13:09:46 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/integer_overflow.SF-1048671.sql
(commit pending...)
## Comment 2066
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1048671 at http://sourceforge.net/support/tracker.php?aid=1048671
| SQL: silent convert to 0 of overflow value? | https://api.github.com/repos/MonetDB/MonetDB/issues/383/comments | 0 | 2020-11-30T08:32:03Z | 2024-06-27T10:59:28Z | https://github.com/MonetDB/MonetDB/issues/383 | 753,270,931 | 383 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-14 11:00:29 +0200
From: Jens Teubner <<teubner>>
To: @njnes
Version: -- development
Last updated: 2004-10-16 09:11:07 +0200
## Comment 2058
Date: 2004-10-14 11:00:29 +0200
From: Jens Teubner <<teubner>>
Bootstrapping Pathfinder returns some messages like
loader/Makefile.am:10: bad macro name `pf-shreddir'
loader/Makefile.am:11: bad macro name `pf-shred_PROGRAMS'
...
Reason is that the hyphen is not allowed in automake macro names.
autogen correctly replaces the hyphen in the _SOURCES macro, but not in
the above two.
(Actually, I don't really understand, why the prefix is used for these two
macros anyway. What difference does it make compared to
bin_PROGRAMS?)
Pathfinder still compiles nicely, so there is no urgent need to fix this. But it
repeatedly irritated people, so it probably makes sense to fix this some
time.
## Comment 2059
Date: 2004-10-14 11:01:47 +0200
From: Jens Teubner <<teubner>>
Logged In: YES
user_id=731390
Assigned this to Niels. AFAIK, he's the autogen person. Lowered priority.
## Comment 2060
Date: 2004-10-16 09:11:07 +0200
From: @njnes
Logged In: YES
user_id=43556
The respective names are now also normalized. The reason for
these variables instead of bin_PROGRAMS is flexibility. The
extra macros make it possible to have install directories
per program/script/....
## Comment 2061
Date: 2005-11-09 13:03:05 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: NO TEST / COMPILATION
## Comment 2062
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1046882 at http://sourceforge.net/support/tracker.php?aid=1046882
| autogen produces ``bad macro names'' | https://api.github.com/repos/MonetDB/MonetDB/issues/382/comments | 0 | 2020-11-30T08:31:59Z | 2024-06-27T10:59:27Z | https://github.com/MonetDB/MonetDB/issues/382 | 753,270,900 | 382 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-11 01:59:29 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-10-11 09:55:54 +0200
## Comment 2053
Date: 2004-10-11 13:59:29 +0200
From: @grobian
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.4.0
Driver: MonetDB Native Driver 1.0 (Dawn)
Type \q to quit, \h for a list of available commands
auto commit mode: on
monetdb-> select * from test where id > 0;
+----------+
| id |
+==========+
+----------+
0 rows
monetdb-> drop table test;
Operation successful
monetdb-> create table test (id int);
Operation successful
monetdb-> insert into test values (2);
1 affected row
monetdb-> insert into test values (1);
1 affected row
monetdb-> delete from test;
2 affected rows
monetdb-> select * from test;
+----------+
| id |
+==========+
+----------+
0 rows
monetdb-> select * from test where id > 0;
+----------+
| id |
+==========+
| 2 |
| 1 |
+----------+
2 rows
monetdb->
## Comment 2054
Date: 2004-10-11 21:55:54 +0200
From: @njnes
Logged In: YES
user_id=43556
fixed. The deletes were only partly forwarded to the global bats
## Comment 2055
Date: 2005-11-09 13:01:38 +0100
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-11-09, meltdown: TEST ADDED / MISSING INFRASTRUCTURE
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/phantom_rows.SF-1044534.sql
I made a RFE to be able to run the test using MapiClient:
http://sourceforge.net/tracker/index.php?func=detail&aid=1352124&group_id=56967&atid=482471
(commit may follow later today, as SourceForge closes the
connection on me)
## Comment 2056
Date: 2005-11-09 20:26:34 +0100
From: @drstmane
Logged In: YES
user_id=572415
- "found" simple way to switch auto_commit on "even" in
MapiClient
- added IMHO correct stable output
Fabian,
could you please check/verify this?
## Comment 2057
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1044534 at http://sourceforge.net/support/tracker.php?aid=1044534
| SQL: MagicMonet(tm) ghosting/phantoms | https://api.github.com/repos/MonetDB/MonetDB/issues/381/comments | 0 | 2020-11-30T08:31:56Z | 2024-06-27T10:59:26Z | https://github.com/MonetDB/MonetDB/issues/381 | 753,270,869 | 381 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-09 01:26:34 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-10-10 09:53:49 +0200
## Comment 2048
Date: 2004-10-09 13:26:34 +0200
From: @grobian
When connecting as a different user than monetdb, the
sql_user and sql_session variables in the env table are
not set accordingly:
java -jar jars/MonetDB_JDBC.jar -uvoc
password:
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.4.0
Driver: MonetDB Native Driver 1.0 (Dawn)
Type \q to quit, \h for a list of available commands
auto commit mode: on
voc-> \d
voc-> select * from sys.env;
+-----------------+-----------------------------------------------------+
| name | value
|
+=================+=====================================================+
| gdk_arch | 32biti686-pc-linux-gnu
|
| gdk_version | 4.4.0
|
|
| gdk_dbname | demo
|
| sql_port | 45123
|
| sql_debug | 0
|
| sql_api | sql(default-schema)
|
| sql_user | monetdb
|
| sql_schema | default-schema
|
|
+-----------------+-----------------------------------------------------+
28 rows
voc->
(for the sake of readability I made the output a bit
smaller by throwing out some lines)
## Comment 2049
Date: 2004-10-10 09:53:49 +0200
From: @njnes
Logged In: YES
user_id=43556
sys.env contains 'default' values only. Use sys.sessions for
session variables.
## Comment 2050
Date: 2004-10-11 08:28:55 +0200
From: @grobian
Logged In: YES
user_id=963970
this is fixed for JDBC's DatabaseMetaData.getUser() method.
## Comment 2051
Date: 2005-11-09 12:57:34 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / INVALID
## Comment 2052
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1043526 at http://sourceforge.net/support/tracker.php?aid=1043526
| SQL: sql_user and sql_schema are not set correctly | https://api.github.com/repos/MonetDB/MonetDB/issues/380/comments | 0 | 2020-11-30T08:31:53Z | 2024-06-28T07:36:36Z | https://github.com/MonetDB/MonetDB/issues/380 | 753,270,835 | 380 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-09 01:06:43 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2004-10-09 03:44:33 +0200
## Comment 2044
Date: 2004-10-09 13:06:43 +0200
From: @grobian
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.4.0
Driver: MonetDB Native Driver 1.0 (Dawn)
Type \q to quit, \h for a list of available commands
auto commit mode: on
monetdb-> \d
monetdb-> create table test (v1 varchar);
Operation successful
monetdb-> insert into test values ('a math assignment');
1 affected row
monetdb-> insert into test values ('pathfinder is fun!');
1 affected row
monetdb-> select * from test where v1 like '%math%';
+--------------------+
| v1 |
+====================+
| a math assignment |
| pathfinder is fun! |
+--------------------+
2 rows
monetdb->
It's obviously this id not correct. Seems like one
char is taken wrongly, as visible below (mind the space):
monetdb-> select * from test where v1 like '% math%';
+-------------------+
| v1 |
+===================+
| a math assignment |
+-------------------+
1 row
monetdb->
## Comment 2045
Date: 2004-10-09 15:44:06 +0200
From: @grobian
Logged In: YES
user_id=963970
fixed in CVS 4.4 branch, also extented test script abit
## Comment 2046
Date: 2005-11-09 12:56:14 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: ALREADY IN TESTWEB
src/test/bugs/like-bug-sf-1043521.sql
## Comment 2047
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1043521 at http://sourceforge.net/support/tracker.php?aid=1043521
| SQL: LIKE matches too much! | https://api.github.com/repos/MonetDB/MonetDB/issues/379/comments | 0 | 2020-11-30T08:31:49Z | 2024-06-27T10:59:24Z | https://github.com/MonetDB/MonetDB/issues/379 | 753,270,795 | 379 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-09 12:47:37 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-10-10 10:16:57 +0200
## Comment 2040
Date: 2004-10-09 12:47:37 +0200
From: @grobian
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.4.0
Driver: MonetDB Native Driver 1.0 (Dawn)
Type \q to quit, \h for a list of available commands
auto commit mode: on
monetdb-> \d
monetdb-> create table test (id int unique);
Operation successful
monetdb-> insert into test values (1);
1 affected row
monetdb-> insert into test values (1);
1 affected row
monetdb-> \d test
CREATE TABLE "sys"."test" (
"id" int(19),
UNIQUE ("id") -- c1006443
);
monetdb->
The unique constraint is obviously not tested in
auto-commit mode, in a transaction the checking is done
correctly:
monetdb-> delete from test;
2 affected rows
monetdb-> start transaction;
auto commit mode: off
monetdb-> insert into test values (1);
1 affected row
monetdb-> insert into test values (1);
Error: ERROR: key constraint test.c1006443 violated
monetdb-> rollback;
auto commit mode: on
monetdb->
## Comment 2041
Date: 2004-10-10 22:16:57 +0200
From: @njnes
Logged In: YES
user_id=43556
Small inserts go into special ibats. As the key check code
reuses the
save bats the old values are missed in the check. Therefor
we insert
tables with keys directly in the 'bat' ie we don't use a
ibat then.
## Comment 2042
Date: 2005-11-09 12:54:03 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: ALREADY IN TESTWEB
src/test/bugs/unique-bug-sf-1043512.sql
## Comment 2043
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1043512 at http://sourceforge.net/support/tracker.php?aid=1043512
| SQL: UNIQUE constraint not checked | https://api.github.com/repos/MonetDB/MonetDB/issues/378/comments | 0 | 2020-11-30T08:31:46Z | 2024-06-27T10:59:23Z | https://github.com/MonetDB/MonetDB/issues/378 | 753,270,774 | 378 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-08 06:53:14 +0200
From: @MarcinZukowski
To: @njnes
Version: -- development
Last updated: 2004-10-16 09:39:17 +0200
## Comment 2036
Date: 2004-10-08 18:53:14 +0200
From: @MarcinZukowski
If a module prints lots of info to stdout/stderr within Mtest framework,
everything hangs.
Provided simple module that shows this problem, use a test script like
this:
module("crash");
crash_me();
I'm not 100% sure, but I think it used to work couple of days ago.
## Comment 2037
Date: 2004-10-16 09:39:17 +0200
From: @njnes
Logged In: YES
user_id=43556
Works just fine.
## Comment 2038
Date: 2005-11-09 12:53:04 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / REDUNDANT
## Comment 2039
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1043189 at http://sourceforge.net/support/tracker.php?aid=1043189
| Mtest does not like printing to stdout/stderr | https://api.github.com/repos/MonetDB/MonetDB/issues/377/comments | 0 | 2020-11-30T08:31:43Z | 2024-06-28T07:36:36Z | https://github.com/MonetDB/MonetDB/issues/377 | 753,270,745 | 377 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-07 06:34:05 +0200
From: @swingbit
To: @njnes
Version: -- development
Last updated: 2004-10-10 10:12:27 +0200
## Comment 2032
Date: 2004-10-07 18:34:05 +0200
From: @swingbit
The rpc() function (it remotely executes a mil string
and gets a result back), the mil statement is correctly
executed on the server side, but then the client hangs:
SERVER:
module(tcpip);
PROC f(): int {
print(10);
RETURN 10;
}
fork(listen());
CLIENT:
module(tcpip);
var cid := open("localhost");
var res := rpc(cid, "f();");
print(res); it never gets here
## Comment 2033
Date: 2004-10-10 22:12:27 +0200
From: @njnes
Logged In: YES
user_id=43556
missing connection waiter thread added. This thread handles
all incoming
data. And does the needed 'sema_up' to resolve the hanging
rpc().
## Comment 2034
Date: 2005-11-09 12:50:21 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: TEST ADDED / SUCCESS
tests/BugDay_2005-11-09_2.9.3/Tests/tcpip_close_crash.SF-1042370.milS
## Comment 2035
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1042381 at http://sourceforge.net/support/tracker.php?aid=1042381
| TCPIP gets stuck on rpc() | https://api.github.com/repos/MonetDB/MonetDB/issues/376/comments | 0 | 2020-11-30T08:31:40Z | 2024-06-27T10:59:21Z | https://github.com/MonetDB/MonetDB/issues/376 | 753,270,705 | 376 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-07 06:18:22 +0200
From: @swingbit
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2006-03-27 05:11:15 +0200
## Comment 2025
Date: 2004-10-07 18:18:22 +0200
From: @swingbit
Trying to close the connection with close(int
connect_id) simply segfaults Mserver:
module("tcpip", "alarm");
VAR tcpip_port := get_mapi_port();
fork(listen(tcpip_port));
sleep(2);
VAR c := open("localhost:"+str(tcpip_port));
sleep(2);
close(c);
## Comment 2026
Date: 2004-10-10 22:13:57 +0200
From: @njnes
Logged In: YES
user_id=43556
only destroy waiter if one exists. This problem was/is
related to the
missing connection waiter bug which is also solved.
## Comment 2027
Date: 2005-11-09 12:39:02 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: TEST ADDED / SUCCESS
tests/BugDay_2005-11-09_2.9.3/Tests/tcpip_close_crash.SF-1042370.milS
## Comment 2028
Date: 2005-12-20 23:32:33 +0100
From: @drstmane
Logged In: YES
user_id=572415
re-open as the respective test fails; cf. e.g.
http://monetdb.cwi.nl/testing/projects/monetdb/Current/MonetDB/.mTests103/GNU.64.64.d-Fedora4/tests_BugDay_2005-11-09_4.9.3/tcpip_close_crash.SF-1042370.out.00.html
## Comment 2029
Date: 2006-01-22 12:07:22 +0100
From: @drstmane
Logged In: YES
user_id=572415
Peter,
could you please check, whether this output is ok,
http://monetdb.cwi.nl/testing/projects/monetdb/Current/MonetDB/.mTests103/GNU.64.64.d-Fedora4/tests_BugDay_2005-11-09_4.9.3/tcpip_hangs_on_close.SF-1070569.out.00.html
and if not, possibly have a look at the bug?
Thanks!
Stefan
ps: This bug might be related
https://sourceforge.net/tracker/index.php?func=detail&aid=1070569&group_id=56967&atid=482468
## Comment 2030
Date: 2006-03-27 17:11:15 +0200
From: @njnes
Logged In: YES
user_id=43556
check output seems correct, ie closing bug.
## Comment 2031
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1042370 at http://sourceforge.net/support/tracker.php?aid=1042370
| TCPIP segfault | https://api.github.com/repos/MonetDB/MonetDB/issues/375/comments | 0 | 2020-11-30T08:31:37Z | 2024-06-27T10:59:20Z | https://github.com/MonetDB/MonetDB/issues/375 | 753,270,680 | 375 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-06 12:35:02 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2004-10-06 07:38:55 +0200
## Comment 2019
Date: 2004-10-06 12:35:02 +0200
From: @grobian
An SQL string containing weird char(s) hang the server.
I don't think via browsers and web protocols the weird
char is being preserved, adding a in CVS.
select
'http://shopping.b92.net/index.php?IDS=11&ID_Group=39&ID_prod=754&PROD_NAME=Belgrade+Coffee+Shop+4&GROUP_NAME=Muziýa+produkcija';
## Comment 2020
Date: 2004-10-06 12:46:46 +0200
From: @grobian
Logged In: YES
user_id=963970
it is a JDBC-only bug, as it seems.
the Java MapiClient seems to be broken when using SQL. (it
simply hangs)
## Comment 2021
Date: 2004-10-06 17:21:30 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
There may be several problems, but one important one is: in
the stream protocol you should specify the number of *bytes*
that follow, not the number of Unicide characters (your 3
"funny" characters are the UTF-8 representation of a single
Unicode character).
## Comment 2022
Date: 2004-10-06 19:37:31 +0200
From: @grobian
Logged In: YES
user_id=963970
fixed in 2-4-0 branch
Indeed, a sloppy implementation neglecting the difference
between a byte representation and a String.
## Comment 2023
Date: 2005-11-09 11:42:51 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: ALREADY IN TESTWEB
src/test/bugs/unicode_varchar-bug-sf-1041324.sql
## Comment 2024
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1041324 at http://sourceforge.net/support/tracker.php?aid=1041324
| SQL: weird chars hang server, maxing to 100% CPU | https://api.github.com/repos/MonetDB/MonetDB/issues/374/comments | 0 | 2020-11-30T08:31:34Z | 2024-06-27T10:59:19Z | https://github.com/MonetDB/MonetDB/issues/374 | 753,270,648 | 374 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-04 07:32:09 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-10-05 01:43:11 +0200
## Comment 2014
Date: 2004-10-04 19:32:09 +0200
From: @grobian
The following create statements do not complete
successfully:
CREATE TABLE "ggle_urls"
(
"id" integer,
"url" varchar(255)
);
ALTER TABLE "ggle_urls" ADD PRIMARY KEY("id");
CREATE TABLE "ggle_url_links"
(
"urls_id" integer,
"link" varchar(255)
);
ALTER TABLE "ggle_url_links" ADD FOREIGN KEY("urls_id")
REFERENCES "ggle_urls"("id");
CREATE TABLE "ggle_url_data"
(
"urls_id" integer,
"doc" varchar(100),
"dom" varchar(100),
"size" varchar(25)
);
ALTER TABLE "ggle_url_data" ADD FOREIGN KEY("urls_id")
REFERENCES "ggle_urls"("id");
The last two ALTER TABLE statements generate an error:
"Error in result"
## Comment 2015
Date: 2004-10-05 01:43:11 +0200
From: @njnes
Logged In: YES
user_id=43556
Seems fixed in 4.4.0 branch
## Comment 2016
Date: 2004-10-05 10:29:39 +0200
From: @grobian
Logged In: YES
user_id=963970
after a cvs update, bootstrap, compile, etc. the error
still persists (on a clean database) for me... :(
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.4.0
Driver: MonetDB Native Driver 1.0 (Dawn)
Type \q to quit, \h for a list of available commands
auto commit mode: on
monetdb-> CREATE TABLE "ggle_urls"
(
"id" integer,
"url" varchar(255)
);monetdb-> monetdb-( monetdb-( monetdb-(
Operation successful
monetdb-> ALTER TABLE "ggle_urls" ADD PRIMARY KEY("id");
Operation successful
monetdb-> CREATE TABLE "ggle_url_links"
(
"urls_id" integer,
"link" varchar(255)
);monetdb-> monetdb-( monetdb-( monetdb-(
Operation successful
monetdb-> ALTER TABLE "ggle_url_links" ADD FOREIGN
KEY("urls_id")
REFERENCES "ggle_urls"("id");monetdb->
Error: Error in result
Can it be related to auto-commit mode, that JDBC by default
uses?
(hint hint hint hint)
## Comment 2017
Date: 2005-11-09 11:39:08 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: TEST ADDED / SUCCESS
src/test/BugDay_2005-11-09_2.9.3/Tests/alter_table.SF-1040057.sql
## Comment 2018
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1040057 at http://sourceforge.net/support/tracker.php?aid=1040057
| SQL: ALTER TABLE results in "Error in result" messages | https://api.github.com/repos/MonetDB/MonetDB/issues/373/comments | 0 | 2020-11-30T08:31:30Z | 2024-06-27T10:59:18Z | https://github.com/MonetDB/MonetDB/issues/373 | 753,270,612 | 373 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-10-01 01:33:59 +0200
From: @grobian
To: MonetDB4 devs <<bugs-monetdb4>>
Version: -- development
Last updated: 2004-10-07 11:39:47 +0200
## Comment 2009
Date: 2004-10-01 13:33:59 +0200
From: @grobian
On windows, if a Mserver window is started, then a
client is started, upon closure of the client session,
the Mserver terminates as well. This is not a
desirable situation.
(experienced on WIN2000)
## Comment 2010
Date: 2004-10-01 23:30:44 +0200
From: Koen van de Sande <<shadowmorpher>>
Logged In: YES
user_id=270334
This works just fine for me on Windows 2000SP4 with 4.4.0-
RC2.
4.4.0-final I have only tried on Windows XPSP2 and have not
experienced server closures of this kind. Exiting the SQL client
(started through the start menu, as is the server) leaves the
server running. But it could be a Win2k/4.4.0 issue, but
haven't upgraded there yet.
## Comment 2011
Date: 2004-10-07 11:39:47 +0200
From: @grobian
Logged In: YES
user_id=963970
Problems seems only to exist on our very crappy Win2000
machine, even after upgrading to SP4.
## Comment 2012
Date: 2005-11-09 11:32:41 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: ALREADY IN TESTWEB
any mapi test will do.
## Comment 2013
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1038432 at http://sourceforge.net/support/tracker.php?aid=1038432
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 on windows closes after connection with client | https://api.github.com/repos/MonetDB/MonetDB/issues/372/comments | 0 | 2020-11-30T08:31:27Z | 2024-06-28T07:36:35Z | https://github.com/MonetDB/MonetDB/issues/372 | 753,270,572 | 372 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-30 09:42:47 +0200
From: Koen van de Sande <<shadowmorpher>>
To: @njnes
Version: -- development
Last updated: 2004-10-05 01:45:53 +0200
## Comment 2002
Date: 2004-09-30 21:42:47 +0200
From: Koen van de Sande <<shadowmorpher>>
OS: Windows 2000 and Windows XP SP2
CPU: Athlon 800 and Centrino 1600 Mhz
Version: MontDB 4.4.0_rc2 Windows Installer dated 20/9
When a lot of tuples are added and in the meantime
selection queries are made from a single Java process
through MonetJDBC-rc2, eventually Mserver crashes
with a runtime error in the Visual C++ runtime.
I saw something like "Batcopy: name must be supplied"
in the server prompt before it closed, but it didn't stay
open long enough.
The moment at which the error occurs is not fixed, but
we have a program which has never been able to run to
full completion yet. It is not a good program (as it runs
insertions in quadratic time at the moment), but it makes
Monet crash consistently.
We have a program available to reproduce this, but we
cannot upload it here as it is part of an assignment. We
can send it by e-mail on request. E-mail kvds at * gmx
dot net
The program consists of tables.sql, which can be used in
the SQL client to create the tables needed for the
program, and the program can be run through "java -cp
MonetJDBC.jar;. DBLoad" and it will try to load the
btv.xml file into the database. Typically it will crash
when there are about 12-13 dots after the "Parsing xml"
prompt, this takes 5-10 minutes.
## Comment 2003
Date: 2004-10-01 00:27:31 +0200
From: @grobian
Logged In: YES
user_id=963970
I sent you an email with a request for the files.
Will further investigate this.
## Comment 2004
Date: 2004-10-01 14:36:32 +0200
From: @grobian
Logged In: YES
user_id=963970
Monet Database Server V4.4.0
Copyright (c) 1993-2004, CWI. All rights reserved.
Compiled for i686-redhat-linux-gnu/32bit; dynamically linked.
Visit http://monetdb.cwi.nl/ for further information.
MonetDB>!ERROR: BBPgetdesc: deleting illegal bat(56034)
15/53/56034
Segmentation fault
The performed test goes nice and fast, till a sudden point
where everything seems to get real slow and finally Mserver
dies with the above BAT administration error.
## Comment 2005
Date: 2004-10-01 23:35:43 +0200
From: Koen van de Sande <<shadowmorpher>>
Logged In: YES
user_id=270334
We are discussing the problem by e-mail, it is reproducible
and there appears to be a problem with BAT
management/SELECT queries inside update transactions.
## Comment 2006
Date: 2004-10-05 01:45:53 +0200
From: @njnes
Logged In: YES
user_id=43556
After fixing the bat leak problems this test finishes now
succesfully
## Comment 2007
Date: 2005-11-09 11:27:15 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / TO LARGE (sources
missing too)
The test would be to large for the nightly testing. Also the
sources for this test are missing.
## Comment 2008
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1037993 at http://sourceforge.net/support/tracker.php?aid=1037993
| Mserver crashes with runtime error | https://api.github.com/repos/MonetDB/MonetDB/issues/371/comments | 0 | 2020-11-30T08:31:24Z | 2024-06-27T10:59:16Z | https://github.com/MonetDB/MonetDB/issues/371 | 753,270,550 | 371 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-30 10:34:09 +0200
From: @swingbit
To: @njnes
Version: -- development
Last updated: 2005-01-11 03:27:32 +0100
## Comment 1995
Date: 2004-09-30 10:34:09 +0200
From: @swingbit
As far as I know,
1/0 = inf
0/0 = NaN
In Monet (same behavior for int and dbl values):
print(1/0) = [nil]
print(0/0) = [nil]
The multiplexed version gives the same:
var a:=new(void,int).seqbase(0@0).insert(nil,
0).insert(nil, 1);
var b := [/](a, a);
b.print();
-------------------------
h tmp_117 name
void int type
-------------------------
[ 0@0, nil ]
[ 1@0, 1 ]
The multiplexed version using bat_arith is much worse:
Mserver exits with floating point exception.
## Comment 1996
Date: 2004-10-10 22:44:22 +0200
From: @njnes
Logged In: YES
user_id=43556
we probably should (like all other script languages) issue a
Division by
zero error.
## Comment 1997
Date: 2004-10-16 10:04:40 +0200
From: @njnes
Logged In: YES
user_id=43556
For single division and modulo, the users will now get an error
"division by zero". Still to do for map operations.
## Comment 1998
Date: 2005-01-11 15:27:31 +0100
From: @njnes
Logged In: YES
user_id=43556
we now correctly give errors.
## Comment 1999
Date: 2005-11-09 11:40:11 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.8/Tests/insert_from_no-existing_tables.SF-1037020.*
## Comment 2000
Date: 2005-11-09 11:44:14 +0100
From: @drstmane
Logged In: YES
user_id=572415
Correction:
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.8/Tests/division_by_zero.SF-1037591.*
## Comment 2001
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1037591 at http://sourceforge.net/support/tracker.php?aid=1037591
| division by 0 (exception or wrong result) | https://api.github.com/repos/MonetDB/MonetDB/issues/370/comments | 0 | 2020-11-30T08:31:21Z | 2024-06-27T10:59:15Z | https://github.com/MonetDB/MonetDB/issues/370 | 753,270,523 | 370 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-29 04:08:50 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-09-29 11:34:37 +0200
## Comment 1990
Date: 2004-09-29 16:08:50 +0200
From: @arjan
Server segfaults on the following:
create table t (i integer);
insert into t select * from fds;
A check for table existence would be appreciated :-)
## Comment 1991
Date: 2004-09-29 16:29:35 +0200
From: @arjan
Logged In: YES
user_id=20087
The crash occurrs on ANY wrong issued query in the select,
e.g. forcing an invalid argument to substring also produces
the crash:
create table t(x varchar(244));
insert into t select substring('aap' from cast(1 as double)
for 3);
## Comment 1992
Date: 2004-09-29 23:34:37 +0200
From: @njnes
Logged In: YES
user_id=43556
Fixed in 'rc branch'
## Comment 1993
Date: 2005-11-09 11:40:35 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.8/Tests/insert_from_no-existing_tables.SF-1037020.*
## Comment 1994
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1037020 at http://sourceforge.net/support/tracker.php?aid=1037020
| SQL: INSERT ... SELECT fro non-existing table crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/369/comments | 0 | 2020-11-30T08:31:17Z | 2024-06-27T10:59:14Z | https://github.com/MonetDB/MonetDB/issues/369 | 753,270,482 | 369 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-27 10:12:03 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-28 02:58:14 +0200
## Comment 1985
Date: 2004-09-27 22:12:03 +0200
From: @grobian
gcc -DHAVE_CONFIG_H -I.
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server
-I../..
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/common
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/gdk
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/monet
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/plain
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/contrib
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/mapi
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/C
-g -O2 -std=c99 -Wall -W -DHWCOUNTERS -DHW_Linux
-DHW_i686 -Werror-implicit-function-declaration -Werror
-Wno-format -Wno-unused-function -Wno-unused-label
-Wno-strict-aliasing -Wno-sign-compare -c sql_bm.c
-fPIC -DPIC -o .libs/sql_bm.o
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:
In function `bm_name2lbid':
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:473:
error: invalid operands to binary -
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:473:
error: invalid operands to binary -
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:473:
warning: left-hand operand of comma expression has no
effect
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:474:
error: invalid operands to binary -
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:474:
error: invalid operands to binary -
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:474:
warning: left-hand operand of comma expression has no
effect
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:
In function `bm_lbid2vid':
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:486:
error: invalid operands to binary -
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:486:
error: invalid operands to binary -
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_bm.mx:486:
warning: left-hand operand of comma expression has no
effect
make[3]: *** [sql_bm.lo] Error 1
make[3]: Leaving directory
`/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/build/src/server'
make[2]: *** [install] Error 2
make[2]: Leaving directory
`/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/build/src/server'
make[1]: *** [install-recursive] Error 1
make[1]: Leaving directory
`/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/build/src'
make: *** [install-recursive] Error 1
(pictor:~) fabian%
I have my suspicions that this error has been
introduced after the icc fixes, as the (to me quite
daft) error appears to happen in one of the various
define parts in gdk.h
## Comment 1986
Date: 2004-09-28 10:19:47 +0200
From: @grobian
Logged In: YES
user_id=963970
After some fixes by Sjoerd, the compilation process stops at
the next location:
gcc -DHAVE_CONFIG_H -I.
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server
-I../..
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/common
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/gdk
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/monet
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/plain
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/contrib
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/mapi
-I/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/include/MonetDB/C
-g -O2 -std=c99 -Wall -W -DHWCOUNTERS -DHW_Linux -DHW_i686
-Werror-implicit-function-declaration -Werror -Wno-format
-Wno-unused-function -Wno-unused-label -Wno-strict-aliasing
-Wno-sign-compare -c sql_mvc.c -fPIC -DPIC -o .libs/sql_mvc.o
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_mvc.mx:
In function `temp_descriptor':
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_mvc.mx:393:
warning: comparison of unsigned expression < 0 is always false
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/src/server/sql_mvc.mx:393:
warning: comparison of unsigned expression < 0 is always false
make[3]: *** [sql_mvc.lo] Error 1
make[3]: Leaving directory
`/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/build/src/server'
make[2]: *** [install] Error 2
make[2]: Leaving directory
`/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/build/src/server'
make[1]: *** [install-recursive] Error 1
make[1]: Leaving directory
`/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/build/src'
make: *** [install-recursive] Error 1
An oid is given as input for somethingt hat requires a bat.
A hard cast solves the compilation problem, but the cast
seems a bit 'hairy'.
## Comment 1987
Date: 2004-09-28 14:25:14 +0200
From: @grobian
Logged In: YES
user_id=963970
Since only the PGI compiler knows how to compile the current
code, someone should either check in the hard cast to bat,
or do something to fix it.
## Comment 1988
Date: 2005-11-09 11:36:49 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: NO TEST / COMPILATION
## Comment 1989
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1035706 at http://sourceforge.net/support/tracker.php?aid=1035706
| SQL head does not compile anymore | https://api.github.com/repos/MonetDB/MonetDB/issues/368/comments | 0 | 2020-11-30T08:31:14Z | 2024-06-27T10:59:13Z | https://github.com/MonetDB/MonetDB/issues/368 | 753,270,449 | 368 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-27 09:31:40 +0200
From: @mlkersten
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2005-11-25 01:50:13 +0100
## Comment 1978
Date: 2004-09-27 21:31:40 +0200
From: @mlkersten
The cleanup of MIL has not been completed.
In particular, the parser still handles the following
statements without a warning.
ls;
var i:=1.int;
ls is a predefined proc, which appears as a TOK_IDENT
in the parser, which then expands it to a possible path
expression. (ls could be used in the expression
ls.fcn(), where ls would denote a variable rather then
a procedure. Despite the fact that it is already
predefined.)
In the second case, int is known to be an atom and
should be used as a proper function call.
## Comment 1979
Date: 2004-09-30 11:05:47 +0200
From: @njnes
Logged In: YES
user_id=43556
Last part of the enfore () introduced a bug in handeling
b.[]() which is
now fixed in the head.
## Comment 1980
Date: 2005-11-09 11:36:21 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / FALIURE
MonetDB/tests/BugDay_2005-11-09_4.8/Tests/MIL_cleanup.SF-1035451.*
Apparently, casts still do not require parantheses:
ls;
!ERROR: interpret: unknown variable 'ls'.
1.int;
1.print;
!ERROR: "1.print;"
! ^
! new MIL syntax (>=4.4) requires parentheses for all operators.
!.
Re-opened bug report.
## Comment 1981
Date: 2005-11-24 23:57:16 +0100
From: @peterboncz
Logged In: YES
user_id=591107
well, indeed atom casts still work without parentheses, but
I do not see the urgency for fixing this.
## Comment 1982
Date: 2005-11-25 01:50:12 +0100
From: @drstmane
Logged In: YES
user_id=572415
Actually, this has just been fixed.
Thanks, Peter!
## Comment 1983
Date: 2005-11-25 02:07:52 +0100
From: @peterboncz
Logged In: YES
user_id=591107
I was going to add that;
go to bed!
## Comment 1984
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1035681 at http://sourceforge.net/support/tracker.php?aid=1035681
| MIL cleanup | https://api.github.com/repos/MonetDB/MonetDB/issues/367/comments | 0 | 2020-11-30T08:31:11Z | 2024-06-27T10:59:12Z | https://github.com/MonetDB/MonetDB/issues/367 | 753,270,418 | 367 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-27 03:14:41 +0200
From: @peterboncz
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-09-30 11:38:50 +0200
## Comment 1974
Date: 2004-09-27 15:14:41 +0200
From: @peterboncz
module(xtables);
var b := new(void,int).CTgroup();
var i := int(b);
bat(sprintf("tmp_%d_histo", i).print;
b := 0;
bat(sprintf("tmp_%d_histo", i).print; bat should not
exist anymore..
the tmp_X_histo bats is leaked, due to a bug in xtables.mx
## Comment 1975
Date: 2004-09-30 11:38:50 +0200
From: @drstmane
Logged In: YES
user_id=572415
According to Peter and Jan, the bug has been fixed, and the
fix has been tested.
## Comment 1976
Date: 2005-11-09 11:27:09 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: NO TEST / OBSOLETE
histolink does not exist anymore
## Comment 1977
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1035451 at http://sourceforge.net/support/tracker.php?aid=1035451
| leak in histolinks (xtables) | https://api.github.com/repos/MonetDB/MonetDB/issues/366/comments | 0 | 2020-11-30T08:31:08Z | 2024-06-27T10:59:11Z | https://github.com/MonetDB/MonetDB/issues/366 | 753,270,384 | 366 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-27 03:11:37 +0200
From: @peterboncz
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2007-11-30 04:20:03 +0100
## Comment 1967
Date: 2004-09-27 15:11:37 +0200
From: @peterboncz
MonetDB has a structure MT_alloc_map that holds for
each 64KB tille
of the address space *until* 3GB a status char (total
size is 49600).
Each memory allocation (typically GDKmmap) and free
administers its region by calling MT_alloc_register()
that marks the allocated region (with 'M' characters,
in case of GDKmmap).
This array is used to administer which regions are
used; basically to monitor address space fragmentation.
The mechanism is disabled on 64-bits systems (where
this is not an issue)
Problem: the new fedora-2 core Linux actually assign VM
addresses between 1GB and 4GB. Now we have to contend
with addresses beyond 3GB.
If the base address is before the 3GB limit, but the
end of the region extends beyond, then
MT_alloc_register() can beyond the 49600 allocated
characters for MT_alloc[]. One of the static data
structures just beyond that is bbptrim[], which then
gets overrun, leading to a crash in BBPtrim.
The fix is to extend MT_allco simply to 65536 slots.
Then it will work
on any 32-bits system no matter how the OS implements
the VM address space.
## Comment 1968
Date: 2004-09-30 11:38:11 +0200
From: @drstmane
Logged In: YES
user_id=572415
According to Peter and Jan, this bug has been fixed, and the
fix has been tested.
## Comment 1969
Date: 2005-11-09 11:20:26 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: NO TEST
memory/address space restriction tests *might* be added later
## Comment 1970
Date: 2006-11-10 04:20:01 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 1971
Date: 2006-11-29 10:04:17 +0100
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
over-ruling "sf-robot":
re-set to "Pending" since adding a proper test has neither been done nor finally been discarded, yet.
User (submitter) & developer (assignee),
please check, again.
## Comment 1972
Date: 2007-11-30 04:20:03 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
Originator: NO
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 1973
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1035450 at http://sourceforge.net/support/tracker.php?aid=1035450
| MonetDB crash: 3>GB addresses on 32-bits systems | https://api.github.com/repos/MonetDB/MonetDB/issues/365/comments | 0 | 2020-11-30T08:31:05Z | 2024-06-27T10:59:10Z | https://github.com/MonetDB/MonetDB/issues/365 | 753,270,356 | 365 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-27 10:43:10 +0200
From: @sjoerdmullender
To: @sjoerdmullender
Version: -- development
Last updated: 2004-09-27 03:21:55 +0200
## Comment 1963
Date: 2004-09-27 10:43:10 +0200
From: @sjoerdmullender
Since my fix to get inline functions on all platforms
that support it, the icc compiler on Linux complains
about certain constructs with the message
error 1173: an entity with internal linkage cannot be
referenced within an inline function with external linkage
This happens in at least gdk_search.mx, gdk_utils.mx,
and gdk_bat.mx.
I analyzed this a bit and saw that at least one of the
casues is the static inline functions that are defined
in the .h part of gdk.mx which are referenced in extern
inline functions in gdk_search.mx.
One fix would be to rewrite the static inline functions
in gdk.mx as define expressions.
I'll also analyze the causes of the other error messages.
## Comment 1964
Date: 2004-09-27 15:21:55 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Fixed in CVS in the head branch.
## Comment 1965
Date: 2005-11-09 11:18:41 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: NO TEST / COMPILATION
## Comment 1966
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1035327 at http://sourceforge.net/support/tracker.php?aid=1035327
| icc not happy with inline functions | https://api.github.com/repos/MonetDB/MonetDB/issues/364/comments | 0 | 2020-11-30T08:31:02Z | 2024-06-27T10:59:09Z | https://github.com/MonetDB/MonetDB/issues/364 | 753,270,329 | 364 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-26 03:58:42 +0200
From: @grobian
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-11-20 12:11:43 +0100
## Comment 1957
Date: 2004-09-26 15:58:42 +0200
From: @grobian
var a, b;
a := 1;
b := 1;
print(a = 1 and b = 1);
!ERROR: interpret: no matching MIL operator to
'and(int, bit)'.
print((a = 1) and (b = 1));
[ true ]
why are the ( ) needed for this to work? Is the and
operator positioned higher in the tree? I feel other
languages (ie. a == 1 && b == 1) have the && operator
on a lower level than ==, !=, etc.
## Comment 1958
Date: 2004-09-27 12:29:28 +0200
From: @grobian
Logged In: YES
user_id=963970
A discussion with Niels and Sjoerd made clear there is no
operator prevalence. However, the parsing is from right to
left, which should be changed to left to right.
## Comment 1959
Date: 2004-11-20 00:11:43 +0100
From: @peterboncz
Logged In: YES
user_id=591107
this was fixed a month ago..
MIL now has operatir precedence!!!
and proper order of parsing
## Comment 1960
Date: 2004-11-20 12:10:39 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
If that is so, why do I get this:
MonetDB>print(a = 1 and b = 1);
!ERROR: interpret: no matching MIL operator to 'and(int, bit)'.
!MAYBE YOU MEAN:
! and(lng, lng) : lng
! and(int, int) : int
! and(sht, sht) : sht
! and(chr, chr) : chr
! and(bit, bit) : bit
!ERROR: interpret_params: =(param 2): evaluation error.
!ERROR: interpret_params: print(param 1): evaluation error.
MonetDB>
## Comment 1961
Date: 2005-11-09 11:18:16 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.8/Tests/operator_precedence.SF-1035021.*
## Comment 1962
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1035021 at http://sourceforge.net/support/tracker.php?aid=1035021
| MIL: brackets required evaluation order/operator prevalence | https://api.github.com/repos/MonetDB/MonetDB/issues/363/comments | 0 | 2020-11-30T08:30:59Z | 2024-06-28T07:36:35Z | https://github.com/MonetDB/MonetDB/issues/363 | 753,270,291 | 363 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-26 11:30:55 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-27 03:29:10 +0200
## Comment 1954
Date: 2004-09-26 11:30:55 +0200
From: @grobian
MonetDB> print("\\" = "_");
MonetDB> ;
[ false ]
the ; on the second line is required to get the answer...
MonetDB> print("_" = "_");
[ true ]
without the escaped slash the answer appears immediately
MonetDB> print("\" = "_");
MonetDB> ;
!ERROR: ";"
! ^
! can't help you here, sorry.
That in this case a second ; is required is ok, I can
live with that.
## Comment 1955
Date: 2005-11-09 11:08:20 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.8/Tests/escaped_backslash.SF-1034927-*
## Comment 1956
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1034927 at http://sourceforge.net/support/tracker.php?aid=1034927
| MIL: escape slash causing delay of execution? | https://api.github.com/repos/MonetDB/MonetDB/issues/362/comments | 0 | 2020-11-30T08:30:56Z | 2024-06-27T10:59:07Z | https://github.com/MonetDB/MonetDB/issues/362 | 753,270,254 | 362 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-26 11:09:01 +0200
From: @grobian
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-26 03:39:41 +0200
## Comment 1950
Date: 2004-09-26 11:09:01 +0200
From: @grobian
both release candidate (rc3) and cvs head
mil>print (1 == 1);
MAPI = anonymous@localhost:50000
ACTION= read_line
QUERY = print (1 == 1);
ERROR = Connection terminated
Ok, it is wrong, because == is not correct, but
crashing is a bit ugly...
Server terminal says:
Myypostprocess: must == to a variable
BATSIGcrash: Mserver internal error (Segmentation
fault), please restart.
## Comment 1951
Date: 2004-09-26 15:39:41 +0200
From: @drstmane
Logged In: YES
user_id=572415
fixed in CVS (both MonetDB_4-4-0 branch and main trunk)
Myypostprocess did correctly detect assignments to
non-variables as errors.
However, these errors were not at all propagated up through
the calling stack.
Fixed that, by making Myypostprocess return -1 in case of
such error, and have
all callers of Myypostprocess check its retrun status and
propagate it where
necessary.
! Maybe someone who *does* know the MIL parser could have a
look whether
! mychanges do indeed make sense ...
! TODO !
"==" is actually no assignment, but a non-valid operator in MIL;
hence, the parser should give a more accurate error message
than the one below.
(I leave this, to someone who *does* know the parser, and
hence, does not
require hours to fix this ... )
========
Monet Database Server V4.4.0_rc3
Copyright (c) 1993-2004, CWI. All rights reserved.
Compiled for i686-redhat-linux-gnu/32bit; dynamically linked.
Visit http://monetdb.cwi.nl/ for further information.
MonetDB>(0:+=0);
!ERROR: Myypostprocess: must += to a variable
MonetDB>print(0:+=0);
!ERROR: Myypostprocess: must += to a variable
MonetDB>(0==0);
!ERROR: Myypostprocess: must == to a variable
MonetDB>print(0==0);
!ERROR: Myypostprocess: must == to a variable
MonetDB>
========
ps: Sorry, today(?) I have no time to add a proper test
script ...
## Comment 1952
Date: 2005-11-09 11:02:09 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.8/Tests/is_equal_operator_in_MIL.SF-1034922.*
## Comment 1953
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1034922 at http://sourceforge.net/support/tracker.php?aid=1034922
| MIL: == operator crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/361/comments | 0 | 2020-11-30T08:30:53Z | 2024-06-27T10:59:06Z | https://github.com/MonetDB/MonetDB/issues/361 | 753,270,221 | 361 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-25 05:24:53 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-25 08:57:05 +0200
## Comment 1946
Date: 2004-09-25 17:24:53 +0200
From: @grobian
when the VOC data set is loaded:
CREATE TABLE "sys"."newvoyages" (
"number" int(9),
"number_sup" char(1),
"trip" int(9),
"trip_sup" char(1),
"boatname" varchar(50),
"master" varchar(50),
"tonnage" int(9),
"type_of_boat" varchar(30),
"built" varchar(15),
"bought" varchar(15),
"hired" varchar(15),
"yard" char(1),
"chamber" char(1),
"departure_date" date,
"departure_harbour" varchar(30),
"cape_arrival" date,
"cape_departure" date,
"cape_call" boolean,
"arrival_date" date,
"arrival_harbour" varchar(30),
"next_voyage" int(9),
"particulars" varchar(530)
);
the following query works:
INSERT INTO "newvoyages" ("number", "number_sup",
"trip", "trip_sup",
"boatname", "master", "tonnage", "type_of_boat",
"built",
"bought", "hired", "yard", "chamber", "departure_date",
"departure_harbour", "cape_arrival", "cape_departure",
"arrival_date", "arrival_harbour", "next_voyage",
"particulars")
SELECT * FROM "voyages";
however, an ordered variant, does not?!?
INSERT INTO "newvoyages" ("number", "number_sup",
"trip", "trip_sup",
"boatname", "master", "tonnage", "type_of_boat",
"built",
"bought", "hired", "yard", "chamber", "departure_date",
"departure_harbour", "cape_arrival", "cape_departure",
"arrival_date", "arrival_harbour", "next_voyage",
"particulars")
SELECT * FROM "voyages" order by "number",
"number_sup", "trip", "trip_sup";
## Comment 1947
Date: 2004-09-25 20:57:05 +0200
From: @njnes
Logged In: YES
user_id=43556
Fixed in head.
The sql engine postpones order by and limit to the output
generation.
In case of subqueries this optimization needs to be reverted.
## Comment 1948
Date: 2005-11-09 11:19:13 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: ALREADY IN TESTWEB
Many tests with order by exist (mainly because we want the
output to have a stable order)
case_cast-bug-sf-1019506.sql:order by non_unique, type,
index_qualifier, index_name, ordinal_position;
drop_foreignkey_selfreferencing.sql:-- simple 'standard'
case (dropping in the right order)
innerjoin_multiple-bug-sf-943661.sql: order by schemas.name,
tables.name, columns.name;
innerjoin_multiple-bug-sf-943661.sql: order by s.name,
t.name, c.name;
update_where-bug-sf-976963.sql:select * from test_update
order by id;
update_where-bug-sf-976963.sql:select * from test_update
order by id;
## Comment 1949
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1034613 at http://sourceforge.net/support/tracker.php?aid=1034613
| SQL: column number change on use of order by? | https://api.github.com/repos/MonetDB/MonetDB/issues/360/comments | 0 | 2020-11-30T08:30:51Z | 2024-06-27T10:59:05Z | https://github.com/MonetDB/MonetDB/issues/360 | 753,270,199 | 360 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-22 11:39:36 +0200
From: @grobian
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-25 05:22:20 +0200
## Comment 1940
Date: 2004-09-22 11:39:36 +0200
From: @grobian
The current state of the SQL testweb cross platform
check site seems to be all but valid. For some test,
red crosses are displayed while the output seems not to
differ. Green little o's on the other hand seem
sometimes to have huge differences with errors and stuff.
For some tests, a 404 appears as the datafiles seem not
to be in place.
Maybe there is no atomic handling of updating the test
results?
Another weird thing is the fact that sometimes an email
"testing sql failed" gets sent, after which testweb
shows the message testing is currently in progress.
These inconsistencies make the process of checking bugs
a little bit hard. Some improvements should be
considered. 1337 man! :D
## Comment 1941
Date: 2004-09-22 12:04:51 +0200
From: @drstmane
Logged In: YES
user_id=572415
A few quick remarls before I look into details:
I made a small mistake yesterday when moving the testing
from db1 to crux which did skrew-up testing ---
unfortunately, I only noticed after testing had started at
midnight, and by than, db1 was already running at load 40(!)
for reasons I don't quite know, yet...
Feeling "responsible", I tried to fix stings before going
home --- this took me a bit more than 2 (two) hours ... |-(
My apologies, for not warning you about the fact that
testing might be a bit "trange" today.
However, this on hold for Current (and Current testing only
started at 10:00am this morning after the (delayed) Stable
testing had finished.
Hence, Current testing is indeed still running, and
available result represent the state of Monday, but
"messed-up" as yesterdays testing
(to be precise: building of the testweb) did not finish due
to the known problems with the TestTools on db1 --- which
are hopefully solved, now, by haveing crux as a new
testing-server.
Stable results should be fine; though.
Since about a week (or even longer?), the testweb should not
say "testing is currently in progress" any more!
If it still does, please tell me where!
I agree, that this "inconsistencies" make the process of
checking bugs
a little hard --- unfortunately, my day also has only 24
hours to make testing more reliable (despite all odds like
hardware, networks, ldap, etc...)
## Comment 1942
Date: 2004-09-22 12:50:44 +0200
From: @drstmane
Logged In: YES
user_id=572415
... maybe this whole story tells us, that (at least) all
"maintainers" need to get four more mails everyday: for each
of Stable and Current one mail when the respective testing
is started, and one as soon as it's done ...
## Comment 1943
Date: 2004-09-22 13:01:17 +0200
From: @grobian
Logged In: YES
user_id=963970
in that case I highly prefer a pull mechanism instead of a
push...
## Comment 1944
Date: 2005-11-09 11:10:37 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / DOCUMENTATION
## Comment 1945
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1032467 at http://sourceforge.net/support/tracker.php?aid=1032467
| Testweb fuxx0r3d? | https://api.github.com/repos/MonetDB/MonetDB/issues/359/comments | 0 | 2020-11-30T08:30:48Z | 2024-06-28T07:36:35Z | https://github.com/MonetDB/MonetDB/issues/359 | 753,270,168 | 359 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-22 02:58:33 +0200
From: @drstmane
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2004-09-23 03:55:32 +0200
## Comment 1933
Date: 2004-09-22 02:58:33 +0200
From: @drstmane
Usually, the MIL interpreter sets new BATs (created by
arbitrary operations) read-only to have subsequent
operations benefit from optimizations like BAT-views.
(cf. macro CATCHVAL in src/monet/yytree.mx)
However, this does not hold for new *empty* BATs. I
suppose the reason herefore is, that empty BATs are
assumed to be created by "new()" or "bat()"; these
should obviously not be set read-only.
Though not being a performance issue, not setting empty
result BATs of other operations than "new()"/"bat()" to
read-only is quite "in-convenient", as the behaviour of
subsequent operations differs between empty and
non-empty BATs; especially, since no views can be used,
void columns cannot be propagated for (writeable) empty
BATs.
I'd like to change to detection of (empty) BATs that
were generated by "new()"/"bat()" in CATCHVAL in order
to be able to set other empty result BATs read-only.
However, given my limited knowledge of the MIL
interpreter, I couldn't find a proper way, how to do it.
Maybe, someone else knows more, and can help...
## Comment 1934
Date: 2004-09-23 12:02:36 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Unfortunately, CATCHVAL won;t be the place to do it, because
it is called repeatedly within the interpreter.
If you make all BATs readonly there, then a majority of test
runs is bound to fail (e.g. new(int,int).insert(1,2))
If you detect a direct call to new() and exclude it then it
fails in the following case
new(int,int).insert(1,2).insert(3,4), because CATCHVAL is
called after the first insert. Such cascaded structures
appear everywhere in the system and is not limited to a
small set of operations (insert,delete)
Moreover, a BAT may become temporarily empty in such a
cascaded expression.
The alternative route, changing kernel routines to be more
restrictive, doesn;t help either due to the same situation.
The best approach is to break cascaded expressions and make
the property required explicit by setting the access mode.
[hmmm, sounds like M5]
## Comment 1935
Date: 2004-09-23 14:08:27 +0200
From: @drstmane
Logged In: YES
user_id=572415
If in the current inplementation of CATHCVAL there was only
a check for empty BATs it would fail just the way Martin
described already now; but it does work, now, and this is
IMHO due to the extra check on the BATstamp, ensuring that
BATsetaccess(BAT_READ) is only called for new BATs, but not
for BATs that were modified by insert, delete, replace, etc.
Hence, changing the current behaviour of calling
BATsetaccess(BAT_READ) only on all *new* non-empty BATs to
calling BATsetaccess(BAT_READ) only on all *new* non-empty
BATs that were generated by "new()" or "bat()" should IMHO
work just fine, provided we can find out in CATCHVAL that we
are dealing with "new()" or "bat()"...
Of course, calling "access(BAT_READ)" explicitely in MIL is
always an option --- but IMHO the "uglier" one as long as
CATCHVAL still does it automatically "in some cases"...
## Comment 1936
Date: 2004-09-23 14:26:03 +0200
From: @mlkersten
Logged In: YES
user_id=490798
The option then is to mark the situation in the BAT descriptor,
which is cleared during the first CATCHVAL. SInce a new BAT
is always dirty (TRUE) and we only test dirtyness for being
non-zero you might consider using a special dirty flag.
This flag should be set in the built-in routine for CMDnew()
and CMDbatcast()
## Comment 1937
Date: 2004-09-23 15:55:32 +0200
From: @drstmane
Logged In: YES
user_id=572415
fixed in CVS:
"new()"/"bat()" (CMDnew & CMDnew_lng) now set the (otherwise
unused) second
bit of "batDirty" in the BAT-header, thus identifying the
BAT as newly made;
CATCHVAL now checks this bit, and calls BATsetaccess(BAT_READ)
only if this bit is not set;
CATCHVAL clears the bit in any case.
Martin,
thank you very much for your hint!
(I only don't see, why we'd need to set the flag in
CMDbatcast, to?)
## Comment 1938
Date: 2005-11-09 11:09:09 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: TEST ADDED / SUCCESS
tests/BugDay_2005-11-09_2.9.3/Tests/empty_bats_readonly.SF-1032273.milS
## Comment 1939
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1032273 at http://sourceforge.net/support/tracker.php?aid=1032273
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.
| new *empty* BATs are not set read-only by MIL interpreter | https://api.github.com/repos/MonetDB/MonetDB/issues/358/comments | 0 | 2020-11-30T08:30:45Z | 2024-06-27T10:59:03Z | https://github.com/MonetDB/MonetDB/issues/358 | 753,270,133 | 358 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-21 11:36:42 +0200
From: @drstmane
To: @mlkersten
Version: -- development
Last updated: 2004-09-22 11:14:52 +0200
## Comment 1928
Date: 2004-09-21 23:36:42 +0200
From: @drstmane
[Jan & Stefan]:
(At least) in the latest MIL parser, "if .. else .."
works only if "else" is on the same line as the last
charater of the "if" command or block (see below; or
respective test script in CVS).
We have no idea, whether this did work before the
latest MIL parser changes or not.
========
Monet Database Server V4.4.0_rc2
Copyright (c) 1993-2004, CWI. All rights reserved.
Compiled for i686-redhat-linux-gnu/32bit; dynamically
linked.
Visit http://monetdb.cwi.nl/ for further information.
if ( 1 = 1 ) print(1); else print(2);
[ 1 ]
if ( 1 = 1 )
print(1)
; else
print(2);
[ 1 ]
if ( 1 = 1 )
print(1);
[ 1 ]
else
!ERROR: " else "
! ^
! can't help you here, sorry.
print(2);
if ( 1 = 1 )
{ print(1);
} else
{ print(2); }
[ 1 ]
if ( 1 = 1 )
{ print(1); }
[ 1 ]
else
!ERROR: " else "
! ^
! can't help you here, sorry.
{ print(2); monet>quit();
========
## Comment 1929
Date: 2004-09-22 11:14:52 +0200
From: @mlkersten
Logged In: YES
user_id=490798
In interactive mode the parser builts a tree and when the
end-of-input it reached executes it.
If you enter "if ( expr) stmt;" then you have an almost
unsolveable problem, because to determine if the statement
is finished you have to look ahead to the next token (line)
or evaluate the expressen and be prepared to accept an 'else
stmt'. Conversely, if the expression succeeds you have to
skip a possible else part. Both greatly complicating the
parser infrastructure.
The solution to end the statement with the ELSE on the same
line, if you mean to extend the expression, is an easy solution.
## Comment 1930
Date: 2004-09-23 08:44:47 +0200
From: @arjenpdevries
Logged In: YES
user_id=572978
We do know, and it was not supported:
Monet Database Server V4.3.14c
Copyright (c) 1993-2003, CWI. All rights reserved.
compiled for i686-redhat-linux-gnu/32bit; dynamically linked.
monet>if ( 1 = 1 )
{ print(1);
} else
{ print(2); }
monet>monet>monet>[ 1 ]
monet>if ( 1 = 1 )
monet> print(1);
[ 1 ]
monet>else
monet>print(2);
!ERROR: "print(2);"
! ^
! can't help you here, sorry.
monet> Monet Database Server V4.3.14c
Copyright (c) 1993-2003, CWI. All rights reserved.
compiled for i686-redhat-linux-gnu/32bit; dynamically linked.
monet>if ( 1 = 1 )
{ print(1);
} else
{ print(2); }
monet>monet>monet>[ 1 ]
monet>if ( 1 = 1 )
monet> print(1);
[ 1 ]
monet>else
monet>print(2);
!ERROR: "print(2);"
! ^
! can't help you here, sorry.
Cheers!
## Comment 1931
Date: 2005-11-09 10:51:53 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / Won't fix
Not concidered a bug, ie no test script
## Comment 1932
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1032186 at http://sourceforge.net/support/tracker.php?aid=1032186
| buggy whitespace-handling with "if .. else .." | https://api.github.com/repos/MonetDB/MonetDB/issues/357/comments | 0 | 2020-11-30T08:30:42Z | 2024-06-28T07:36:34Z | https://github.com/MonetDB/MonetDB/issues/357 | 753,270,106 | 357 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-20 10:25:07 +0200
From: Koen van de Sande <<shadowmorpher>>
To: @sjoerdmullender
Version: -- development
Last updated: 2004-09-21 11:26:34 +0200
## Comment 1922
Date: 2004-09-20 22:25:07 +0200
From: Koen van de Sande <<shadowmorpher>>
MKnife.jar in Installer-4.4.0-rc2 has "failed to load
Main-Class manifest attribute from MKnife.jar".
This is a packaging problem (the manifest is incomplete).
Note: Separate MKnife 4.4.0-rc0 does not have this
problem, but is not compatible (not completely I think)
with RC2 (schema changes make some of the queries fail).
## Comment 1923
Date: 2004-09-20 22:40:39 +0200
From: Koen van de Sande <<shadowmorpher>>
Logged In: YES
user_id=270334
The manifest of a working Mknife:
Manifest-Version: 1.0
Created-By: 1.4.2_02 (Sun Microsystems Inc.)
Main-Class: MonetDB.Mknife.main
and the one included in the Windows 4.4-rc2 distro:
Manifest-Version: 1.0
Created-By: 1.4.2_04 (Sun Microsystems Inc.)
So it's just missing a single line.
## Comment 1924
Date: 2004-09-21 11:25:48 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Thanks for the report.
There was a bug in the script that generated the Mknife.jar
file on Windows. The bug is now fixed in CVS.
I also replaced the installer on SourceForge with an updated
version. The only difference being the MANIFEST file inside
Mknife.jar.
## Comment 1925
Date: 2004-09-21 11:26:34 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Closing.
## Comment 1926
Date: 2005-11-09 10:50:13 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / INSTALLATION
## Comment 1927
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1031508 at http://sourceforge.net/support/tracker.php?aid=1031508
| MKnife.jar in Installer-4.4.0-rc2 invalid | https://api.github.com/repos/MonetDB/MonetDB/issues/356/comments | 0 | 2020-11-30T08:30:38Z | 2024-06-27T10:59:01Z | https://github.com/MonetDB/MonetDB/issues/356 | 753,270,079 | 356 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-20 08:43:08 +0200
From: @grobian
To: Fabian Groffen <<fabian>>
Version: -- development
Last updated: 2004-09-26 05:14:16 +0200
## Comment 1917
Date: 2004-09-20 20:43:08 +0200
From: @grobian
the _ character is not recognised properly in the SQL
LIKE clause.
monetdb-> select * from like_test;
+-----+
| str |
+=====+
| |
| t |
| ts |
| tsz |
+-----+
4 rows
monetdb-> select * from like_test where str like 't_';
+----------+
| str |
+==========+
+----------+
0 rows
monetdb->
the last select should have returned the 'ts' record.
## Comment 1918
Date: 2004-09-26 10:24:51 +0200
From: @grobian
Logged In: YES
user_id=963970
the escaped % char is also seen as (unescaped) %.
Trying to fix this...
## Comment 1919
Date: 2004-09-26 17:14:16 +0200
From: @grobian
Logged In: YES
user_id=963970
fixed in release branch
## Comment 1920
Date: 2005-11-09 10:48:46 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: ALREADY IN TESTWEB
src/test/bugs/like-bug-sf-1031435.sql.sql
## Comment 1921
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1031435 at http://sourceforge.net/support/tracker.php?aid=1031435
| SQL: _ char not recognised in LIKE string | https://api.github.com/repos/MonetDB/MonetDB/issues/355/comments | 0 | 2020-11-30T08:30:36Z | 2024-06-27T10:59:00Z | https://github.com/MonetDB/MonetDB/issues/355 | 753,270,046 | 355 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-19 11:21:17 +0200
From: Koen van de Sande <<shadowmorpher>>
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-19 01:34:39 +0200
## Comment 1911
Date: 2004-09-19 11:21:17 +0200
From: Koen van de Sande <<shadowmorpher>>
Acquiring the MonetJDBC.jar driver under Windows is a
major problem, as the binary version is only included
in the MonetDB-SQL-client*.rpm files. The source only
includes Linux makefiles, so building the Jar file
yourself under Windows is troublesome.
Perhaps the binary JDBC driver could get it's own
platform independant distribution instead of residing
in Linux-only RPMs?
## Comment 1912
Date: 2004-09-19 11:46:58 +0200
From: @grobian
Logged In: YES
user_id=963970
Thanks for your comments. I think the jar-file should be
available as separate download.
If you download the source of the SQL tree, you will find an
ant build.xml file which allows you to build the driver
using the ant-tool which is available for windows as well.
Then again, I think we should just make it available for
download as well.
## Comment 1913
Date: 2004-09-19 13:30:49 +0200
From: @drstmane
Logged In: YES
user_id=572415
I made MonetDB's latest JDBC-driver & -client
(MonetJDBC-1.0_rc2.jar) separately availbale for download.
Note, however, that you need to have access to a running
Mserver including SQL support (i.e., MonetDB + MonetDB-SQL)
to be able to use the MonetDB JDBC-driver & -client...
Apart from that, the source tar-ball also contains the
Makefiles for Windows, hence, (provided you have the
required tools installed,) you can also compile the
MonetJDBC.jar on native Windows --- you have to compile the
whole MonetDB-SQL package, though. See the
"HowToStart-SQL-Win32.txt" in the MonetDB-SQL source
tar-ball for details.
## Comment 1914
Date: 2004-09-19 13:34:39 +0200
From: @drstmane
Logged In: YES
user_id=572415
Furthermore, the MonetJDBC.jar wil of course also be
included in the Windows binary installer(s) for MonetDB+SQL
that will be available as of Tuesday, Sep 21 2004.
## Comment 1915
Date: 2005-11-09 10:46:47 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / INSTALLATION
## Comment 1916
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1030614 at http://sourceforge.net/support/tracker.php?aid=1030614
| MonetJDBC.jar only available in RPM files | https://api.github.com/repos/MonetDB/MonetDB/issues/354/comments | 0 | 2020-11-30T08:30:33Z | 2024-06-27T10:58:59Z | https://github.com/MonetDB/MonetDB/issues/354 | 753,270,003 | 354 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-18 12:44:45 +0200
From: @drstmane
To: @mlkersten
Version: -- development
Last updated: 2004-09-18 07:00:14 +0200
## Comment 1904
Date: 2004-09-18 12:44:45 +0200
From: @drstmane
Mserver's "--dbinit" option is implemented as follows:
The MIL statments given with all "--dbinit" options are
gathered in one file, which is then source as if it was
the first MIL script given on Mserver's command line.
Well, as such, there's nothing wrong with this;
BUT:
- the file has a fixed path and name: "/tmp/Mserver.init";
hence, there obviously will be a conflict if two
Mservers "happen to be started concurrently on the same
machine (e.g., once there will be an "Mguardian", or
even be different users...)
- the (supposed to be temporary) file
("/tmp/Mserver.init") is not removed, once it has been
sourced;
hence, if first user A (with standard umask 022)
starts an Mserver with some "--dbinit" and later (but
before some tmpwatch has removed "/tmp/Mserver.init")
user B wants to start an Mserver also with some
"--dbinit", user B's Mserver cannot (over-)write user
A's left-over "/tmp/Mserver.init", resulting no only in
an error message,
but also in user B's Mserver sourcing user A's
left-over "/tmp/Mserver.init" (provided user B is
allowed to read it)!
The small exercise below should make clear, that this
is not only "in-convenient", but highly DANGEROUS!
Solution:
- Use a unique filename for the temporary file
(cf. mkdtemp(3), mktemp(3), tmpnam(3), tempnam(3),
tmpfile(3))
- Make sure, the temporary file is indeed removed, once
it has been sourced.
========
$ echo 'printf("\n ... loading my secret module that
allows me to open a root shell from MIL ... \n\n");' >
/tmp/Mserver.init
$ cat /tmp/Mserver.init
printf("\n ... loading my secret module that allows me
to open a root shell from MIL ... \n\n");
$ chmod a-w,a+r /tmp/Mserver.init
$ ls -l /tmp/Mserver.init
-r--r--r-- 1 manegold ins 97 Sep 18 12:21
/tmp/Mserver.init
manegold@draco:~ $ Mserver --dbfarm=/var/tmp/
--dbinit='printf("\nHello!\n\n");'
ERROR: could not access /tmp/Mserver.init
Monet Database Server V4.4.0_rc1
Copyright (c) 1993-2004, CWI. All rights reserved.
Compiled for i686-redhat-linux-gnu/32bit; dynamically
linked.
Visit http://monetdb.cwi.nl/ for further information.
... loading my secret module that allows me to open a
root shell from MIL ...
monet>
$
========
## Comment 1905
Date: 2004-09-18 14:07:41 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Correct analysis. Actually, this was known to the original
developer, but the implementation was focussed on getting
it work first. There is no scheme in place to capture these
open-end and 'todo' issues in a consistent way.
## Comment 1906
Date: 2004-09-18 18:10:42 +0200
From: @grobian
Logged In: YES
user_id=963970
Some further analysis on the problem
The functions mktemp and tmpname should not be used as they
are not really much more safe than using a fixed filename in
/tmp. The better alternatives mkstemp and tmpfile (which 1.
deletes the file afterwards itself! and 2. is available on
windows as well) return a filepointer instead of a filename.
They do this to overcome the security risks involved when
the file is not kept open (and in exclusive mode) from the
creation moment.
The current hack of --dbinit is just to create a tempfile,
fill it with the commandline data, then adding that file to
the default array of script files to execute. This requires
to add a filename, instead of a pointer, as the actual
reading of the script seems to happen relatively late.
Looking up the filename for a filepointer is not possible in
an elegant way, and absolutely not in a portable way. UNIX
simply was not designed in such a way that a filename can be
resolved for a filepointer. (This is in UNIX a problem
anyway, as there would be for instance a problem which name
to return in case of multiple hard-links for one file.)
Conclusions:
- doing it completely safe requires some deep digging coding
IMHO
- two directions are left on the other hand:
* obsolete/deprecate the feature as you just should create
a file yourself and feed that to Mserver.
* do it, but a bit unsafe, for instance using mktemp, and
cleaning the file just before Mserver terminates.
## Comment 1907
Date: 2004-09-18 18:37:46 +0200
From: @drstmane
Logged In: YES
user_id=572415
what about:
gather all "--dbinit" statements in a (single or array of)
string(s) (say "monet_dbinit") iso. in a tmp. file
convey this this string to "scheduleClient" (either nicely
as a parameter, or as global variable just like
"monet_script", now)
have "monet_dbinit" (resp. "monet_dbinit[*]" interpreted
(directly) by "interpret_str" in "scheduleClient", before
"scheduleClient" calles "scheduleFile" to have the
"monet_script[*]" interpreted by "interpret_file"
(maybe, we could even have a "scheduleStr" in between, just
like "scheduleFile")
I'm not sure, whether this will work as expected, resp.
which "tricks" we need to use to get this working, corretly
respecting things like client contexts, parser/interpreter
states, etc. ...
at least, this would solve the tmp file problem "elegantly"
by avoiding it completely ;-)
## Comment 1908
Date: 2004-09-18 19:00:14 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Removed the dependency on an external temporary file.
Simply called the monet evaluator using the string.
## Comment 1909
Date: 2005-11-09 10:43:50 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / SECURITY
We no longer us a temporary file
## Comment 1910
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1030294 at http://sourceforge.net/support/tracker.php?aid=1030294
| Implementation of "--dbinit" is DANGEROUS! | https://api.github.com/repos/MonetDB/MonetDB/issues/353/comments | 0 | 2020-11-30T08:30:30Z | 2024-06-27T10:58:58Z | https://github.com/MonetDB/MonetDB/issues/353 | 753,269,972 | 353 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-18 10:23:48 +0200
From: @drstmane
To: @sjoerdmullender
Version: -- development
Last updated: 2004-09-24 02:49:43 +0200
## Comment 1899
Date: 2004-09-18 10:23:48 +0200
From: @drstmane
Due to the following code in src/common/monet_utils.mx
and consistent usage of "INLINE" (iso. "inline")
throughout the code, we seem to use function inlining
only with gcc... !
This needs to be checked!
if defined (__GNUC__) && defined (__STDC__)
define INLINE inline
else
define INLINE
endif
(Was introduced by Sjoerd's changes on 2003/08/13.)
## Comment 1900
Date: 2004-09-21 17:03:56 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
A bit unkind to attribute those lines to me. I merely moved
them to monet_utils.mx from gdk.mx. They were in revision
1.1 of that file. ;-)
According to the autoconf documentation, when you use
AC_C_INLINE (which we do), you can just use "inline". It
will get defined to whatever makes the C compiler happy.
The easiest solution seems to be to just put
define INLINE inline
in monet_utils.mx (i.e. without the if stuff).
## Comment 1901
Date: 2004-09-24 14:49:43 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Implemented my suggestion.
## Comment 1902
Date: 2005-11-09 10:34:42 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / COMPILATION
## Comment 1903
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1030261 at http://sourceforge.net/support/tracker.php?aid=1030261
| function inlining only used with gcc! | https://api.github.com/repos/MonetDB/MonetDB/issues/352/comments | 0 | 2020-11-30T08:30:26Z | 2024-06-27T10:58:57Z | https://github.com/MonetDB/MonetDB/issues/352 | 753,269,927 | 352 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-16 11:44:35 +0200
From: @drstmane
To: @sjoerdmullender
Version: -- development
Last updated: 2004-09-21 04:50:23 +0200
## Comment 1893
Date: 2004-09-16 11:44:35 +0200
From: @drstmane
Plain .class files (i.e., those that are not gathered
in a .jar) are not installed into $prefix// under Windows.
## Comment 1894
Date: 2004-09-21 16:50:23 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Fixed in CVS.
## Comment 1895
Date: 2004-09-22 12:18:46 +0200
From: @drstmane
Logged In: YES
user_id=572415
my latest checkin of src/utils/autogen/msc.py should now
indeed fix this bug...
## Comment 1896
Date: 2004-09-22 13:04:38 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
I was convinced I had checked in the fix, but I guess I
missed the fact that CVS complained when I did so.
Anyway, there was one little thing missing in Stefan's fix
(which was identical to part of my fix) for the case that
HAVE_JAVA was not defined. I have now made sure that my
checkin worked.
## Comment 1897
Date: 2005-11-09 10:32:21 +0100
From: @njnes
Logged In: YES
user_id=43556
BugDay_2005-11-09, Niels Nes: NO TEST / INSTALLATION
## Comment 1898
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1029099 at http://sourceforge.net/support/tracker.php?aid=1029099
| autogen(?): .class files do not get installed under Windows | https://api.github.com/repos/MonetDB/MonetDB/issues/351/comments | 0 | 2020-11-30T08:30:23Z | 2024-06-27T10:58:56Z | https://github.com/MonetDB/MonetDB/issues/351 | 753,269,896 | 351 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-16 10:31:23 +0200
From: @drstmane
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-17 07:09:05 +0200
## Comment 1888
Date: 2004-09-16 10:31:23 +0200
From: @drstmane
(sometimes/in some cases?) the color/font encoding
(green,ornage,red/normal,bold,italic) in Mtest.py's
output is not correct: although one of the tests in a
directory shows minor/major differences, the directory
name in the top-most navigation frame stays green.
Note, this is about plain Mtest.py runs, not about the
TestWeb.
## Comment 1889
Date: 2004-09-17 11:40:50 +0200
From: @drstmane
Logged In: YES
user_id=572415
(hopefully) fixed in CVS (MonetDB_4-4-0 branch):
fixed typo that sneaked-in when Niels'\'' added extra result
information
to times.lst on 2004/09/12.
leaving report open until I have time to indeed verify the fix
(ro someone else does this...)
## Comment 1890
Date: 2004-09-17 19:09:05 +0200
From: @drstmane
Logged In: YES
user_id=572415
checked it: fix was ok.
## Comment 1891
Date: 2005-11-09 11:43:25 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: NO TEST / TESTING
## Comment 1892
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1029058 at http://sourceforge.net/support/tracker.php?aid=1029058
| Mtest: results are not properly "aggregated" | https://api.github.com/repos/MonetDB/MonetDB/issues/350/comments | 0 | 2020-11-30T08:30:20Z | 2024-06-27T10:58:55Z | https://github.com/MonetDB/MonetDB/issues/350 | 753,269,864 | 350 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-16 09:40:35 +0200
From: @drstmane
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-18 10:26:44 +0200
## Comment 1883
Date: 2004-09-16 09:40:35 +0200
From: @drstmane
With non-MonetDB packages (i.e., SQL), native Windows
uses wrong pkg*dir for installation: ".../MonetDB" iso.
".../sql" (see below).
Haven't analyzed it any further, yet, but I suppose
this is due to a bug in autogen.
========
$ ls ~monet/Stable/sql/.???_??-?/*/lib/*/Tests/odbcsample1*
/ufs/monet/Stable/sql/.GNU_32-d/Darwin7.5.0/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_32-d/Linux2.4.21-241-smp/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_32-d/Linux2.6.8-1.521smp/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_32-d/linux/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_32-s/CYGWIN1.5.11/lib/sql/Tests/odbcsample1.exe
/ufs/monet/Stable/sql/.GNU_32-s/CYGWIN1.5.7/lib/sql/Tests/odbcsample1.exe
/ufs/monet/Stable/sql/.GNU_32-s/linux/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_64-d/IRIX646.5/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_64-d/Linux2.4.18-e.25smp/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_64-d/Linux2.4.26/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_64-d/linux/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.GNU_64-d/SunOS5.8/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.ntv_32-d/Linux2.4.26/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.ntv_32-d/Linux2.6.8-1.521smp/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.ntv_32-d/linux/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.ntv_32-d/WindowsNT5.0/lib/MonetDB/Tests/odbcsample1.exe
^^^^^^
/ufs/monet/Stable/sql/.ntv_64-d/IRIX646.5/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.ntv_64-d/Linux2.4.18-e.25smp/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.ntv_64-d/linux/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.ntv_64-d/SunOS5.8/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.PGI_64-d/Linux2.4.26/lib/sql/Tests/odbcsample1
/ufs/monet/Stable/sql/.PGI_64-d/linux/lib/sql/Tests/odbcsample1
========
## Comment 1884
Date: 2004-09-17 22:13:16 +0200
From: @drstmane
Logged In: YES
user_id=572415
most probably fixed in CVS:
replaced
pkg = MonetDB
by
pkg = sql
in sql/NT/rules.msc
leaving report open, until tomorrows testing results do
confirm the fix.
## Comment 1885
Date: 2004-09-18 10:26:44 +0200
From: @drstmane
Logged In: YES
user_id=572415
testing confirmed fix.
closing.
## Comment 1886
Date: 2005-11-09 11:41:43 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: NO TEST / COMPILATION
## Comment 1887
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1029031 at http://sourceforge.net/support/tracker.php?aid=1029031
| autogen(?): wrong "pkg*dir" on native Windows | https://api.github.com/repos/MonetDB/MonetDB/issues/349/comments | 0 | 2020-11-30T08:30:17Z | 2024-06-27T10:58:54Z | https://github.com/MonetDB/MonetDB/issues/349 | 753,269,832 | 349 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-15 11:39:51 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-25 07:12:35 +0200
## Comment 1879
Date: 2004-09-15 11:39:51 +0200
From: @grobian
monetdb-> update bond_demo set image = 'file://'||image;
Error: Types varchar(307,0) (str) and varchar(300,0)
(str) are not equal
monetdb-> select max(length(image)) from bond_demo;
+------------------+
| max_length_image |
+==================+
| 69 |
+------------------+
1 row
monetdb-> update bond_demo set image =
cast('file://'||image as varchar(300));
Error: Types varchar(307,0) (str) and varchar(300,0)
(str) are not equal
So, the actual contents of the table will not exceed
the varchar length, but the calculations are not done
on the content, but the signature. Even with a cast
(which should chop off the overflow part) the update
still doesn't work.
## Comment 1880
Date: 2004-09-25 19:12:18 +0200
From: @njnes
Logged In: YES
user_id=43556
fixed by new dynamic type checks. Thanks to stefan for the
[] semantic fixes.
## Comment 1881
Date: 2005-11-09 11:39:55 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/varchar_length_calculation.SF-1028461.sql
## Comment 1882
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1028461 at http://sourceforge.net/support/tracker.php?aid=1028461
| SQL: inappropriate varchar length behaviour | https://api.github.com/repos/MonetDB/MonetDB/issues/348/comments | 0 | 2020-11-30T08:30:15Z | 2024-06-27T10:58:53Z | https://github.com/MonetDB/MonetDB/issues/348 | 753,269,801 | 348 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-14 06:42:48 +0200
From: Koen van de Sande <<shadowmorpher>>
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-14 11:19:41 +0200
## Comment 1875
Date: 2004-09-14 18:42:48 +0200
From: Koen van de Sande <<shadowmorpher>>
At the momet only MonetDB 4.4RC0 is available for
download (and of those only the linux version and
source). There is no prebuilt version of RC0 yet, but I
think that is by choice.
The older stable release, 4.3.18, is no longer
available (I think somebody accidentally made that
release invisible).
## Comment 1876
Date: 2004-09-14 23:19:41 +0200
From: @drstmane
Logged In: YES
user_id=572415
MonetDB-4.3.18 & MonetDB-SQL-2.0.18 are back.
Sorry for the inconvenience.
Please note though, that we do not "promise"/"guarantee" any
support for these anymore!
Users of MonetDB-4.3.18 & MonetDB-SQL-2.0.18 are advised to
upgrade to MonetDB-4.4.0_rc0 & MonetDB-SQL-2.4.0_rc0 (or at
least the respective final releases once they become available).
(Windows binary installers for MonetDB-4.4.0_rc0 &
MonetDB-SQL-2.4.0_rc0 will follow next Monday, Sep. 20 2004.)
## Comment 1877
Date: 2005-11-09 11:32:16 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: NO TEST / DISTRIBUTION
## Comment 1878
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1028040 at http://sourceforge.net/support/tracker.php?aid=1028040
| MonetDB 4.3.x no longer amongst visible downloads | https://api.github.com/repos/MonetDB/MonetDB/issues/347/comments | 0 | 2020-11-30T08:30:12Z | 2024-06-27T10:58:52Z | https://github.com/MonetDB/MonetDB/issues/347 | 753,269,764 | 347 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-13 08:32:57 +0200
From: Jan Rittinger <<jan.rittinger>>
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-17 08:44:15 +0200
## Comment 1869
Date: 2004-09-13 20:32:57 +0200
From: Jan Rittinger <<jan.rittinger>>
var a := 1;
{ var a := a + 1; }
produces an error a unknown because it's translated into
var a; a := 1;
{ var a; a := a + 1; }
In my opinion the two queries are not the same and the
first example should result in a bound to 2 instead of
an error.
## Comment 1870
Date: 2004-09-14 16:07:23 +0200
From: @drstmane
Logged In: YES
user_id=572415
Jan,
IMHO, the two queries are the same, and it is correct that
they do not work --- at least I don't want them to work.
If the two version were considered different, I guess we
agree, that the second is indeed wrong. If then the first
one would work, this would mean, that in a (then) single
statement, "a" would be used in/from to differen contexts
--- this sound very "ugly" to me,
and I don't know, if any other programming language would
allow such things...
Thus, this is IMHO not a "bug" but a "feature", and hence
won't fix/change --- I leave this report open for now,
waiting for further comments.
Stefan
## Comment 1871
Date: 2004-09-14 18:02:13 +0200
From: @MarcinZukowski
Logged In: YES
user_id=607094
I guess the semantic should be that both 'a' become 1.
For me this actually happens, I don't get any error.
## Comment 1872
Date: 2004-09-18 11:27:49 +0200
From: @arjenpdevries
Logged In: YES
user_id=572978
Stefan, Jan,
The two are not identical, for in Jan's example the
semantics are not unambiguous from the statement, whereas in
the translated version it is clear that it should give an
undefined error.
I think I would follow Jan's expectation that the "a" used
after the assignment symbol is the one from the outer scope.
Yet, this would lead to many programming mistakes, exactly
of the type why MIL was being cleaned up, so I'd go with
Stefan in saying it's a feature.
Arjen
## Comment 1873
Date: 2005-11-09 11:29:43 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.9.3/Tests/variable_nesting.SF-1027421.milC
## Comment 1874
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1027421 at http://sourceforge.net/support/tracker.php?aid=1027421
| wrong variable nesting | https://api.github.com/repos/MonetDB/MonetDB/issues/346/comments | 0 | 2020-11-30T08:30:09Z | 2024-06-28T07:36:34Z | https://github.com/MonetDB/MonetDB/issues/346 | 753,269,734 | 346 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-13 07:41:45 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2007-05-29 02:02:52 +0200
## Comment 1861
Date: 2004-09-13 19:41:45 +0200
From: @grobian
$ java -jar jars/MonetJDBC.jar
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.4.0_rc0 (4.4)
Driver: MonetDB Native Driver 0.9 1.0 Release Candidate
2 (0.9)
Type \q to quit, \h for a list of available commands
auto commit mode: on
monetdb-> CREATE SCHEMA "voc";
Operation successful
monetdb-> CREATE USER "voc" WITH PASSWORD 'voc' NAME
'VOC Explorer' SCHEMA "voc";
Error: Protocol violation: header sent before start of
header was issued!
Executed query: CREATE USER "voc" WITH PASSWORD 'voc'
NAME 'VOC Explorer' SCHEMA "voc";
monetdb-> SELECT * FROM "users";
+---------+---------+---------------+----------------+
| name | passwd | fullname | default_schema |
+=========+=========+===============+================+
| monetdb | monetdb | MonetDB Admin | 1004877 |
| voc | voc | VOC Explorer | 1008061 |
+---------+---------+---------------+----------------+
2 rows
monetdb-> \q
$ java -jar jars/MonetJDBC.jar -uvoc
password:
Welcome to the MonetDB
interactive JDBC terminal!
Database: MonetDB 4.4.0_rc0 (4.4)
Driver: MonetDB Native Driver 0.9 1.0 Release Candidate
2 (0.9)
Type \q to quit, \h for a list of available commands
auto commit mode: on
voc-> CREATE TABLE voc_test (id int);
Error: End of stream reached (Mserver still alive?)
Executed query: CREATE TABLE voc_test (id int);
voc->
The running MonetDB instance indeed crashed... probably
a permission/schema maintenance thing?
(btw I don't know how to check in a test for this, only
a JDBC application can do the trick at the moment I think)
## Comment 1862
Date: 2004-09-26 09:53:56 +0200
From: @njnes
Logged In: YES
user_id=43556
The crash is fixed, current problem is now the user/role
isn't getting permission to change the schema (ie add table's)
Current script
create role voc_admin;
CREATE SCHEMA "voc" authorization voc_admin;
CREATE USER "voc" WITH PASSWORD 'voc' NAME 'VOC Explorer'
SCHEMA "voc";
grant voc_admin to voc;
select * from sys.users;
commit;
Then a reconnect as user voc
set role voc_admin;
create table ..
## Comment 1863
Date: 2004-09-26 20:25:07 +0200
From: @njnes
Logged In: YES
user_id=43556
fixed by implementing more of the role support and role
based priviliges.
## Comment 1864
Date: 2005-11-09 11:18:45 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: REASSIGN niels SQL schemata.
## Comment 1865
Date: 2006-11-10 04:20:00 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 1866
Date: 2006-11-29 10:04:10 +0100
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
over-ruling "sf-robot":
re-set to "Pending" since adding a proper test has neither been done nor finally been discarded, yet.
User (submitter) & developer (assignee),
please check, again.
## Comment 1867
Date: 2007-05-29 14:02:52 +0200
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
closing as the VOC demo is part of sql testing.
## Comment 1868
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1027388 at http://sourceforge.net/support/tracker.php?aid=1027388
| SQL: login as different user than monetdb == meltdown! | https://api.github.com/repos/MonetDB/MonetDB/issues/345/comments | 0 | 2020-11-30T08:30:06Z | 2024-06-27T10:58:50Z | https://github.com/MonetDB/MonetDB/issues/345 | 753,269,697 | 345 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-12 09:41:54 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2005-11-09 11:42:07 +0100
## Comment 1857
Date: 2004-09-12 21:41:54 +0200
From: @grobian
sql>create schema voc;
6 querytype
sql>set schema voc;
7 querytype
affected table_name
rows name
mediumint type
0 length
1 tuplecount
1 id
[ 1 ]
sql>create table voctest (id int);
MAPI = monetdb@localhost:45123
ACTION= read_line
QUERY = create table voctest (id int);
ERROR = Connection terminated
the server crashes (Killed) after a schema change and
schema manipulation.
however:
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.3.19 (4.3)
Driver: MonetDB Native Driver 0.9 1.0 Release Candidate
2 (0.9)
Type \q to quit, \h for a list of available commands
auto commit mode: on
monetdb-> create schema voc;
Operation successful
monetdb-> set schema voc;
monetdb-> create table voctest (id int);
Operation successful
monetdb-> \d
TABLE voc.voctest
monetdb->
WITH auto-commit modus (jdbcclient) it DOES work...
## Comment 1858
Date: 2004-09-26 09:39:56 +0200
From: @njnes
Logged In: YES
user_id=43556
fixed in HEAD. Crash was caused by old privilege columns.
## Comment 1859
Date: 2005-11-09 11:05:23 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: TEST ADDED / FAILURE
sql/src/test/BugDay_2005-11-09_2.9.3/Tests/schema_change.SF-1026845.sql
## Comment 1860
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1026845 at http://sourceforge.net/support/tracker.php?aid=1026845
| SQL: schema change makes server prone for kill (meltdown) | https://api.github.com/repos/MonetDB/MonetDB/issues/344/comments | 0 | 2020-11-30T08:30:03Z | 2024-06-27T10:58:49Z | https://github.com/MonetDB/MonetDB/issues/344 | 753,269,661 | 344 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-12 05:36:20 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2006-12-04 09:05:08 +0100
## Comment 1849
Date: 2004-09-12 17:36:20 +0200
From: @grobian
the following scenario does not please me much:
monetdb-> create user voc with password voc name 'voc'
schema voc;
Error: Schema voc doesn't exist
Executed query: create user voc with password voc name
'voc' schema voc;
monetdb-> select * from schemas;
+----------+----------------+---------------+
| id | name | authorization |
+==========+================+===============+
| 1004877 | sys | sqladmin |
| 1005162 | default-schema | sqladmin |
| 1006997 | voc | sqladmin |
+----------+----------------+---------------+
3 rows
monetdb->
From this scenario two things bother me:
1. the password may not be quoted (it's an ident, not a
string)
2. the create does not recognise the voc schema which
is available (see the select afterwards)
3. what is name element, and why does it require a
string (a username was given, wasn't it?)
Passwords might contain spaces or other chars that
might interfere with SQL syntax. Therefore, the
password should be entered as a string.
It might be obvious the CREATE USER statement should
recognise the schemas available in the database.
## Comment 1850
Date: 2004-09-12 18:51:43 +0200
From: @njnes
Logged In: YES
user_id=43556
Fixed the schema bug. Also changed the syntax, ie expect a
password
string. The string after the NAME specifies the full name
for the user.
## Comment 1851
Date: 2005-11-09 10:54:12 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: REASSIGN niels SQL schemata.
## Comment 1852
Date: 2006-11-10 04:20:00 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 1853
Date: 2006-11-29 10:03:59 +0100
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
over-ruling "sf-robot":
re-set to "Pending" since adding a proper test has neither been done nor finally been discarded, yet.
User (submitter) & developer (assignee),
please check, again.
## Comment 1854
Date: 2006-12-04 21:05:08 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
tested by the VOC and skyserver tests. So we can close this.
## Comment 1855
Date: 2006-12-04 21:06:31 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
tested by the VOC and skyserver tests. So we can close this.
## Comment 1856
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1026730 at http://sourceforge.net/support/tracker.php?aid=1026730
| SQL: schema management | https://api.github.com/repos/MonetDB/MonetDB/issues/343/comments | 0 | 2020-11-30T08:30:00Z | 2024-06-27T10:58:48Z | https://github.com/MonetDB/MonetDB/issues/343 | 753,269,626 | 343 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-09 02:21:45 +0200
From: @MarcinZukowski
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-09-09 03:43:43 +0200
## Comment 1843
Date: 2004-09-09 14:21:45 +0200
From: @MarcinZukowski
print(0 - 1 + 1);
produces -2
It is an old issue, but since Peter fixed the
(in)famous -1 problem, maybe this can be fixed as well? :)
## Comment 1844
Date: 2004-09-09 14:32:29 +0200
From: @drstmane
Logged In: YES
user_id=572415
assigned to Peter,
though others are also more than welcome to take care;
who ever start first should first re-assigne it to himself,
indicating that he's busy with it.
## Comment 1845
Date: 2004-09-09 15:43:43 +0200
From: @peterboncz
Logged In: YES
user_id=591107
properly properly, of course they work properly
operator precedence is not going to be implemented.
## Comment 1846
Date: 2004-09-09 15:53:18 +0200
From: @MarcinZukowski
Logged In: YES
user_id=607094
any good reasons why? :)
bison supports it...
## Comment 1847
Date: 2005-11-09 10:43:31 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, sjoerd: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.9.3/Tests/operator_precedence.SF-1025069.milC
## Comment 1848
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1025069 at http://sourceforge.net/support/tracker.php?aid=1025069
| arithmetics still don't work properly | https://api.github.com/repos/MonetDB/MonetDB/issues/342/comments | 0 | 2020-11-30T08:29:56Z | 2024-06-28T07:36:34Z | https://github.com/MonetDB/MonetDB/issues/342 | 753,269,584 | 342 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-09 02:13:42 +0200
From: @MarcinZukowski
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-09-09 03:53:02 +0200
## Comment 1837
Date: 2004-09-09 14:13:42 +0200
From: @MarcinZukowski
This code shows the problem:
<<<
var x,y;
x:=new(int,int).insert(1,1);
x.print;
y:=x.mark(nil).reverse;
y.print;
y:=y.seqbase(100@0);
y.print;
>>>
first y.print output:
<<<
-------------------------
t tmp_198 name
oid int type
-------------------------
[ nil, 1 ]
>>>
then, seqbase doesn't work.
If I use mark(0@0) it creates proper bat.
Hint: if the bat is BAT_READ it also works correctly
## Comment 1838
Date: 2004-09-09 14:32:12 +0200
From: @drstmane
Logged In: YES
user_id=572415
assigned to Peter,
though others are also more than welcome to take care;
who ever start first should first re-assigne it to himself,
indicating that he's busy with it.
## Comment 1839
Date: 2004-09-09 15:42:07 +0200
From: @MarcinZukowski
Logged In: YES
user_id=607094
when closed, please fix tests/BugReports/no.106
## Comment 1840
Date: 2004-09-09 15:53:02 +0200
From: @peterboncz
Logged In: YES
user_id=591107
fixed
of course, this is not a bug, but a feature request anyway
## Comment 1841
Date: 2005-11-09 10:26:21 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
BugDay_2005-11-09, <your name>: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.9.3/Tests/mark_nil_oid.SF-1025062.milC
## Comment 1842
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1025062 at http://sourceforge.net/support/tracker.php?aid=1025062
| mark(nil) produces [any,oid] and not [any,void] | https://api.github.com/repos/MonetDB/MonetDB/issues/341/comments | 0 | 2020-11-30T08:29:53Z | 2024-06-28T07:36:33Z | https://github.com/MonetDB/MonetDB/issues/341 | 753,269,548 | 341 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-09 12:21:21 +0200
From: Jens Teubner <<teubner>>
To: @njnes
Version: -- development
Last updated: 2004-09-13 04:32:48 +0200
## Comment 1826
Date: 2004-09-09 12:21:21 +0200
From: Jens Teubner <<teubner>>
I just tried to installed CVS versions of MonetDB and the SQL frontend as of
today. (On Gentoo Linux, Pentium M, with `--enable-optimize')
MonetDB compiled and installed fine. SQL also compiled fine. Installation,
however, gives me
/bin/install -c benchblockmode.html benchblockmode.ps benchblockmode.
pdf /opt/MonetDB-cvs/share/MonetDB/docs/SQL/jdbc/
/bin/install: cannot stat `benchblockmode.html': No such file or directory
make[4]: *** [install-data-local-bench] Error 1
make[4]: Leaving directory `/home/teubner/work/monet/sql/src/jdbc'
and equivalently for ./src/jdbc/jdbcmanual.html and ./SQLfeatures.html.
By the way: What is `split 0 -no_images -info 0 -no_subdir XXX.tex'
supposed to be doing. My machine only says `split: invalid option -- n'
## Comment 1827
Date: 2004-09-09 12:51:44 +0200
From: @drstmane
Logged In: YES
user_id=572415
Jens,
I noticed this problem yesterday, too, and it seems to be
caused by problems with make rules and/or detection of tools
(latex2html in this case) during configure.
`split 0 -no_images -info 0 -no_subdir XXX.tex'
results from rules.mk's
"$$ -split 0 -no_mages -info 0 -no_subdir $(<F)"
with $(LATEX2HTML) being empty;
apparently, your system does not have latex2html, or
configure fails to find it, AND our lates make rules do not
skip the doc-creation properly in this case (anymore; I
thing it used to work...).
unfortunately, we're currently too busy to take care of this
immediately (i.e., this week). our apologies for the
inconveniences!
for now, you could simply remove the doc_* sections from
sql/Makefile.ag and sql/src/jdbc/Makefile.ag.
alternatively, you could try-out the versions of both
MonetDB and SQL that are tagged "PetersMILcleanup_before" or
"BeforeBoston2_Current", but I cannot guarantee that they do
work better...
Stefan
## Comment 1828
Date: 2004-09-09 12:54:33 +0200
From: @grobian
Logged In: YES
user_id=963970
since the generation of the HTML files fails, it is
reasonable that they cannot be installed.
Do you have latex2html installed? (check emerge -s
latex2html) By default Gentoo does not ship you with any
TeX tools, do you have it installed?
the split 0 -no... should be the arguments of latex2html.
Will look into that, if you haven't emerged latex2html
and/or TeTeX, then the document creation should not happen
at all.
Please send me your output for emerge -p latex2html
## Comment 1829
Date: 2004-09-09 13:39:56 +0200
From: Jens Teubner <<teubner>>
Logged In: YES
user_id=731390
Okay, I haven't installed latex2html, so your interpretations of the problems make
sense.
Stefan, don't worry, I found ways to get the stuff installed on my system. Of
course somebody should still fix the detection of latex2html in ./configure at
some point in time.
(Unfortunately, although I got the SQL frontend installed on my system, nothing is
actually running here. Jan just sent me Peter's commit messages from yesterday.
His changes really seem to have broken everything. I'm happy that it is not my job
to fix all the stuff that has been broken by this...)
## Comment 1830
Date: 2004-09-09 13:46:00 +0200
From: @drstmane
Logged In: YES
user_id=572415
ok.
sure, we'll eventually take care of the latex2html
detection/handling issue; hence thanks for the bug report!
but now you know, why we (Peter, Niels, Sjoerd, Marcin, me)
don't have any time right now...
tag "PetersMILcleanup_before" should explain itself...
## Comment 1831
Date: 2004-09-10 13:18:22 +0200
From: @grobian
Logged In: YES
user_id=963970
update:
SQL now builds perfectly on my Gentoo system with latex2html
installed,
but fails to build on Fedora Core 2 with latex2html
installed... :?
It seems to detect it, but the latex2html is never called on
the Fedora machine (pictor) thus no html file generation.
The problems seems to persist in a double declaration of the
html target in the Makefile.
On the (clean compiling) Gentoo machine, the Makefile in the
build dir does not include the html target, so the html
target from html.mk is used, pictor on the other hand has a
html target defined which has dependency html-recursive.
This way the right html target from html.mk is never called
I think.
## Comment 1832
Date: 2004-09-12 15:30:40 +0200
From: @njnes
Logged In: YES
user_id=43556
make DOCs only when
latex,latex2html,pdflatex,fig2dev and dvips exist
Jens could you test this?
## Comment 1833
Date: 2004-09-13 15:40:39 +0200
From: @grobian
Logged In: YES
user_id=963970
pictor does still not manage to install CVS:
/usr/bin/install -c SQLfeatures.html SQLfeatures.ps
SQLfeatures.pdf
/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/program/share/MonetDB/docs/SQL
/usr/bin/install: cannot stat `SQLfeatures.html': No such
file or directory
make[2]: *** [install-data-local-sql] Error 1
make[2]: Leaving directory
`/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/build'
make[1]: *** [install-am] Error 2
make[1]: Leaving directory
`/net/pictor.ins.cwi.nl/export/scratch1/fabian/monet/sql/build'
make: *** [install-recursive] Error 1
## Comment 1834
Date: 2004-09-14 10:10:16 +0200
From: Jens Teubner <<teubner>>
Logged In: YES
user_id=731390
Niels' fix works for me. Docs are no longer
installed here, as I don't have latex2html.
## Comment 1835
Date: 2005-11-09 10:37:27 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: NO TEST / COMPILATION
## Comment 1836
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1025005 at http://sourceforge.net/support/tracker.php?aid=1025005
| Install fails due to missing .html files | https://api.github.com/repos/MonetDB/MonetDB/issues/340/comments | 0 | 2020-11-30T08:29:50Z | 2024-06-27T10:58:45Z | https://github.com/MonetDB/MonetDB/issues/340 | 753,269,514 | 340 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-08 11:26:24 +0200
From: @njnes
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-09-09 10:37:16 +0200
## Comment 1822
Date: 2004-09-08 23:26:24 +0200
From: @njnes
Seems the $(i+1) fails after the resent mil cleanup,
mkey.mx/ds_link
monet>ds_link(bat("user_role_grantor_1"),bat("user_role_grantor_1"),bat("user_role_grantor_1"),bat("user_role_grantor_1"));
!ERROR: interpret: no matching MIL operator to 'i()'.
!ERROR: <(param 1): evaluation error.
## Comment 1823
Date: 2004-09-09 10:37:16 +0200
From: @peterboncz
Logged In: YES
user_id=591107
fixed
(i < 1)
did not work, due to the context-aware parser not knowing
about the '<' a starter char for MIL operators.
## Comment 1824
Date: 2005-11-09 10:36:43 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-11-09_4.8/Tests/MIL_var_args.SF-1024691.*
## Comment 1825
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1024691 at http://sourceforge.net/support/tracker.php?aid=1024691
| problem with var args | https://api.github.com/repos/MonetDB/MonetDB/issues/339/comments | 0 | 2020-11-30T08:29:47Z | 2024-06-28T07:36:33Z | https://github.com/MonetDB/MonetDB/issues/339 | 753,269,477 | 339 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-08 09:24:18 +0200
From: @grobian
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-12 05:47:28 +0200
## Comment 1815
Date: 2004-09-08 21:24:18 +0200
From: @grobian
The following error turns up on various JDBC queries...
My previous report about this was closed as invalid,
but after testing on tree different setups and seeing
it happening more and more, I am sure the bug DOES exist.
The error message is:
ERROR: CMDsort: returned BAT has unexpected head.
ERROR: reverse(param 1): evaluation error.
For instance JdbcClient's most important function fails
(\d <obj>).
One of the queries that seem to be lethal is:
SELECT cast(null AS varchar) AS "TABLE_CAT",
"schemas"."name" AS "TABLE_SCHEM", "tables"."name" AS
"TABLE_NAME", "keycolumns"."column" AS "COLUMN_NAME",
"keys"."type" AS "KEY_SEQ", "keys"."name" AS "PK_NAME"
FROM "sys"."keys" AS "keys", "sys"."keycolumns" AS
"keycolumns", "sys"."tables" AS "tables",
"sys"."schemas" AS "schemas" WHERE "keys"."id" =
"keycolumns"."id" AND "keys"."table_id" = "tables"."id"
AND "tables"."schema_id" = "schemas"."id" AND
"keys"."type" = 0 AND "schemas"."name" LIKE 'sys' AND
"tables"."name" LIKE 'x' ORDER BY "COLUMN_NAME"
## Comment 1816
Date: 2004-09-09 22:41:01 +0200
From: @drstmane
Logged In: YES
user_id=572415
is there a test in CVS that triggers the bug?
## Comment 1817
Date: 2004-09-11 21:08:15 +0200
From: @grobian
Logged In: YES
user_id=963970
the bug is reproducable with MapiClient, commiting above
mentioned query in CVS.
## Comment 1818
Date: 2004-09-12 16:20:26 +0200
From: @njnes
Logged In: YES
user_id=43556
Following mil code also produces this error, ie. its not sql
specific
var b2 := bat(void,str);
b2.seqbase(oid(10));
var b3 := mirror(b2);
var b4 := b3.reverse().join(b2);
b4.reverse().sort().reverse().print();
## Comment 1819
Date: 2004-09-12 17:47:28 +0200
From: @drstmane
Logged In: YES
user_id=572415
fixed in CVS:
empty join results should only be marked dense,
if they are type "void" or "oid" ...
## Comment 1820
Date: 2005-10-06 15:28:57 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
test script available as:
MonetDB/tests/BugsViaSourgeforce/Tests/ID.1024615.*
sql/src/test/bugs/Tests/select_orderby_alias-bug-sf-1024615.*
sql/src/test/bugs/select_orderby_alias-bug-sf-1024615.sql
## Comment 1821
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1024615 at http://sourceforge.net/support/tracker.php?aid=1024615
| SQL: CMDsort error | https://api.github.com/repos/MonetDB/MonetDB/issues/338/comments | 0 | 2020-11-30T08:29:44Z | 2024-06-27T10:58:43Z | https://github.com/MonetDB/MonetDB/issues/338 | 753,269,445 | 338 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-08 12:02:35 +0200
From: bdekker <<bdekker>>
To: @mlkersten
Version: -- development
Last updated: 2004-09-12 01:51:54 +0200
## Comment 1811
Date: 2004-09-08 12:02:35 +0200
From: bdekker <<bdekker>>
BATrename returns BBPRENAME_LONG for logical names
longer than IDLENGTH. IDLENGTH is 64 bytes. Even
names shorter than 64 bytes are not allowed since the
length of the directory name is added to the length of
the (logical) name of the bat.
## Comment 1812
Date: 2004-09-12 13:51:54 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Indeed, each subdirectory costs an additional 3 characters.
Changing this is currently not acceptable, because it would
invalidate the BAT descriptor, thereby causing major backward
compatibility issues.
## Comment 1813
Date: 2005-11-09 10:24:16 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: NO TEST
source code insufficiency that is set/said to "Wont Fix".
## Comment 1814
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1024266 at http://sourceforge.net/support/tracker.php?aid=1024266
| BBPrename restricts logical names to IDLENGTH | https://api.github.com/repos/MonetDB/MonetDB/issues/337/comments | 0 | 2020-11-30T08:29:41Z | 2024-06-28T07:36:33Z | https://github.com/MonetDB/MonetDB/issues/337 | 753,269,405 | 337 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 10:19:27 +0200
From: @drstmane
To: Arjan Scherpenisse <<arjan>>
Version: -- development
Last updated: 2004-09-09 10:40:13 +0200
## Comment 1807
Date: 2004-09-07 22:19:27 +0200
From: @drstmane
obviously, I cannot write in /usr/lib/php4/.
all installs should go to (some subdirectory of)
$MONETDB_PREFIX ...
make[5]: Entering directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php/Cimpl'
make[6]: Entering directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php/Cimpl'
make[6]: Nothing to be done for `install-exec-am'.
/bin/sh
/ufs/manegold/Monet/4.3/MonetDB/conf/mkinstalldirs
/usr/lib/php4
/bin/sh ../../../../../libtool --mode=install
/usr/bin/install -c monetdb.la /usr/lib/php4/monetdb.la
libtool: install: warning: relinking `monetdb.la'
(cd
/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php/Cimpl;
/bin/sh ../../../../../libtool --mode=relink gcc -g -O2
-ansi -std=c99 -Wall -W -DHWCOUNTERS -DHW_Linux
-DHW_i686 -DCOMPILE_DL_MONETDB=1 -D_GNU_SOURCE
-D_REENTRANT -o monetdb.la -rpath /usr/lib/php4 -module
php_monetdb.lo ../../C/libMapi.la )
gcc -shared php_monetdb.lo -Wl,--rpath
-Wl,/var/tmp/_PREFIX_MONETDB_GNU_32_d__/lib
-L/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/common/.libs
-L/var/tmp/_PREFIX_MONETDB_GNU_32_d__/lib -lMapi
-Wl,-soname -Wl,monetdb.so.0 -o .libs/monetdb.so.0.0.0
/usr/bin/install -c .libs/monetdb.so.0.0.0T
/usr/lib/php4/monetdb.so.0.0.0
/usr/bin/install: cannot create regular file
`/usr/lib/php4/monetdb.so.0.0.0': Permission denied
make[6]: *** [install-monetdbLTLIBRARIES] Error 1
make[6]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php/Cimpl'
make[5]: *** [install-am] Error 2
make[5]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php/Cimpl'
make[4]: *** [install-recursive] Error 1
make[4]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php'
make[3]: *** [install-recursive] Error 1
make[3]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients'
make[2]: *** [install-recursive] Error 1
make[2]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi'
make[1]: *** [install-recursive] Error 1
make[1]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src'
make: *** [install-recursive] Error 1
## Comment 1808
Date: 2004-09-09 22:40:13 +0200
From: @drstmane
Logged In: YES
user_id=572415
Arjan fixed the bug.
the open issue of how to get monetdb.so to /usr/lib/php4/ is
documented in a thread on the Monetdb-develpoers list.
## Comment 1809
Date: 2005-11-09 10:20:45 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: NO TEST / COMPILATION
## Comment 1810
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023957 at http://sourceforge.net/support/tracker.php?aid=1023957
| "make install" fails for src/mapi/clients/php/Cimpl | https://api.github.com/repos/MonetDB/MonetDB/issues/336/comments | 0 | 2020-11-30T08:29:38Z | 2024-06-27T10:58:41Z | https://github.com/MonetDB/MonetDB/issues/336 | 753,269,379 | 336 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 06:42:03 +0200
From: @MarcinZukowski
To: @mlkersten
Version: -- development
Last updated: 2004-09-09 09:59:18 +0200
## Comment 1801
Date: 2004-09-07 18:42:03 +0200
From: @MarcinZukowski
if table() gets >15 parameters, segfault occurs.
test script provided.
## Comment 1802
Date: 2004-09-09 21:00:03 +0200
From: @drstmane
Logged In: YES
user_id=572415
apparently, this holds for all procs/commands with variable
number of parameters ... needs to be fixed!
## Comment 1803
Date: 2004-09-09 21:59:18 +0200
From: @mlkersten
Logged In: YES
user_id=490798
The kernel module did not test for array overflow while
handling variable argument lists.
## Comment 1804
Date: 2004-09-09 22:37:50 +0200
From: @drstmane
Logged In: YES
user_id=572415
Martin,
thank you very much for the bug fix AND the respective test!
I felt free, though, to approve also the intended/error
messages... ;-)
## Comment 1805
Date: 2005-10-06 15:27:49 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
test script available as:
MonetDB/tests/BugsViaSourgeforce/Tests/ID.1023816.*
## Comment 1806
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023816 at http://sourceforge.net/support/tracker.php?aid=1023816
| table() argument count limited | https://api.github.com/repos/MonetDB/MonetDB/issues/335/comments | 0 | 2020-11-30T08:29:35Z | 2024-06-27T10:58:40Z | https://github.com/MonetDB/MonetDB/issues/335 | 753,269,352 | 335 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 03:22:47 +0200
From: @drstmane
To: Arjan Scherpenisse <<arjan>>
Version: -- development
Last updated: 2004-09-07 05:28:50 +0200
## Comment 1797
Date: 2004-09-07 15:22:47 +0200
From: @drstmane
with freshly updated CVS tree, compilation from scratch
fails for src/mapi/clients/php/Cimpl (RH 9, gcc-3.2.2,
standard configure):
make[7]: Entering directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php/Cimpl'
/bin/sh ../../../../../libtool --mode=compile gcc
-DHAVE_CONFIG_H -I.
-I/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl
-I../../../../..
-I/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl
-I../../C -I -I/usr/include/php -I/usr/include/php/main
-I/usr/include/php/Zend -I/usr/include/php/TSRM
-I/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/
-I/usr/include/php -I/usr/include/php/main
-I/usr/include/php/Zend -I/usr/include/php/TSRM -g
-O2 -ansi -std=c99 -Wall -W -DHWCOUNTERS -DHW_Linux
-DHW_i686 -DCOMPILE_DL_MONETDB=1 -D_BSD_SOURCE
-D_REENTRANT -c -o php_monetdb.lo `test -f
'/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c'
|| echo
'/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/'`/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c
gcc -DHAVE_CONFIG_H -I.
-I/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl
-I../../../../..
-I/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl
-I../../C -I -I/usr/include/php -I/usr/include/php/main
-I/usr/include/php/Zend -I/usr/include/php/TSRM
-I/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/
-I/usr/include/php -I/usr/include/php/main
-I/usr/include/php/Zend -I/usr/include/php/TSRM -g -O2
-ansi -std=c99 -Wall -W -DHWCOUNTERS -DHW_Linux
-DHW_i686 -DCOMPILE_DL_MONETDB=1 -D_BSD_SOURCE
-D_REENTRANT -c
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c
-fPIC -DPIC -o php_monetdb.lo
In file included from /usr/include/php/Zend/zend.h:180,
from /usr/include/php/main/php.h:34,
from
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:24:
/usr/include/php/Zend/zend_hash.h:191: syntax error
before "ulong"
In file included from
/usr/include/php/Zend/zend_globals.h:30,
from /usr/include/php/Zend/zend_list.h:25,
from /usr/include/php/Zend/zend_API.h:26,
from /usr/include/php/main/php.h:38,
from
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:24:
/usr/include/php/Zend/zend_ptr_stack.h:41: syntax error
before "void"
/usr/include/php/Zend/zend_ptr_stack.h:51: syntax error
before "void"
In file included from /usr/include/php/Zend/zend_API.h:28,
from /usr/include/php/main/php.h:38,
from
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:24:
/usr/include/php/Zend/zend_operators.h:61: syntax error
before "int"
In file included from /usr/include/php/Zend/zend_API.h:30,
from /usr/include/php/main/php.h:38,
from
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:24:
/usr/include/php/Zend/zend_execute.h:57: syntax error
before "void"
/usr/include/php/Zend/zend_execute.h:67: syntax error
before "int"
/usr/include/php/Zend/zend_execute.h:107: syntax error
before "void"
/usr/include/php/Zend/zend_execute.h:119: syntax error
before "int"
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zm_startup_monetdb':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:172:
warning: unused parameter `type'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zm_shutdown_monetdb':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:192:
warning: unused parameter `type'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zm_activate_monetdb':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:202:
warning: unused parameter `type'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:202:
warning: unused parameter `module_number'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zm_deactivate_monetdb':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:211:
warning: unused parameter `type'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:211:
warning: unused parameter `module_number'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_connect':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:254:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:254:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_close':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:391:
warning: empty body in an if-statement
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:365:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:365:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_query':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:443:
warning: implicit declaration of function
`php_error_docref'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:409:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:409:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_num_rows':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:485:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:485:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_num_fields':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:508:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:508:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_next_result':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:532:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:532:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_field_name':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:564:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:564:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_field_type':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:603:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:603:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_errno':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:642:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:642:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_error':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:671:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:671:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `php_monetdb_fetch_hash':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:708:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:708:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_fetch_object':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:820:
`ZEND_STANDARD_CLASS_DEF_PTR' undeclared (first use in
this function)
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:820:
(Each undeclared identifier is reported only once
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:820:
for each function it appears in.)
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_data_seek':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:845:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:845:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_escape_string':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:868:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:868:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_ping':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:890:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:890:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_free_result':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:921:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:921:
warning: unused parameter `return_value_used'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:
In function `zif_monetdb_info':
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:942:
warning: unused parameter `this_ptr'
/ufs/manegold/Monet/4.3/MonetDB/src/mapi/clients/php/Cimpl/php_monetdb.c:942:
warning: unused parameter `return_value_used'
make[7]: *** [php_monetdb.lo] Error 1
make[7]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php/Cimpl'
make[6]: *** [all-recursive] Error 1
make[6]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients/php'
make[5]: *** [all-recursive] Error 1
make[5]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi/clients'
make[4]: *** [all-recursive] Error 1
make[4]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi'
make[3]: *** [all] Error 2
make[3]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src/mapi'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__/src'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory
`/var/tmp/_BUILD_MONETDB_GNU_32_d__'
make: *** [all] Error 2
## Comment 1798
Date: 2004-09-07 17:28:49 +0200
From: @arjan
Logged In: YES
user_id=20087
Included monetdb_config.h so inline gets properly defined
Also modified code so it works with PHP versions below 4.3.
We should have a check on the PHP version in the configure
script, ultimately.
## Comment 1799
Date: 2005-11-09 10:18:47 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: NO TEST / COMPILATION
## Comment 1800
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023672 at http://sourceforge.net/support/tracker.php?aid=1023672
| compilation of src/mapi/clients/php/Cimpl fails | https://api.github.com/repos/MonetDB/MonetDB/issues/334/comments | 0 | 2020-11-30T08:29:31Z | 2024-06-28T07:36:33Z | https://github.com/MonetDB/MonetDB/issues/334 | 753,269,312 | 334 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 02:59:27 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-07 04:23:11 +0200
## Comment 1793
Date: 2004-09-07 14:59:27 +0200
From: @grobian
monetdb-> SELECT * FROM "keys", "keycolumns", "tables",
"schemas" WHERE "keys"."id" = "keycolumns"."id" AND
"keys"."table_id" = "tables"."id" AND
"tables"."schema_id" = "schemas"."id" AND "keys"."type"
= 0 ORDER BY "keycolumns"."column";
Error: ERROR: CMDsort: returned BAT has unexpected head.
ERROR: reverse(param 1): evaluation error.
Without the ORDER BY:
monetdb-> SELECT * FROM "keys", "keycolumns", "tables",
"schemas" WHERE "keys"."id" = "keycolumns"."id" AND
"keys"."table_id" = "tables"."id" AND
"tables"."schema_id" = "schemas"."id" AND "keys"."type"
= 0;
+----------+----------+----------+----------+----------+----------+----------+----------+----------+----------+----------+-----------+----------+---------+--------+-------+------------+----------+----------+---------------+
| id | table_id | type | name | rkey
| id | column | trunc | nr | id
| name | schema_id | query | istable | system |
clear | temporary | id | name | authorization |
+==========+==========+==========+==========+==========+==========+==========+==========+==========+==========+==========+===========+==========+=========+========+=======+============+==========+==========+===============+
+----------+----------+----------+----------+----------+----------+----------+----------+----------+----------+----------+-----------+----------+---------+--------+-------+------------+----------+----------+---------------+
0 rows
monetdb->
JdbcClient uses this query to "describe" a table, and
currently fails to do so since this query can't be
executed anymore.
(this is ANNOYING, b.t.w.)
## Comment 1794
Date: 2004-09-07 16:23:11 +0200
From: @njnes
Logged In: YES
user_id=43556
Cannot reproduce the problem on the HEAD MonetDB/SQL
## Comment 1795
Date: 2005-11-09 10:17:57 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.8/Tests/ORDER_BY_evaluation_error.SF-1023658.*
## Comment 1796
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023658 at http://sourceforge.net/support/tracker.php?aid=1023658
| SQL: ORDER BY clause causes evaluation error | https://api.github.com/repos/MonetDB/MonetDB/issues/333/comments | 0 | 2020-11-30T08:29:28Z | 2024-06-28T07:36:32Z | https://github.com/MonetDB/MonetDB/issues/333 | 753,269,262 | 333 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 02:47:56 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-08 12:35:53 +0200
## Comment 1789
Date: 2004-09-07 14:47:56 +0200
From: @grobian
The savepoints JDBC test now no longer fails on BAT key
dense, but on a BATins.
FAILED :( Unknown table table_test_csavepoints
ERROR: BATins: access denied to tmp_419, aborting.
the performed test is like:
create table x (id int);
savepoint savepoint_1;
insert into x values (1);
select * from x;
rollback to savepoint savepoint_1;
select * from x;
Seems like the rollback, rolls too much back.
## Comment 1790
Date: 2004-09-08 00:35:52 +0200
From: @njnes
Logged In: YES
user_id=43556
savepoint names where save in the active transaction, now we
save it
in the parent. Also dropping objects wasn't rolled forward
from savepoint to
savepoint (to central catalog)
## Comment 1791
Date: 2005-11-09 10:11:17 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: ALREADY IN TESTWEB
sql/src/test/bugs/Tests/savepoint_rollback-bug-sf-1023652
## Comment 1792
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023652 at http://sourceforge.net/support/tracker.php?aid=1023652
| SQL: ROLLBACK TO SAVEPOINT broken? | https://api.github.com/repos/MonetDB/MonetDB/issues/332/comments | 0 | 2020-11-30T08:29:25Z | 2024-06-27T10:58:37Z | https://github.com/MonetDB/MonetDB/issues/332 | 753,269,230 | 332 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 02:37:46 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-09-07 02:48:41 +0200
## Comment 1785
Date: 2004-09-07 14:37:46 +0200
From: @arjan
SQL should check if an insert or update is applied to a
view (as I did by mistake), and should return an error
(for now). Currently the server crashes on both updates
and inserts on views.
## Comment 1786
Date: 2004-09-07 14:48:41 +0200
From: @njnes
Logged In: YES
user_id=43556
added checks
## Comment 1787
Date: 2005-11-09 10:06:51 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-11-09, Stefan: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-11-09_2.8/Tests/updating_views.SF-1023645.*
## Comment 1788
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023645 at http://sourceforge.net/support/tracker.php?aid=1023645
| SQL: Modification of views | https://api.github.com/repos/MonetDB/MonetDB/issues/331/comments | 0 | 2020-11-30T08:29:22Z | 2024-06-27T10:58:36Z | https://github.com/MonetDB/MonetDB/issues/331 | 753,269,196 | 331 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 02:35:41 +0200
From: @drstmane
To: @njnes
Version: -- development
Last updated: 2004-09-07 03:04:13 +0200
## Comment 1782
Date: 2004-09-07 14:35:41 +0200
From: @drstmane
the output of tests src/test/Tests/{count,null} needs
to be checked and approved; cf.
http://www.ins.cwi.nl/intern/TestWeb/Current/sql/.mTests103/GNU_32-d_Linux2.4.20-31.9smp/src_test/count.out.00.html
http://www.ins.cwi.nl/intern/TestWeb/Current/sql/.mTests103/GNU_32-d_Linux2.4.20-31.9smp/src_test/null.out.00.html
## Comment 1783
Date: 2005-10-06 22:49:49 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
BugDay_2005-10-06: NO TEST (not necessary)
## Comment 1784
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023642 at http://sourceforge.net/support/tracker.php?aid=1023642
| SQL: check/approve output of src/test/Tests/{count,null} | https://api.github.com/repos/MonetDB/MonetDB/issues/330/comments | 0 | 2020-11-30T08:29:20Z | 2024-06-27T10:58:35Z | https://github.com/MonetDB/MonetDB/issues/330 | 753,269,171 | 330 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 12:40:10 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-07 05:20:28 +0200
## Comment 1775
Date: 2004-09-07 12:40:10 +0200
From: @grobian
monetdb-> \d
TABLE sys.craftsmen
TABLE sys.impotenten
TABLE sys.invoices
TABLE sys.passengers
TABLE sys.seafarers
TABLE sys.soldiers
TABLE sys.total
TABLE sys.voyages
monetdb-> update voyages set master = trim(master);
8141 affected rows
monetdb-> select master from voyages where number=7948;
+----------+
| master |
+==========+
| Volkers |
+----------+
1 row
monetdb-> select trim(master) from voyages where
number=7948;
+-------------+
| trim_master |
+=============+
| Volkers |
+-------------+
1 row
monetdb->
Even though the update should have replaced each value
from the master column with the trimmed version, it
seems it just has neglected to do so; which the two
selects prove...
## Comment 1776
Date: 2004-09-07 12:41:38 +0200
From: @grobian
Logged In: YES
user_id=963970
for web reference (HTML reduces multiple spaces and combines
them into one):
the first select returns Volkers with a space in front, the
second select returns Volkers without that space.
## Comment 1777
Date: 2004-09-07 15:00:55 +0200
From: @grobian
Logged In: YES
user_id=963970
this bug only seems to work for the VOC dataset, not for
small tests. Due to some other 'bugs' further downtracing
of this problem takes some extra time...
## Comment 1778
Date: 2004-09-07 16:05:55 +0200
From: @grobian
Logged In: YES
user_id=963970
this bug is auto-commit related. test script checked in in CVS.
## Comment 1779
Date: 2004-09-07 17:20:28 +0200
From: @njnes
Logged In: YES
user_id=43556
forwarding of updates to catalog was broke
## Comment 1780
Date: 2005-10-06 15:26:57 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
test script available as:
sql/src/test/bugs/Tests/trimtest-bug-sf-1023588.*
sql/src/test/bugs/trimtest-bug-sf-1023588.sql
## Comment 1781
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023588 at http://sourceforge.net/support/tracker.php?aid=1023588
| SQL: UPDATE does not update table? | https://api.github.com/repos/MonetDB/MonetDB/issues/329/comments | 0 | 2020-11-30T08:29:16Z | 2024-06-27T10:58:34Z | https://github.com/MonetDB/MonetDB/issues/329 | 753,269,134 | 329 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 11:04:46 +0200
From: @drstmane
To: @sjoerdmullender
Version: -- development
Last updated: 2004-09-07 01:11:34 +0200
## Comment 1772
Date: 2004-09-07 11:04:46 +0200
From: @drstmane
make[5]: Entering directory
`/var/tmp/monet.ntv_32-d.10134/sql/Linux2.4.26/src/backends/monet5'
/bin/sh ../../../libtool --mode=compile icc -no-gcc
-DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10134/sql/src/backends/monet5
-I../../..
-I/var/tmp/monet.ntv_32-d.10134/sql/src/backends/monet5
-I../
icc -no-gcc -DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10134/sql/src/backends/monet5
-I../../..
-I/var/tmp/monet.ntv_32-d.10134/sql/src/backends/monet5
-I../../common -I/var/tmp/monet.ntv_32-d.1013
/var/tmp/monet.ntv_32-d.10134/sql/src/backends/monet5/sql_gencode.mx(952):
error 593: variable "iconv" was set but never used
int codeset,iconv;
^
compilation aborted for sql_gencode.c (code 2)
make[5]: *** [sql_gencode.lo] Error 1
make[5]: Entering directory
`/var/tmp/monet.GNU_64-d.10134/sql/Linux2.4.26/src/backends/monet5'
/bin/sh ../../../libtool --mode=compile gcc -m64
-DHAVE_CONFIG_H -I.
-I/var/tmp/monet.GNU_64-d.10134/sql/src/backends/monet5
-I../../..
-I/var/tmp/monet.GNU_64-d.10134/sql/src/backends/monet5
-I../../
gcc -m64 -DHAVE_CONFIG_H -I.
-I/var/tmp/monet.GNU_64-d.10134/sql/src/backends/monet5
-I../../..
-I/var/tmp/monet.GNU_64-d.10134/sql/src/backends/monet5
-I../../common -I/var/tmp/monet.GNU_64-d.10134/s
/var/tmp/monet.GNU_64-d.10134/sql/src/backends/monet5/sql_gencode.mx:
In function `backend_dumpstmt':
/var/tmp/monet.GNU_64-d.10134/sql/src/backends/monet5/sql_gencode.mx:245:
warning: `q' might be used uninitialized in this function
make[5]: *** [sql_gencode.lo] Error 1
## Comment 1773
Date: 2005-10-06 22:43:23 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done
BugDay_2005-10-06: NO TEST / COMPILATION
## Comment 1774
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023527 at http://sourceforge.net/support/tracker.php?aid=1023527
| sql/src/backends/monet5/sql_gencode.mx fails to compile | https://api.github.com/repos/MonetDB/MonetDB/issues/328/comments | 0 | 2020-11-30T08:29:13Z | 2024-06-28T13:36:30Z | https://github.com/MonetDB/MonetDB/issues/328 | 753,269,108 | 328 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-07 08:56:21 +0200
From: @drstmane
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-08 07:32:26 +0200
## Comment 1768
Date: 2004-09-07 08:56:21 +0200
From: @drstmane
the output some tests (see list below AND check
TestWeb) is affected by the change that made the
"default seqbase value of non-void columns 0 (iso
nil)"; hence, the respective output needs to be approved.
tests/BugsViaSourgeforce/Tests/ID.823516
tests/from_Twente/Bugs/Tests/*
## Comment 1769
Date: 2004-09-08 19:32:26 +0200
From: @drstmane
Logged In: YES
user_id=572415
Peter confirmed that I should approve the respective output.
And I did as he said.
## Comment 1770
Date: 2005-10-06 22:42:27 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done / NO TEST
Because this bug is caused by out-of-date stable output of
some test scripts.
## Comment 1771
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023479 at http://sourceforge.net/support/tracker.php?aid=1023479
| approval of stable output | https://api.github.com/repos/MonetDB/MonetDB/issues/327/comments | 0 | 2020-11-30T08:29:10Z | 2024-06-27T10:58:32Z | https://github.com/MonetDB/MonetDB/issues/327 | 753,269,081 | 327 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-06 02:56:58 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-09-07 01:17:35 +0200
## Comment 1764
Date: 2004-09-06 14:56:58 +0200
From: @arjan
Really should have submitted this one first, but
anyhow. The last SELECT operation gives the error
message. Attached below code as well.
--
CREATE TABLE _metadata (
artist VARCHAR(255),
album VARCHAR(255),
track VARCHAR(255),
tracknum INT(10),
duration INT(10)
);
CREATE VIEW metadata AS
SELECT artist, album, track, tracknum, duration,
((CASE WHEN artist IS NULL OR
char_length(trim(artist))=0 THEN 0 ELSE 16 END) +
(CASE WHEN album IS NULL OR
char_length(trim(album))=0 THEN 0 ELSE 8 END) +
(CASE WHEN track IS NULL OR
char_length(trim(track))=0 THEN 0 ELSE 4 END) +
(CASE WHEN tracknum IS NULL OR tracknum=0 THEN 0
ELSE 2 END) +
(CASE WHEN duration IS NULL OR duration=0 THEN 0
ELSE 1 END)) AS weightindex
FROM _metadata;
SELECT * FROM metadata;
-
MAPI = monetdb@localhost:45123
QUERY = SELECT * FROM metadata;
ERROR = !ERROR: batbat_sht_add: BATs l & r must be
void-headed and head-aligned.
!ERROR: batbat_sht_add: operation failed.
## Comment 1765
Date: 2004-09-07 13:17:35 +0200
From: @njnes
Logged In: YES
user_id=43556
this bug was a side effect of the empty bat bug
## Comment 1766
Date: 2005-10-06 15:26:08 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
test script available as:
sql/src/test/bugs/Tests/view_case-bug-sf-1023058.*
sql/src/test/bugs/view_case-bug-sf-1023058.sql
## Comment 1767
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023058 at http://sourceforge.net/support/tracker.php?aid=1023058
This bug was marked DUPLICATE in the database it was moved from.
Changing resolution to "MOVED"
| SQL: Case bug (1) | https://api.github.com/repos/MonetDB/MonetDB/issues/326/comments | 0 | 2020-11-30T08:29:07Z | 2024-06-28T13:36:29Z | https://github.com/MonetDB/MonetDB/issues/326 | 753,269,040 | 326 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-06 02:54:02 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-09-07 01:54:17 +0200
## Comment 1760
Date: 2004-09-06 14:54:02 +0200
From: @arjan
Trying to find a bug i stumble upon another:
select ((case when name='tables' or name<>'views' then
1 else 0 end)) from tables;
-
MAPI = monetdb@localhost:45123
QUERY = select ((case when name='tables' or
name<>'views' then 1 else 0 end)) from tables;
ERROR = !ERROR: interpret: no matching MIL operator to
'kunion(BAT[oid,void], BAT[void,str])'.
!MAYBE YOU MEAN:
! kunion(BAT[any::1,any::2],
BAT[any::1,any::2]) : BAT[any::1,any::2]
## Comment 1761
Date: 2004-09-07 13:54:17 +0200
From: @njnes
Logged In: YES
user_id=43556
The optimization of OR's could lead to union of select and
basetable,
ie oid,'void' and void,'something'. Now we simply create
void,0 in this case.
## Comment 1762
Date: 2005-10-06 15:25:29 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
test script available as:
sql/src/test/bugs/Tests/case_parenthesis-bug-sf-1023056.*
sql/src/test/bugs/case_parenthesis-bug-sf-1023056.sql
## Comment 1763
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1023056 at http://sourceforge.net/support/tracker.php?aid=1023056
| SQL: CASE bug (2) | https://api.github.com/repos/MonetDB/MonetDB/issues/325/comments | 0 | 2020-11-30T08:29:05Z | 2024-06-27T10:58:29Z | https://github.com/MonetDB/MonetDB/issues/325 | 753,269,001 | 325 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-06 10:58:12 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-12 01:55:34 +0200
## Comment 1756
Date: 2004-09-06 10:58:12 +0200
From: @grobian
the auto-commit test (Test_Cautocommit.java) fails on
Darwin:
Unknown table table_test_cautocommit
Monet returns to have auto-commit enabled when it
performs these two queries on two separate statements
(on the same connection):
CREATE TABLE table_Test_Cautocommit ( id int );
SELECT * FROM table_Test_Cautocommit;
The darwin machine fails on the select...
## Comment 1757
Date: 2004-09-12 13:55:34 +0200
From: @njnes
Logged In: YES
user_id=43556
Seems the recent changes in transaction support have fixed
this bug
## Comment 1758
Date: 2005-10-06 22:36:14 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done / ALREADY IN TESTWEB
## Comment 1759
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1022952 at http://sourceforge.net/support/tracker.php?aid=1022952
| Testweb: scaryness with Darwin: unknown table | https://api.github.com/repos/MonetDB/MonetDB/issues/324/comments | 0 | 2020-11-30T08:29:01Z | 2024-06-27T10:58:28Z | https://github.com/MonetDB/MonetDB/issues/324 | 753,268,961 | 324 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-06 10:45:23 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-06 11:48:36 +0200
## Comment 1749
Date: 2004-09-06 10:45:23 +0200
From: @grobian
The JDBC tests which involve timestamps now report
different output (a 2 hour difference). This is
probably due to the recent time zone preparations.
Either JDBC does not handle correctly, or the server
has changed its output/behvaviour on input.
It used to work... Performed test is roughly an insert
of a timestamp via INSERT INTO and then a reparse of
that date on a SELECT.
(Re)compiling monet to dive further into details.
## Comment 1750
Date: 2004-09-06 11:05:54 +0200
From: @grobian
Logged In: YES
user_id=963970
monetdb-> create table x (id timestamp);
Operation successful
monetdb-> insert into x values ('2004-10-06 16:00:00');
1 affected row
monetdb-> select * from x;
+----------+
| id |
+==========+
| 2004-10-06 18:00:00.000 |
+----------+
1 row
monetdb->
If above would be true, I would be very unhappy.
I think there needs some additional tweaking as I don't give
any timezone information, thus no timezone needs to be
applied to the given timestamp.
## Comment 1751
Date: 2004-09-06 11:37:58 +0200
From: @drstmane
Logged In: YES
user_id=572415
apparently, without explicite specific timezone
specification, MonetDB/SQL resp. monettime assumes UTC on
cast from string, but then seems to assume local timezone
(CEST == UTC + 2h) on output/cast to string:
16:00:00 UTC = 18:00:00 CEST ... ???
## Comment 1752
Date: 2004-09-06 11:42:28 +0200
From: @grobian
Logged In: YES
user_id=963970
well, if I recall correct the darwin machine (which is in
another time zone) has the same 2 hour difference on the
tests, which makes me think something is not going
completely well with guessing the local time zone.
## Comment 1753
Date: 2004-09-06 11:48:36 +0200
From: @njnes
Logged In: YES
user_id=43556
fixed bug in monettime
## Comment 1754
Date: 2005-10-06 22:35:45 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done / ALREADY IN TESTWEB
## Comment 1755
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1022945 at http://sourceforge.net/support/tracker.php?aid=1022945
| Testweb: regression on time tests | https://api.github.com/repos/MonetDB/MonetDB/issues/323/comments | 0 | 2020-11-30T08:28:58Z | 2024-06-27T10:58:27Z | https://github.com/MonetDB/MonetDB/issues/323 | 753,268,926 | 323 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-04 10:54:41 +0200
From: @sjoerdmullender
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-09-07 08:30:48 +0200
## Comment 1745
Date: 2004-09-04 10:54:41 +0200
From: @sjoerdmullender
A bunch of tests fail since a bunch of new string
functions have been added. The output needs to be
checked and approved.
## Comment 1746
Date: 2004-09-07 08:30:48 +0200
From: @drstmane
Logged In: YES
user_id=572415
fixed. couldn't resist ...
## Comment 1747
Date: 2005-10-06 22:34:52 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: NO TEST (maybe already in testweb)
Because this bug is about a collection of failures in
testweb. I believe that necessary tests have already been
added at the time this bug was fixed.
## Comment 1748
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1022262 at http://sourceforge.net/support/tracker.php?aid=1022262
| new string functions should be approved | https://api.github.com/repos/MonetDB/MonetDB/issues/322/comments | 0 | 2020-11-30T08:28:56Z | 2024-06-27T10:58:26Z | https://github.com/MonetDB/MonetDB/issues/322 | 753,268,910 | 322 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-09-02 09:01:46 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-09-12 01:57:14 +0200
## Comment 1736
Date: 2004-09-02 21:01:46 +0200
From: @grobian
The Sun test machine reports the following error when
trying to run any of the JDBC tests:
Exception in thread "main"
java.lang.NoClassDefFoundError: Test_Cautocommit
This indicates Java is available/configured properly,
but the classpath is somehow set wrong.
I think this has to do some different paths setting on
the Solaris machine somehow, but I can't check it since
I don't know how to reproduce it (and I don't got
Solaris installed anymore).
## Comment 1737
Date: 2004-09-02 21:12:06 +0200
From: @njnes
Logged In: YES
user_id=43556
could this be related to the old java version on the sun?
## Comment 1738
Date: 2004-09-02 21:22:38 +0200
From: @grobian
Logged In: YES
user_id=963970
I just noticed that compilation of those tests fail, with a
small orange x. I don't know what that means, for it's not
a large red X.
I assume it means that the compilation has not even been
attempted.
If the Java version is < 1.4 then compilation of the driver
itself won't succeed. (which is the case)
(apps:~) fabian% java -version
java version "1.3.0_02"
Java(TM) 2 Runtime Environment, Standard Edition (build
1.3.0_02)
Java HotSpot(TM) Client VM (build 1.3.0_02, mixed mode)
Is it possible to even think of an 'upgrade' ?
## Comment 1739
Date: 2004-09-02 21:37:49 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
The small orange x is a new feature (today was the first
time). It appears for missing files in directories that are
optionally compiled because, since it is optional, it can be
expected that for some systems the files will be missing.
## Comment 1740
Date: 2004-09-02 22:29:37 +0200
From: @drstmane
Logged In: YES
user_id=572415
since recently, MonetDB and MonetDB/SQL require java 1.4 but
on our solaris machine (apps) weonly have java 1.3, hence,
non of the java stuff will get compiled --- and thanks to
sjoerd, small orange 'x's now indicate, that we do not
expect them to compile (large red 'X's still indicate files
that "unexpectedly" do not compile.
## Comment 1741
Date: 2004-09-03 19:32:13 +0200
From: @grobian
Logged In: YES
user_id=963970
Fix (upgrade of Java installation) is out of controllable
reach. Need to keep an eye on it, or alternatively (if
possible) use userland Java installation.
## Comment 1742
Date: 2004-09-12 13:57:14 +0200
From: @njnes
Logged In: YES
user_id=43556
we need a java >= 1.4, which we installed on the solaris system
## Comment 1743
Date: 2005-10-06 22:30:52 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done / ALREADY IN TESTWEB
## Comment 1744
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1021377 at http://sourceforge.net/support/tracker.php?aid=1021377
| TestWeb: Solaris JDBC tests | https://api.github.com/repos/MonetDB/MonetDB/issues/321/comments | 0 | 2020-11-30T08:28:54Z | 2024-06-27T10:58:25Z | https://github.com/MonetDB/MonetDB/issues/321 | 753,268,888 | 321 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-31 11:48:03 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-08-31 02:11:23 +0200
## Comment 1732
Date: 2004-08-31 11:48:03 +0200
From: @arjan
select name from tables where name='doesnotexist';
gives no result, which is correct.
However:
select 1 from tables where name='doesnotexist';
gives:
[ 1 ]
which is not correct. A related issue: 'select 1 from
tables' gives only a single row...
## Comment 1733
Date: 2004-08-31 14:11:23 +0200
From: @njnes
Logged In: YES
user_id=43556
Improved the check if a const to column conversion is needed.
## Comment 1734
Date: 2005-10-06 15:24:30 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
test script available as:
sql/src/test/bugs/Tests/select_constant_from_where_1=1-bug-sf-1019529.*
sql/src/test/bugs/select_constant_from_where_1=1-bug-sf-1019529.sql
## Comment 1735
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1019529 at http://sourceforge.net/support/tracker.php?aid=1019529
| SQL: Selecting only constants from table | https://api.github.com/repos/MonetDB/MonetDB/issues/320/comments | 0 | 2020-11-30T08:28:51Z | 2024-06-27T10:58:24Z | https://github.com/MonetDB/MonetDB/issues/320 | 753,268,859 | 320 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-31 11:03:40 +0200
From: @sjoerdmullender
To: @njnes
Version: -- development
Last updated: 2004-08-31 02:10:15 +0200
## Comment 1728
Date: 2004-08-31 11:03:40 +0200
From: @sjoerdmullender
This little gem causes an assertion to fail in
sql_optimize.mx (line 870). It was concocted by the
ODBC SQLStatisctics function.
select cast(null as varchar) as table_cat, cast(s.name
as varchar) as table_schem, cast(t.name as varchar) as
table_name, case when k.name is null then cast(1 as
smallint) else cast(0 as smallint) end as non_unique,
cast(null as varchar) as index_qualifier, cast(i.name
as varchar) as index_name, case i.type when 0 then
cast(2 as smallint) else cast(3 as smallint) end as
type, cast(kc.nr as smallint) as ordinal_position,
cast(c.name as varchar) as column_name, cast(null as
char(1)) as asc_or_desc, cast(null as integer) as
cardinality, cast(null as integer) as pages, cast(null
as varchar) as filter_condition from sys.idxs i,
sys.schemas s, sys.tables t, sys.columns c,
sys.keycolumns kc, sys.keys k where i.table_id = t.id
and t.schema_id = s.id and i.id = kc.id and t.id =
c.table_id and kc."column" = c.name and (k.type is null
or k.type = 1) and t.name = 'c1006309' and s.name =
'sys' order by non_unique, type, index_qualifier,
index_name, ordinal_position;
## Comment 1729
Date: 2004-08-31 14:10:15 +0200
From: @njnes
Logged In: YES
user_id=43556
The keys table is not joined with the other tables, ie an
implicite cross product is needed. This case wasn't recognized.
## Comment 1730
Date: 2005-10-06 15:23:18 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
test script available as:
sql/src/test/bugs/Tests/case_cast-bug-sf-1019506.*
sql/src/test/bugs/case_cast-bug-sf-1019506.sql
## Comment 1731
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1019506 at http://sourceforge.net/support/tracker.php?aid=1019506
| assertion failure | https://api.github.com/repos/MonetDB/MonetDB/issues/319/comments | 0 | 2020-11-30T08:28:49Z | 2024-06-27T10:58:23Z | https://github.com/MonetDB/MonetDB/issues/319 | 753,268,828 | 319 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-30 04:51:50 +0200
From: @sjoerdmullender
To: @njnes
Version: -- development
Last updated: 2004-08-31 10:28:00 +0200
## Comment 1724
Date: 2004-08-30 16:51:50 +0200
From: @sjoerdmullender
The following command gives an error. From the grammar
in sql_parser.mx I expect this to be correct syntax.
create table test (m interval month);
The error is !Create Column: type or name
## Comment 1725
Date: 2004-08-31 10:28:00 +0200
From: @njnes
Logged In: YES
user_id=43556
added mising interval qualifier parse code
## Comment 1726
Date: 2005-10-06 22:26:28 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
BugDay_2005-10-06: TEST ADDED / SUCCESS
src/test/BugDay_2005-10-06_2.9.3/Tests/error_in_sql_parser.SF-1019122.*
## Comment 1727
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1019122 at http://sourceforge.net/support/tracker.php?aid=1019122
| error in SQL parser | https://api.github.com/repos/MonetDB/MonetDB/issues/318/comments | 0 | 2020-11-30T08:28:46Z | 2024-06-27T10:58:22Z | https://github.com/MonetDB/MonetDB/issues/318 | 753,268,784 | 318 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-30 03:31:18 +0200
From: @sjoerdmullender
To: @njnes
Version: -- development
Last updated: 2004-08-31 10:28:38 +0200
## Comment 1720
Date: 2004-08-30 15:31:18 +0200
From: @sjoerdmullender
The SQL command
select cast('0' as interval year);
causes a crash in the server.
## Comment 1721
Date: 2004-08-31 10:28:38 +0200
From: @njnes
Logged In: YES
user_id=43556
added missing interval qualifier parse code
## Comment 1722
Date: 2005-10-06 21:40:26 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/cast_to_interval.SF-1019079.*
## Comment 1723
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1019079 at http://sourceforge.net/support/tracker.php?aid=1019079
| cast to interval causes crash | https://api.github.com/repos/MonetDB/MonetDB/issues/317/comments | 0 | 2020-11-30T08:28:43Z | 2024-06-27T10:58:21Z | https://github.com/MonetDB/MonetDB/issues/317 | 753,268,751 | 317 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-26 03:16:05 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-08-31 10:29:47 +0200
## Comment 1716
Date: 2004-08-26 15:16:05 +0200
From: @arjan
create table t1 (id int, primary key(id));
create table t2 (id int, t1 int, primary key(id),
foreign key (t1) references t1);
insert into t1 values (1);
insert into t2 values (1,1);
second insert crashes server with:
Mserver: sql_mem.mx:87: sql_ref_inc: Assertion
`r->refcnt > 0' failed.
## Comment 1717
Date: 2004-08-31 10:29:47 +0200
From: @njnes
Logged In: YES
user_id=43556
added stmt_dup (stmt was used twice with only one refcnt)
## Comment 1718
Date: 2005-10-06 21:32:25 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/fk_without_notnull_constraint.SF-1016842.*
## Comment 1719
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1016842 at http://sourceforge.net/support/tracker.php?aid=1016842
| SQL: foreign keys without not-null constraint -> meltdown | https://api.github.com/repos/MonetDB/MonetDB/issues/316/comments | 0 | 2020-11-30T08:28:41Z | 2024-06-27T10:58:20Z | https://github.com/MonetDB/MonetDB/issues/316 | 753,268,725 | 316 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-24 04:02:40 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2006-12-04 09:02:48 +0100
## Comment 1709
Date: 2004-08-24 16:02:40 +0200
From: @arjan
1) Start Mserver with sql_server module
2) Open MapiClient -lsql and create a table, for instance
CREATE TABLE test (x int); COMMIT;
3) Quit client
4) Quit server
5) Start server with sql_server module
6) Start MapiClient -lsql
7) Server crashes! Message:
monet>Mserver: sql_mvc.mx:3013: mvc_bind: Assertion `t'
failed.
Aborted
Then, the only way to reconnect to the server is to
remove the dbfarm...
## Comment 1710
Date: 2005-11-06 11:04:43 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 1711
Date: 2005-11-21 04:20:21 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
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 1712
Date: 2006-11-22 04:20:02 +0100
From: Pseudo user for Sourceforge import <<sfimport>>
---- Original comment by: sf-robot@users.sourceforge.net ----
Logged In: YES
user_id=1312539
Originator: NO
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 365 days (the time period specified by
the administrator of this Tracker).
## Comment 1713
Date: 2006-11-29 10:05:01 +0100
From: @drstmane
Logged In: YES
user_id=572415
Originator: NO
over-ruling "sf-robot":
re-set to "Pending" since adding a proper test has neither been done nor finally been discarded, yet.
User (submitter) & developer (assignee),
please check, again.
## Comment 1714
Date: 2006-12-04 21:02:48 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
Tested by allmost any sql test. So many tests exists. Closing..
## Comment 1715
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1015273 at http://sourceforge.net/support/tracker.php?aid=1015273
| Server crash after restart | https://api.github.com/repos/MonetDB/MonetDB/issues/315/comments | 0 | 2020-11-30T08:28:38Z | 2024-06-27T10:58:19Z | https://github.com/MonetDB/MonetDB/issues/315 | 753,268,691 | 315 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-24 01:39:44 +0200
From: Jens Teubner <<teubner>>
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-09-06 06:40:38 +0200
## Comment 1703
Date: 2004-08-24 13:39:44 +0200
From: Jens Teubner <<teubner>>
I started to implement arithmetics in Pathfinder. With MIL scripts becoming
a bit larger, MonetDB starts to segfault here.
I'm using 4.3.18 with Stefan's fixes for http://sourceforge.net/support/tracker.php?aid=996570 (end July) here. The
attached MIL script produces the segfault (I cut the remaining lines of my
script off.). Replacing `unused' by `nil' does not help. After throwing out all
lines containing `unused' (grep -v unused), however, the script runs nicely.
(Well, then the line that caused the error is gone also.)
Sometimes (but only sometimes), the Mserver console shows messages
like `free(): invalid pointer 0x85299e0!' and will fail lateron (not
immediately).
Let me know how I can help you track the problem down.
## Comment 1704
Date: 2004-08-25 18:14:11 +0200
From: Jens Teubner <<teubner>>
Logged In: YES
user_id=731390
I investigated a bit into the problem (as far as a
MonetDB novice can ;-)):
-- The problems are not related to the arithmetics
I started to implement. The behavior can be
reproduced with the current CVS version of
Pathfinder.
-- GDB tells me that the segmentation fault is
triggered from MT_create_thread(), it finally
happens in the pthread library.
-- You can easily reproduce the problem by using
the code generated by Pathfinder given the
query `for $i in (1,2) return $i':
echo 'for $i in (1,2) return $i' | pf | MapiClient
-- If you remove the `unused' assignments that I
generate to ``free'' variables, the script runs
nicely and prints the correct query result. In that
case, however, the Mserver will crash as soon
as the MapiClient terminates. (GDB reports the
same location.)
echo 'for $i in (1,2) return $i' | pf | grep -v unused | MapiClient
The attached file is the MIL output of the above
XQuery expression with the `unused' lines removed.
The problem is triggered as soon as the script is
executed up to line 117.
## Comment 1705
Date: 2004-09-06 18:40:38 +0200
From: @peterboncz
Logged In: YES
user_id=591107
I added the following lines to the scripts:
module(xtables);
module(pf_support);
module(pathfinder);
to stop them causing syntax errors.
then they run fine - no crash in sight.
Maybe this bug was fixed after I fixed Jan's crashes three
weeks ago?
## Comment 1706
Date: 2004-09-06 18:50:15 +0200
From: @drstmane
Logged In: YES
user_id=572415
[repeating the info that I only gave as checkin-message when
adding Jens' scripts to the test suite in CVS --- sorry for
not adding this info here in time!]
this bug still exists in MonetDB-4.3.18 ("Stable"), but
seems to be fixed in MonetDB-4.3.19 ("Current") --- most
probably by some recent fixed made by Peter.
since MonetDB-4.3.18 will become obsolete as soon as
MonetDB-4.4_rc00 will be released later this week, the bug
will not be fixed in MonetDB-4.3.18.
## Comment 1707
Date: 2005-10-06 15:20:56 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
test script available as:
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1015172a.*
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1015172b.*
pathfinder/tests/BugsViaSourgeforce/Tests/ID.1015172c.*
## Comment 1708
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1015172 at http://sourceforge.net/support/tracker.php?aid=1015172
| Mserver segfaults (``internal error'') | https://api.github.com/repos/MonetDB/MonetDB/issues/314/comments | 0 | 2020-11-30T08:28:34Z | 2024-06-28T13:36:28Z | https://github.com/MonetDB/MonetDB/issues/314 | 753,268,658 | 314 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-21 02:26:37 +0200
From: @grobian
To: Arjan Scherpenisse <<arjan>>
Version: -- development
Last updated: 2004-08-23 05:09:50 +0200
## Comment 1696
Date: 2004-08-21 14:26:37 +0200
From: @grobian
MonetDB doesn't compile anymore on machines with
up-to-date software.
SWIG seems to throw in a lot of dust.
SWIG on CWI RedHat 9:
(pavo:~) fabian% swig -version
SWIG Version 1.1 (Patch 5)
Copyright (c) 1995-98
University of Utah and the Regents of the University of
California
Compiled with c++
SWIG on CWI Fedora Core 2
(pictor:~) fabian% swig -version
SWIG Version 1.3.19
Copyright (c) 1995-1998
University of Utah and the Regents of the University of
California
Copyright (c) 1998-2002
University of Chicago
Compiled with i386-redhat-linux-g++
Please see http://www.swig.org for reporting bugs and
further information
SWIG on an Gentoo up-to-date system:
$ swig -version
SWIG Version 1.3.21
Copyright (c) 1995-1998
University of Utah and the Regents of the University of
California
Copyright (c) 1998-2003
University of Chicago
Compiled with g++ [i686-pc-linux-gnu]
Please see http://www.swig.org for reporting bugs and
further information
With SWIG 1.3.19 the compilation seems to go, but with
the heavy production of zillions of warnings. SWIG
1.3.21 on the other hand errors on generated code.
MapiLib.pm.c:2786 error: `na' undeclared (first use in
this function).
It's annoying that it doesn't compile out of the box
anymore.
## Comment 1697
Date: 2004-08-23 10:13:26 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Swig below 1.3.21 is not supported. Any reports for lower
versions will be ignored. The MonetDB configure program
checks the version number of swig, and if it is below 1.3.20
it will say swig is not available. (If the check fails,
that's a bug.)
It is a known problem that swig doesn't produce clean code
(third party bug). Therefore the Makefile.ag in both the
Python and Perl Cimpl directories overrides the X_CFLAGS
variable to turn off the flag to abort on warnings.
When I compile the Perl Cimpl directory I get lots of
(expected) warnings, but no errors. Please provide more
details about your environment (version numbers of relevant
software such as Perl).
From the error message I gather that PL_na is not defined
and so defined to na in MapiLib.pm.c (i.e. MapiLib.pm.i).
Assigning to Arjan since this seems to have to do with Perl.
## Comment 1698
Date: 2004-08-23 10:21:22 +0200
From: @grobian
Logged In: YES
user_id=963970
Autoconf: sys-devel/autoconf-2.59-r4
Automake: sys-devel/automake-1.8.3
SWIG: Version 1.3.21
Perl: v5.8.4 built for i586-linux
gcc: (GCC) 3.3.3 20040412
Python: 2.3.3 (1, May 17 2004, 20:53:44)
what else do you want more?
## Comment 1699
Date: 2004-08-23 17:09:50 +0200
From: @arjan
Logged In: YES
user_id=20087
Removed useless define (caused error on gentoo systems),
muted some of swigs warnings by replacing deprecated
constructs in typemaps
## Comment 1700
Date: 2004-08-23 19:58:32 +0200
From: @grobian
Logged In: YES
user_id=963970
delicious!
## Comment 1701
Date: 2005-10-06 20:08:55 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
BugDay_2005-10-06: NO TEST / COMPILATION
## Comment 1702
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1013373 at http://sourceforge.net/support/tracker.php?aid=1013373
| SWIG annoyances | https://api.github.com/repos/MonetDB/MonetDB/issues/313/comments | 0 | 2020-11-30T08:28:32Z | 2024-06-27T10:58:17Z | https://github.com/MonetDB/MonetDB/issues/313 | 753,268,630 | 313 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-20 01:03:40 +0200
From: @njnes
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-09-07 09:09:07 +0200
## Comment 1691
Date: 2004-08-20 13:03:40 +0200
From: @njnes
Queries 16/17 of the tpch benchmark/test fail because of
the resently changed handling of oid-empty bats.
Err output of q16
MAPI = monetdb@medusa:53050
QUERY = select
ERROR = !ERROR: batbat_int_sub: BATs l & r must be
void-headed and head-aligned.
!ERROR: batbat_int_sub: operation failed.
!ERROR: [string](param 3): evaluation error.
!WARNING: BBPdecref: range error 199
!WARNING: BBPdecref: range error 199
## Comment 1692
Date: 2004-08-20 15:00:39 +0200
From: @njnes
Logged In: YES
user_id=43556
also tests no.107 and all from tests/from_Twente/Bugs seem
affected by this change
## Comment 1693
Date: 2004-09-07 09:09:07 +0200
From: @drstmane
Logged In: YES
user_id=572415
Peter rolled-back this changes; hence the bug is fixed.
## Comment 1694
Date: 2005-10-06 21:09:53 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
Tests already exist in test-web
## Comment 1695
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1012839 at http://sourceforge.net/support/tracker.php?aid=1012839
| void/oid empty bats handling | https://api.github.com/repos/MonetDB/MonetDB/issues/312/comments | 0 | 2020-11-30T08:28:29Z | 2024-06-27T10:58:16Z | https://github.com/MonetDB/MonetDB/issues/312 | 753,268,605 | 312 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-08-20 09:36:24 +0200
From: @grobian
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-09-07 09:07:47 +0200
## Comment 1682
Date: 2004-08-20 09:36:24 +0200
From: @grobian
> (pictor:~/develop/monetjdbc) fabian% java -jar
jars/MonetJDBC.jar
> Welcome to the MonetDB interactive JDBC terminal!
> Database: MonetDB 4.3.19 (4.3)
> Driver: MonetDB Native Driver 0.8 1.0 Release
Candidate 1 (0.8)
> Type \q to quit, \h for a list of available commands
> auto commit mode: on
> monetdb-> \d
> TABLE sys.craftsmen
> TABLE sys.impotenten
> TABLE sys.invoices
> TABLE sys.passengers
> TABLE sys.seafarers
> TABLE sys.soldiers
> TABLE sys.total
> TABLE sys.voyages
> monetdb-> \d voyages;
> A fatal exception occurred: java.sql.SQLException:
ERROR: BATpropcheck:
> BAT -538 is dense but not key!?
> ERROR: CMDsort: returned BAT has unexpected head.
> ERROR: reverse(param 1): evaluation error.
>
executed SQL query + server response:
> << sSELECT null AS "TABLE_CAT", "schemas"."name" AS
"TABLE_SCHEM",
> "tables"."nam
> e" AS "TABLE_NAME", "keycolumns"."column" AS
"COLUMN_NAME",
> "keys"."type" AS "KE
> Y_SEQ", "keys"."name" AS "PK_NAME" FROM "keys",
"keycolumns", "tables",
> "schemas
> " WHERE "keys"."id" = "keycolumns"."id" AND
"keys"."table_id" =
> "tables"."id" AN
> D "tables"."schema_id" = "schemas"."id" AND
"keys"."type" = 0 AND
> "schemas"."nam
> e" LIKE 'sys' AND "tables"."name" LIKE 'voyages'
ORDER BY "COLUMN_NAME";
> >> -
> >> !ERROR: BATpropcheck: BAT -857 is dense but not key!?
> >> !ERROR: CMDsort: returned BAT has unexpected head.
> >> !ERROR: reverse(param 1): evaluation error.
> >> ^A^A
## Comment 1683
Date: 2004-08-20 14:56:21 +0200
From: @njnes
Logged In: YES
user_id=43556
also the tests moa/list and by tests/BugsViaSourgeforce/
{ID.823516,ID.978814,ID.981954} have the propcheck error
## Comment 1684
Date: 2004-08-20 16:16:58 +0200
From: @grobian
Logged In: YES
user_id=963970
this error also appeared on the savepoints test:
9. rollback...passed :)
FAILED :( Unknown table table_test_csavepoints
ERROR: BATpropcheck: BAT -365 is dense but not key!?
## Comment 1685
Date: 2004-08-30 14:31:35 +0200
From: @grobian
Logged In: YES
user_id=963970
The latest error message in the savepoints test has an
additional error:
FAILED :( Unknown table table_test_csavepoints
ERROR: BATpropcheck: BAT -351 is dense but not key!?
ERROR: BATins: access denied to tmp_405, aborting.
## Comment 1686
Date: 2004-09-03 19:29:32 +0200
From: @grobian
Logged In: YES
user_id=963970
JdbcClient fails to describe a table due to the same error:
!ERROR: BATpropcheck: BAT -226 is dense but not key!?
!ERROR: CMDsort: returned BAT has unexpected head.
!ERROR: reverse(param 1): evaluation error.
I think it's time to increase the importance of this bug,
since it influences many things.
## Comment 1687
Date: 2004-09-06 18:41:30 +0200
From: @drstmane
Logged In: YES
user_id=572415
bug is most probably fixed by my fix of Peter's changes in
[left]fetchjoin:
empty results should not only be marked sorted & dense, but
also key
(dense implies sorted and key; hence, setting dense without
key was inconsistent, i.e. wrong)
let's check was testing tells us tomorrow, before we close
this bug report...
## Comment 1688
Date: 2004-09-07 09:07:47 +0200
From: @drstmane
Logged In: YES
user_id=572415
testing seems to confirm that the bug is indeed fixed.
## Comment 1689
Date: 2005-10-06 21:04:39 +0200
From: @yzchang
Logged In: YES
user_id=341633
BugDay_2005-10-06: Claimed by jennie
BugDay_2005-10-06: Done.
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/BATpropcheck_error.SF-1012739.*
## Comment 1690
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 1012739 at http://sourceforge.net/support/tracker.php?aid=1012739
| BATpropcheck error | https://api.github.com/repos/MonetDB/MonetDB/issues/311/comments | 0 | 2020-11-30T08:28:26Z | 2024-06-27T10:58:15Z | https://github.com/MonetDB/MonetDB/issues/311 | 753,268,570 | 311 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.