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: 2004-04-28 01:59:45 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-05-01 03:04:28 +0200
## Comment 1145
Date: 2004-04-28 13:59:45 +0200
From: @arjan
sql>select * from test a inner join test b on (a.i = b.i);
i, i name
mediumint, mediumint type
3, tuplecount
3, id
[ 1, 1 ]
[ 2, 2 ]
[ 3, 3 ]
sql>select * from test a inner join test b on (a.i =
b.i) inner join test c on (a.id = c.id);
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select * from test a inner join test b on (a.i
= b.i) inner join test c
on (a.id = c.id);
ERROR = !Column: a.id unknown
## Comment 1146
Date: 2004-05-01 15:04:28 +0200
From: @njnes
Logged In: YES
user_id=43556
Properly store the used table alias with each basetable.
The table_name function returns this value for
st_basetable
## Comment 1147
Date: 2005-10-06 17:32:31 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
BugDay_2005-10-06: TEST ADDED / SUCCESS
SQL script existed
(sql/src/test/bugs/innerjoin_multiple-bug-sf-943661.sql)
but was not yet called from the Tests subdir
## Comment 1148
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 943661 at http://sourceforge.net/support/tracker.php?aid=943661
| SQL: column references in JOIN | https://api.github.com/repos/MonetDB/MonetDB/issues/210/comments | 0 | 2020-11-30T08:18:15Z | 2024-06-27T10:56:36Z | https://github.com/MonetDB/MonetDB/issues/210 | 753,261,689 | 210 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-28 01:37:51 +0200
From: @menzowindhouwer
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-04-28 11:31:28 +0200
## Comment 1141
Date: 2004-04-28 13:37:51 +0200
From: @menzowindhouwer
The HTML page:
monetdb.cwi.nl/TechDocs/Services/gdk/index.html
doesn't contain the bibliography, i.e. references like [1]
can't be looked up. The latex in MonetDB/src/gdk.mx
wants to include gdk.bib, but it isn't there anymore ...
## Comment 1142
Date: 2004-04-28 23:31:28 +0200
From: @drstmane
Logged In: YES
user_id=572415
fixed:
(re-?) added missing (lost?) gdk.bib
and fixed mkdoc.py to properly handle Mx + .bib
## Comment 1143
Date: 2005-10-06 17:17:44 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
BugDay_2005-10-06: NO TEST / DOCUMENTATION
## Comment 1144
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 943655 at http://sourceforge.net/support/tracker.php?aid=943655
| kernel documentation misses bibliography | https://api.github.com/repos/MonetDB/MonetDB/issues/209/comments | 0 | 2020-11-30T08:18:13Z | 2024-06-27T10:56:35Z | https://github.com/MonetDB/MonetDB/issues/209 | 753,261,662 | 209 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-28 10:29:05 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-05-01 03:04:53 +0200
## Comment 1137
Date: 2004-04-28 10:29:05 +0200
From: @arjan
I think the query below should give a result instead of
the error message:
sql>select a,b, char_length(b), count(*) from test
group by a,b having count(a)= char_length(b);
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select a,b, char_length(b), count(*) from test
group by a,b having count(a) = char_length(b);
ERROR = !Cannot compare sets with values, probably a
aggregate function missing
sql>rollback;
Whereas using a constant in the char_length function
yields correct behaviour:
sql>select a,b, char_length(b), count(*) from test
group by a,b having count(a)
= char_length('a');
a, b, char_length_b, count_b name
mediumint, varchar, mediumint,
mediumint type
8, , , tuplecount
35, , , id
[ 1, "aloha", 5, 1 ]
[ 1, "aloha!!!!!", 10, 1 ]
[ 2, "aloha!!!!!", 10, 1 ]
[ 2, "hoi", 3, 1 ]
[ 2, "x", 1, 1 ]
## Comment 1138
Date: 2004-05-01 15:04:53 +0200
From: @njnes
Logged In: YES
user_id=43556
Now we join variables (groupby columns only) with
group result,
which are then allowed (ie are keyed) in the HAVING
clause without
a aggregation
## Comment 1139
Date: 2005-10-06 17:15:23 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
BugDay_2005-10-06: TEST ADDED / SUCCESS
added link to
sql/src/test/bugs/groupby_having_charlength-bug-sf-943566.sql
into the corresponding Tests directory, so that the bug is
tested by testweb
## Comment 1140
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 943566 at http://sourceforge.net/support/tracker.php?aid=943566
| SQL: variable arguments in functions in HAVING clause | https://api.github.com/repos/MonetDB/MonetDB/issues/208/comments | 0 | 2020-11-30T08:18:10Z | 2024-06-27T10:56:34Z | https://github.com/MonetDB/MonetDB/issues/208 | 753,261,623 | 208 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-27 02:40:28 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-28 10:46:42 +0200
## Comment 1134
Date: 2004-04-27 14:40:28 +0200
From: @arjan
SELECT SUBSTRING('hallo' FROM i FOR 3) FROM N WHERE i
<= CHAR_LENGTH('hallo') + 2;
Yields a parse error...
The N table consists of a single column with values for
i in range (1 ... 1000)
## Comment 1135
Date: 2005-10-06 16:59:30 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/variable_arg_substring.SF_942981.*
## Comment 1136
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 942981 at http://sourceforge.net/support/tracker.php?aid=942981
| SQL: cannot use variable argument in SUBSTRING | https://api.github.com/repos/MonetDB/MonetDB/issues/207/comments | 0 | 2020-11-30T08:18:07Z | 2024-06-27T10:56:33Z | https://github.com/MonetDB/MonetDB/issues/207 | 753,261,601 | 207 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-27 07:36:51 +0200
From: @drstmane
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-04-27 11:07:15 +0200
## Comment 1129
Date: 2004-04-27 07:36:51 +0200
From: @drstmane
Don Quichote speaking:
Yesterday's changes do break compilation of GDK and
hence everything else:
========
icc -no-gcc -DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk -I../..
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk
-I../common -DNDEBUG -we140 -c99 -Wall -w2 -mp1 -O3
-restrict -unroll -tpp6 -axKWNPB -DHWCOUNTERS
-DHW_Linux -DHW_i686 -we266 -Werror
-wd1418,1419,279,310,981,810,444,193,111,177,171,181,764,269,108,188,1357,102,70
-D_REENTRANT -c gdk_bbp.c -Kpic -DPIC
/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/gdk_bbp.mx(499):
error 165: too few arguments in function call
BBP = (BBPrec*) GDKvmalloc(size, &maxsize);
^
compilation aborted for gdk_bbp.c (code 2)
make[4]: *** [gdk_bbp.lo] Error 1
/bin/sh ../../libtool --mode=compile icc -no-gcc
-DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk -I../..
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk
-I../common -DNDEBUG -we140 -c99 -Wall -w2 -mp1 -O3
-restrict -unroll
-tpp6 -axKWNPB -DHWCOUNTERS -DHW_Linux
-DHW_i686 -we266 -Werror
-wd1418,1419,279,310,981,810,444,193,111,177,171,181,764,269,108,188,1357,102,70
-D_REENTRANT -c -o gdk_heap.lo `test -f 'gdk_heap.c'
|| echo
'/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/'`gdk_heap.c
icc -no-gcc -DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk -I../..
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk
-I../common -DNDEBUG -we140 -c99 -Wall -w2 -mp1 -O3
-restrict -unroll -tpp6 -axKWNPB -DHWCOUNTERS
-DHW_Linux -DHW_i686 -we266 -Werror
-wd1418,1419,279,310,981,810,444,193,111,177,171,181,764,269,108,188,1357,102,70
-D_REENTRANT -c gdk_heap.c -Kpic -DPIC
/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/gdk_heap.mx(118):
error 165: too few arguments in function call
h->base = (char *)
GDKmallocmax(h->size, &h->maxsize);
^
compilation aborted for gdk_heap.c (code 2)
make[4]: *** [gdk_heap.lo] Error 1
/bin/sh ../../libtool --mode=compile icc -no-gcc
-DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk -I../..
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk
-I../common -DNDEBUG -we140 -c99 -Wall -w2 -mp1 -O3
-restrict -unroll
-tpp6 -axKWNPB -DHWCOUNTERS -DHW_Linux
-DHW_i686 -we266 -Werror
-wd1418,1419,279,310,981,810,444,193,111,177,171,181,764,269,108,188,1357,102,70
-D_REENTRANT -c -o gdk_setop.lo `test -f 'gdk_setop.c'
|| echo
'/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/'`gdk_setop.c
icc -no-gcc -DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk -I../..
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk
-I../common -DNDEBUG -we140 -c99 -Wall -w2 -mp1 -O3
-restrict -unroll -tpp6 -axKWNPB -DHWCOUNTERS
-DHW_Linux -DHW_i686 -we266 -Werror
-wd1418,1419,279,310,981,810,444,193,111,177,171,181,764,269,108,188,1357,102,70
-D_REENTRANT -c gdk_setop.c -Kpic -DPIC
mv -f gdk_setop.o gdk_setop.lo
/bin/sh ../../libtool --mode=compile icc -no-gcc
-DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk -I../..
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk
-I../common -DNDEBUG -we140 -c99 -Wall -w2 -mp1 -O3
-restrict -unroll
-tpp6 -axKWNPB -DHWCOUNTERS -DHW_Linux
-DHW_i686 -we266 -Werror
-wd1418,1419,279,310,981,810,444,193,111,177,171,181,764,269,108,188,1357,102,70
-D_REENTRANT -c -o gdk_utils.lo `test -f 'gdk_utils.c'
|| echo
'/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/'`gdk_utils.c
icc -no-gcc -DHAVE_CONFIG_H -I.
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk -I../..
-I/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk
-I../common -DNDEBUG -we140 -c99 -Wall -w2 -mp1 -O3
-restrict -unroll -tpp6 -axKWNPB -DHWCOUNTERS
-DHW_Linux -DHW_i686 -we266 -Werror
-wd1418,1419,279,310,981,810,444,193,111,177,171,181,764,269,108,188,1357,102,70
-D_REENTRANT -c gdk_utils.c -Kpic -DPIC
/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/gdk_utils.mx(898):
error 147: declaration is incompatible with "void
*GDKmallocmax(size_t={unsigned int}, size_t={unsigned
int} *, int)" (declared at line 2158 of
"/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/gdk.mx")
void* GDKmallocmax(size_t size, size_t *maxsize) {
^
/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/gdk_utils.mx(912):
error 165: too few arguments in function call
s = (ssize_t*) GDKvmalloc(newsize,
&newmax);
^
/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/gdk_utils.mx(1090):
error 147: declaration is incompatible with "void
*GDKvmalloc(size_t={unsigned int}, size_t={unsigned
int} *, int)" (declared at line 2164 of
"/var/tmp/monet.ntv_32-d.10938/MonetDB/src/gdk/gdk.mx")
void* GDKvmalloc(size_t size, size_t *maxsize) {
^
compilation aborted for gdk_utils.c (code 2)
make[4]: *** [gdk_utils.lo] Error 1
========
Hence, there is no testweb and no generated
documentation on the website, today...
This applies to both the "stable"(???) release branch
(4.3.16) and the development branch (4.3.17).
Don Quichote has spoken...
## Comment 1130
Date: 2004-04-27 07:40:38 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Peter missed removal of the patched definitions in gdk.mx.
Moreover, I suggest to leave the changes I had made for
some time in 4.3.17, because I am close to finishing the
protection against BATnew failures
## Comment 1131
Date: 2004-04-27 11:07:15 +0200
From: @drstmane
Logged In: YES
user_id=572415
could not resist:
for those of you longing for a *compileable* "stable"
MonetDB branch:
finished Peter'\''s half-way roll-back of Martin'\''s recent
changes
related to BATnew/GDKmalloc.
Now, MonetDB should compile (and work) again ...
(Cetero Censeo: Someone should add ACID properties to CVS ...)
## Comment 1132
Date: 2005-10-06 16:00:57 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
BugDay_2005-10-06: NO TEST / COMPILATION
## Comment 1133
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 942797 at http://sourceforge.net/support/tracker.php?aid=942797
| latest check-ins break compilation of GDK ! |-( | https://api.github.com/repos/MonetDB/MonetDB/issues/206/comments | 0 | 2020-11-30T08:18:04Z | 2024-06-27T10:56:32Z | https://github.com/MonetDB/MonetDB/issues/206 | 753,261,559 | 206 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-25 05:43:43 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-26 07:26:40 +0200
## Comment 1123
Date: 2004-04-25 17:43:43 +0200
From: @grobian
select col1 from (
select 'a' as col1
union
select 'b' as col1
) as a;
results in:
Identifier: col1 unknown or ambigious
the same holds for variants like
select "col1" from (
select 'a' as "col1"
union
select 'b' as "col1"
) as a;
or
select col1 from (
select 'a' as col1
union
select 'b' as col1
) as a where col1 like 'a';
## Comment 1124
Date: 2004-04-25 17:45:57 +0200
From: @grobian
Logged In: YES
user_id=963970
ohw and by the way, postgres of course swallows it like
candy and spits the right answer back.
## Comment 1125
Date: 2004-04-26 19:26:40 +0200
From: @njnes
Logged In: YES
user_id=43556
added extra alias stmt's
## Comment 1126
Date: 2005-10-06 15:25:09 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
## Comment 1127
Date: 2005-10-06 15:56:41 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: TEST ADDED / SUCCESS
A sql example to trigger this bug already existed in
src/tests/bugs//subselect_union-bug-sf-941826.sql,
but it was not integrated into testweb. fixed this.
## Comment 1128
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 941826 at http://sourceforge.net/support/tracker.php?aid=941826
| SQL: unable to use column names in outer select | https://api.github.com/repos/MonetDB/MonetDB/issues/205/comments | 0 | 2020-11-30T08:18:00Z | 2024-06-27T10:56:31Z | https://github.com/MonetDB/MonetDB/issues/205 | 753,261,520 | 205 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-25 04:44:41 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-26 07:28:43 +0200
## Comment 1118
Date: 2004-04-25 16:44:41 +0200
From: @grobian
select * from (
select 'a' as col1
) as a;
produces an error:
ERROR: interpret: no matching MIL operator to
'mark(str, oid)'.
Postgres thinks it is delicious SQL.
## Comment 1119
Date: 2004-04-26 19:28:43 +0200
From: @njnes
Logged In: YES
user_id=43556
set aggr,key, nrcols correct for all stmts
## Comment 1120
Date: 2005-10-06 13:35:31 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
## Comment 1121
Date: 2005-10-06 15:24:07 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: TEST ADDED / SUCCESS
The sql source for this test
(src/tests/bugs/subselect_single_value-bug-sf-941791.sql)
was already in cvs, but it was not called from testweb and
output had to be
added
## Comment 1122
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 941791 at http://sourceforge.net/support/tracker.php?aid=941791
| SQL: Single subquery query inpossible? | https://api.github.com/repos/MonetDB/MonetDB/issues/204/comments | 0 | 2020-11-30T08:17:57Z | 2024-06-27T10:56:30Z | https://github.com/MonetDB/MonetDB/issues/204 | 753,261,486 | 204 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-25 04:40:22 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-26 07:27:45 +0200
## Comment 1112
Date: 2004-04-25 16:40:22 +0200
From: @grobian
Since the recent checkins a bug has snooped in, which
at fist sight broke my Niels-safety(tm) approach.
Error on line 13: Aiieeee!!! Inconsistant data! Monet
said there were 2 tuples, but it only returned 0
Executed query: SELECT * FROM ( SELECT null AS
"TABLE_CAT", "schemas"."name" AS "TABLE_SCHEM",
"tables"."name" AS "TABLE_NAME", 'SYSTEM TABLE' AS
"TABLE_TYPE", '' AS "REMARKS", null AS "TYPE_CAT", null
AS "TYPE_SCHEM", null AS "TYPE_NAME", 'id' AS
"SELF_REFERENCING_COL_NAME", 'SYSTEM' AS
"REF_GENERATION" FROM "tables", "schemas" WHERE
"tables"."schema_id" = "schemas"."id" AND
"tables"."type" = 1 UNION ALL SELECT null AS
"TABLE_CAT", "schemas"."name" AS "TABLE_SCHEM",
"tables"."name" AS "TABLE_NAME", 'TABLE' AS
"TABLE_TYPE", '' AS "REMARKS", null AS "TYPE_CAT", null
AS "TYPE_SCHEM", null AS "TYPE_NAME", 'id' AS
"SELF_REFERENCING_COL_NAME", 'SYSTEM' AS
"REF_GENERATION" FROM "tables", "schemas" WHERE
"tables"."schema_id" = "schemas"."id" AND
"tables"."type" = 0 ) AS "tables" WHERE 1 = 1 ;
Weirdest thing is
missing table name 4: temp
missing table name 4: temp
missing table name 4: temp
appears in my Mserver console and the query
SELECT * FROM (
SELECT null AS "TABLE_CAT", "schemas"."name" AS
"TABLE_SCHEM", "tables"."name" AS "TABLE_NAME",
'SYSTEM TABLE' AS "TABLE_TYPE", '' AS "REMARKS", null
AS "TYPE_CAT", null AS "TYPE_SCHEM",
null AS "TYPE_NAME", 'id' AS
"SELF_REFERENCING_COL_NAME", 'SYSTEM' AS "REF_GENERATION"
FROM "tables", "schemas" WHERE "tables"."schema_id" =
"schemas"."id" AND "tables"."type" = 1
UNION ALL
SELECT null AS "TABLE_CAT", "schemas"."name" AS
"TABLE_SCHEM", "tables"."name" AS "TABLE_NAME",
'TABLE' AS "TABLE_TYPE", '' AS "REMARKS", null AS
"TYPE_CAT", null AS "TYPE_SCHEM",
null AS "TYPE_NAME", 'id' AS
"SELF_REFERENCING_COL_NAME", 'SYSTEM' AS "REF_GENERATION"
FROM "tables", "schemas" WHERE "tables"."schema_id" =
"schemas"."id" AND "tables"."type" = 0
UNION ALL
SELECT null AS "TABLE_CAT", "schemas"."name" AS
"TABLE_SCHEM", "tables"."name" AS "TABLE_NAME",
'VIEW' AS "TABLE_TYPE", '' AS "REMARKS", null AS
"TYPE_CAT", null AS "TYPE_SCHEM",
null AS "TYPE_NAME", 'id' AS
"SELF_REFERENCING_COL_NAME", 'SYSTEM' AS "REF_GENERATION"
FROM "tables", "schemas" WHERE "tables"."schema_id" =
"schemas"."id" AND "tables"."type" = 2
UNION ALL
SELECT null AS "TABLE_CAT", "schemas"."name" AS
"TABLE_SCHEM", "tables"."name" AS "TABLE_NAME",
'SESSION TEMPORARY TABLE' AS "TABLE_TYPE", '' AS
"REMARKS", null AS "TYPE_CAT", null AS "TYPE_SCHEM",
null AS "TYPE_NAME", 'id' AS
"SELF_REFERENCING_COL_NAME", 'SYSTEM' AS "REF_GENERATION"
FROM "tables", "schemas" WHERE "tables"."schema_id" =
"schemas"."id" AND "tables"."type" = 3
UNION ALL
SELECT null AS "TABLE_CAT", "schemas"."name" AS
"TABLE_SCHEM", "tables"."name" AS "TABLE_NAME",
'TEMPORARY TABLE' AS "TABLE_TYPE", '' AS "REMARKS",
null AS "TYPE_CAT", null AS "TYPE_SCHEM",
null AS "TYPE_NAME", 'id' AS
"SELF_REFERENCING_COL_NAME", 'SYSTEM' AS "REF_GENERATION"
FROM "tables", "schemas" WHERE "tables"."schema_id" =
"schemas"."id" AND "tables"."type" = 4
) AS "tables" WHERE 1 = 1
ORDER BY "TABLE_TYPE", "TABLE_SCHEM", "TABLE_NAME";
complains about that it can't find the column TABLE_TYPE...
So something is really broken I think...
## Comment 1113
Date: 2004-04-25 17:37:30 +0200
From: @grobian
Logged In: YES
user_id=963970
Due to the recent mergejoin problems I reverted back to a
pre-fix monet state with
cvs update -D 2004-04-22
in the MonetDB tree.
This didn't change anything to the situation. I went back in
the sql tree before niels' fixes of the 22nd (2004-04-22
12:42 checkin) with
cvs update -D 2004-04-21
and found that the above queries work fine on that version.
The bug is introduced in that 22nd checkin
## Comment 1114
Date: 2004-04-26 19:27:45 +0200
From: @njnes
Logged In: YES
user_id=43556
set key,aggr,nrcols for each stmt correctly
## Comment 1115
Date: 2005-10-06 13:24:38 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
## Comment 1116
Date: 2005-10-06 13:27:41 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: ALREADY IN TESTWEB
src/test/bugs/union_all-bug-sf-941788
## Comment 1117
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 941788 at http://sourceforge.net/support/tracker.php?aid=941788
| SQL: Monet lies, cheats and bugs on UNION subquery | https://api.github.com/repos/MonetDB/MonetDB/issues/203/comments | 0 | 2020-11-30T08:17:54Z | 2024-06-27T10:56:29Z | https://github.com/MonetDB/MonetDB/issues/203 | 753,261,450 | 203 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-25 12:44:01 +0200
From: @drstmane
To: @mlkersten
Version: -- development
Last updated: 2004-04-28 01:15:56 +0200
## Comment 1102
Date: 2004-04-25 12:44:01 +0200
From: @drstmane
Martin's latest code changes do not compile with GNU
64-bit:
[...]
gcc -m64 -DHAVE_CONFIG_H -I.
-I/var/tmp/monet.GNU_64-d.28893/MonetDB/src/gdk -I../..
-I/var/tmp/monet.GNU_64-d.28893/MonetDB/src/gdk
-I../common -DNDEBUG -O2 -std=c99 -D_POSIX_SOURCE
-D_POSIX_C_SOURCE=199506L -D_XOPEN_SOURCE=600 -Wall -W
-O6 -fomit-frame-pointer -finline-functions
-falign-loops=4 -falign-jumps=4 -falign-functions=4
-fexpensive-optimizations -funroll-loops
-frerun-cse-after-loop -frerun-loop-opt
-Werror-implicit-function-declaration -Werror
-Wno-format -Wno-unused-function -Wno-unused-label
-Wno-strict-aliasing -D_REENTRANT -c gdk_bat.c -fPIC
-DPIC -o gdk_bat.lo
/var/tmp/monet.GNU_64-d.28893/MonetDB/src/gdk/gdk_bat.mx:
In function `BATnewstorage':
/var/tmp/monet.GNU_64-d.28893/MonetDB/src/gdk/gdk_bat.mx:180:
warning: comparison is always false due to limited
range of data type
/var/tmp/monet.GNU_64-d.28893/MonetDB/src/gdk/gdk_bat.mx:180:
warning: comparison is always false due to limited
range of data type
/var/tmp/monet.GNU_64-d.28893/MonetDB/src/gdk/gdk_bat.mx:180:
warning: comparison is always false due to limited
range of data type
make[4]: *** [gdk_bat.lo] Error 1
[...]
see
http://monetdb.cwi.nl/Development/TestWeb/Stable/MonetDB/MonetDB/.GNU_64-d/Linux2.4.24/make-k.out
for details (look for "***").
## Comment 1103
Date: 2004-04-25 13:01:43 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Indeed, correct warning for an intentional action.
This piece of code should be enclosed by a ifdef
which is only set for 64-bit compilation.
Didn't (don't) know which one. Anyone ?
## Comment 1104
Date: 2004-04-25 13:03:48 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Sorry, should be defined for 32-bit compilation.
It is not needed in 64-bit case
## Comment 1105
Date: 2004-04-25 16:37:55 +0200
From: @drstmane
Logged In: YES
user_id=572415
Well, the respective code does have some more problems,
*especially* on 32-bit systems:
========
src/gdk/gdk_bat.mx:180:
--------
if( abs(((long)@1)* @2) > ((long)2147483647L) &&
sizeof(long*)==4 ){
========
1)
"long" is concidered eval, since there is no standard, how
big it is; usually, it's 32-bit on 32-bit archs and 64-bit
on 64-bit archs, but there are exceptions, hence, you should
*not* use it.
As a "growing" integer type, you should use "size_t"
(unsigned) or "ssize_t" (signed); though in this case, a
real 64-bit type ("lng") might be a good idea even
(especially) in the 32-bit case:
2)
the product on the left side of your predicate can easily
overflow (especially on 32-bit systems!), and hence, the
predicate incorrectly evaluates to "false"...
I might try to fix this later, but I first have to
understand all of your changes...
ps:
you might use either of the following to check for 64-bit at
compile time:
if SIZEOF_VOID_P==8
if SIZEOF_SIZE_T==8
## Comment 1106
Date: 2004-04-25 16:43:01 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Actually, this size test can be completely removed, because
the changes made to HEAPalloc ensure that you can react on a
failure to of GDKvmalloc
## Comment 1107
Date: 2004-04-26 09:42:32 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
As a followup to Stefan's suggestion, there are a whole
bunch of ways to check for 64-bit-ness, all with slightly
different semantics. It is important to choose the
appropriate one, i.e. the one that tests what you need to test:
SIZEOF_INT==8
SIZEOF_VOID_P==8
SIZEOF_SIZE_T==8
SIZEOF_OID==8
All of the above left-hand sides == SIZEOF_LNG
## Comment 1108
Date: 2004-04-28 13:15:56 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Code has been dropped.
## Comment 1109
Date: 2005-10-06 13:14:24 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
## Comment 1110
Date: 2005-10-06 13:14:59 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: NO TEST / COMPILATION
## Comment 1111
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 941698 at http://sourceforge.net/support/tracker.php?aid=941698
| latest code changes do not compile with GNU 64-bit | https://api.github.com/repos/MonetDB/MonetDB/issues/202/comments | 0 | 2020-11-30T08:17:51Z | 2024-06-27T10:56:28Z | https://github.com/MonetDB/MonetDB/issues/202 | 753,261,418 | 202 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-25 12:28:39 +0200
From: @drstmane
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-04-26 05:31:31 +0200
## Comment 1097
Date: 2004-04-25 12:28:39 +0200
From: @drstmane
Since Peter's fix of BUG #936469 "BATpropcheck
incorrect after outerjoin", then "join" command seems
to be broken, at least it now gives different/empty
output in tests:
src/modules/contrib/Tests/mprof
src/modules/contrib/Tests/mprof2
src/modules/plain/Tests/decimal
tests/BugReports/Tests/no.024
tests/BugReports/Tests/no.145
tests/BugReports/Tests/no.162
tests/from_Twente/moa/Tests/list
tests/from_Twente/moa/Tests/test
tests/suite_02/Tests/test_70
This Bug report applies to MonetDB 4.3.16 & MonetDB 4.3.17.
## Comment 1098
Date: 2004-04-26 17:31:31 +0200
From: @drstmane
Logged In: YES
user_id=572415
Peter's "HOTFIX" does solve the problem.
Thanks!
## Comment 1099
Date: 2005-10-06 13:02:27 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: Claimed by Sandor.
## Comment 1100
Date: 2005-10-06 13:11:28 +0200
From: @BitGrit
Logged In: YES
user_id=1292024
BugDay_2005-10-06: ALREADY IN TESTWEB
## Comment 1101
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 941693 at http://sourceforge.net/support/tracker.php?aid=941693
| "join" seems broken since bug fix #936469 | https://api.github.com/repos/MonetDB/MonetDB/issues/201/comments | 0 | 2020-11-30T08:17:48Z | 2024-06-27T10:56:27Z | https://github.com/MonetDB/MonetDB/issues/201 | 753,261,382 | 201 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-24 09:31:21 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-04-24 01:43:56 +0200
## Comment 1093
Date: 2004-04-24 09:31:21 +0200
From: @mlkersten
start Mserver and MapiClient -l sql and issue
set explain="mil";
## Comment 1094
Date: 2004-04-24 13:43:56 +0200
From: @njnes
Logged In: YES
user_id=43556
added check if right handside is valid (ie. not NULL)
## Comment 1095
Date: 2005-10-06 19:13:51 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / ALREADY IN TESTWEB
other test about set, that include this, exist
## Comment 1096
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 941194 at http://sourceforge.net/support/tracker.php?aid=941194
| SQL set instruction crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/200/comments | 0 | 2020-11-30T08:17:44Z | 2024-06-27T10:56:26Z | https://github.com/MonetDB/MonetDB/issues/200 | 753,261,338 | 200 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-22 09:41:57 +0200
From: @grobian
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2007-11-30 04:20:04 +0100
## Comment 1076
Date: 2004-04-22 09:41:57 +0200
From: @grobian
consider the following dialogue:
Welcome to the MonetDB interactive JDBC terminal!
Database: MonetDB 4.3.17 (4.3)
Driver: MonetDB Native Driver 0.4 beta release (0.4)
Type \q to quit, \h for a list of available commands
auto commit mode: on
monetdb-> select count(*) from voyages;
+----------
| count_mediumint
+----------
| 8141
+----------
1 rows
monetdb-> select count(*) from invoices;
+----------
| count_mediumint
+----------
| 3836
+----------
1 rows
monetdb-> select count(*) from craftsmen;
+----------
| count_mediumint
+----------
| 2372
+----------
1 rows
monetdb-> select count(*) from voyages, invoices,
craftsmen;
Error on line 4: java.io.IOException: End of stream reached
Executed query: select count(*) from voyages, invoices,
craftsmen;
monetdb->
Mserver dies without a notice after a minute or two.
top reports a memory usage of around 500MB, but my
systems is swapping heavily making it unusable for the
time that Mserver is alive.
Dunno if this can make any difference, but I'm using
blocked mode Mapi interface. Will try with line mode as
well using MapiClient.
## Comment 1077
Date: 2004-04-22 09:47:48 +0200
From: @grobian
Logged In: YES
user_id=963970
MapiClient does exactly the same:
sql>select count(*) from voyages, invoices, craftsmen;
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = select count(*) from voyages, invoices, craftsmen;
ERROR = Battle control terminated
## Comment 1078
Date: 2004-04-22 10:03:54 +0200
From: @drstmane
Logged In: YES
user_id=572415
hm, be aware of the fact, that this creates a temp. BAT of
31,228,876 tuples, i.e., about 238 MB in case the BAT is
[oid,oid], hence, there *might* be a memory ...
It does work in a simple test, though:
=======
mil>module(radix);
mil>a:=uniform(8141).project(0);
mil>a.slice(0,1).print;
-------------------------
h tmp_303 name
oid int type
-------------------------
[ 0@0, 0 ]
[ 1@0, 0 ]
mil>b:=uniform(3836).project(0).reverse;
mil>b.slice(0,1).print;
-------------------------
t tmp_303 name
int oid type
-------------------------
[ 0, 0@0 ]
[ 0, 1@0 ]
mil>c:=join(a,b);
mil>c.count.print;
[ 31228876 ]
mil>
========
## Comment 1079
Date: 2004-04-22 10:11:49 +0200
From: @grobian
Logged In: YES
user_id=963970
Yup, I'm aware of that I'm generating a HUGE result set...
but hey consider me to be an SQL starter, then why does the
DB crash on me?!? What did I do wrong? In the worst case I
expected to get three columns with rowcounts for each table!
(LOL)
I would be happy if Mserver terminated with an out of memory
error.
I would be ultimate happy if Mserver would cut off my client
connection with an too much resources used error.
I would be in the seventh heaven if Mserver terminated the
current query execution and returned an error to the client
saying 'current execution terminated: out of memory' and
maintained the connection to my client.
## Comment 1080
Date: 2004-04-22 12:38:02 +0200
From: @njnes
Logged In: YES
user_id=43556
The memory model of monet is fare from save. There is no way the
sql frontend could return an error as the underlying BATnew
calls
will not return a error (NULL) but just crash! Adding checks
for all
returns of BATnew (and other BAT returning functions) is
currently considered a to large job.
## Comment 1081
Date: 2004-04-23 10:53:21 +0200
From: @peterboncz
Logged In: YES
user_id=591107
there are two approaches to this problem:
(a) fix all GDK/module primitives that receive BATs directly
from other GDK/module C primitives. They should properly
collect their garbage and also return NULL
(b) implement a Mguardian that at least keeps all Mapiclients
open, restarts the Mserver, reconnects and just tells all
clients that their transactions where aborted.
I have no specific preference as of yet.
## Comment 1082
Date: 2004-04-23 11:01:15 +0200
From: @grobian
Logged In: YES
user_id=963970
solution b is dirty.
It implicates that all sessions to the particular server are
invalid which more or less requires a reconnect of the
client(s) involved.
## Comment 1083
Date: 2004-04-24 09:13:17 +0200
From: @mlkersten
Logged In: YES
user_id=490798
The original query would produced 8141*3836*2372
= 74074893872 results.
It is unclear in which stage of the MIL execution the server
crashes. I suspect after the first join upon allocation of
the temporaries for the second halve.
A first line of defense is to issue an emergency stop if a
BAT request size goes beyond a highwatermark level.
Possible responses:
1) GDKfatal, too crude
2) GDKerror is set and the actual size is trimmed to an
acceptable level and making the BAT readonly to avoid any
further update (tryout with a patch in gdk_bat.mx)
## Comment 1084
Date: 2004-05-06 19:51:55 +0200
From: @mlkersten
Logged In: YES
user_id=490798
The source code has been patched to react to failures of
this kind.
An error is produced and the system continous working.
## Comment 1085
Date: 2004-05-07 11:36:14 +0200
From: @grobian
Logged In: YES
user_id=963970
seems to work:
monetdb-> select count(*) from voyages, craftsmen, passengers;
Error on line 2: ERROR: GDKload: cannot mmap(): name=01/102,
ext=buns.priv
OS: Cannot allocate memory
ERROR: BATpropcheck: BAT parameter: BAT required.
ERROR: CMDjoin: operation failed.
Executed query: select count(*) from voyages, craftsmen,
passengers;
monetdb->
BUT....
something gets messed up, because issuing the same statement
again does:
monetdb-> select count(*) from voyages, craftsmen, passengers;
Error on line 3: ERROR: interpret: no matching MIL operator
to 'sql2()'.
Executed query: select count(*) from voyages, craftsmen,
passengers;
So, life is better, but it looks like the SQL session gets
affected after this error. (Simple) other select statements
do work.
## Comment 1086
Date: 2005-10-06 19:10:12 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / POSTPONED
The test also depends on the available amount of memory...
## Comment 1087
Date: 2005-11-06 10:56:27 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 1088
Date: 2005-11-21 04:20:19 +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 1089
Date: 2006-11-22 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 1090
Date: 2006-11-29 10:05:21 +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 1091
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 1092
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 939842 at http://sourceforge.net/support/tracker.php?aid=939842
| SQL: Meltdown: cartesian product crashes server? | https://api.github.com/repos/MonetDB/MonetDB/issues/199/comments | 0 | 2020-11-30T08:17:40Z | 2024-06-27T10:56:25Z | https://github.com/MonetDB/MonetDB/issues/199 | 753,261,299 | 199 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-21 09:41:30 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-04-22 12:47:26 +0200
## Comment 1072
Date: 2004-04-21 21:41:30 +0200
From: @mlkersten
select distinct count(columns.name) from columns;
produces CTgroup error message
## Comment 1073
Date: 2004-04-22 12:47:26 +0200
From: @njnes
Logged In: YES
user_id=43556
Duplicate of
939397 'SELECT DISTINCT 1 from tables' fails
## Comment 1074
Date: 2005-10-06 18:49:34 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / duplicate
## Comment 1075
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 939554 at http://sourceforge.net/support/tracker.php?aid=939554
This bug was marked DUPLICATE in the database it was moved from.
Changing resolution to "MOVED"
| SQL distinct count error | https://api.github.com/repos/MonetDB/MonetDB/issues/198/comments | 0 | 2020-11-30T08:17:36Z | 2024-06-28T13:36:53Z | https://github.com/MonetDB/MonetDB/issues/198 | 753,261,247 | 198 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-21 06:22:16 +0200
From: Jan Rittinger <<jan.rittinger>>
To: @njnes
Version: -- development
Last updated: 2004-04-22 12:47:02 +0200
## Comment 1067
Date: 2004-04-21 18:22:16 +0200
From: Jan Rittinger <<jan.rittinger>>
No cartesian product for constants is made. The answer should
contain 10+ times '1'.
sql>select * from tables;
id, name, schema_id, query, type name
mediumint, varchar, mediumint, varchar,
mediumint type
14, , , , tuplecount
9, , , , id
[ 1004581, "modules", 1004571, "", 1 ]
[ 1004626, "schemas", 1004571, "", 1 ]
[ 1004630, "tables", 1004571, "", 1 ]
[ 1004636, "columns", 1004571, "", 1 ]
[ 1004646, "keys", 1004571, "", 1 ]
[ 1004652, "idxs", 1004571, "", 1 ]
[ 1004657, "keycolumns", 1004571, "", 1 ]
[ 1004662, "types", 1004571, "", 1 ]
[ 1004670, "users", 1004571, "", 1 ]
[ 1004693, "user_role", 1004571, "", 1 ]
[ 1004716, "auths", 1004571, "", 1 ]
[ 1004734, "privileges", 1004571, "", 1 ]
[ 1004763, "history", 1004571, "", 1 ]
[ 1004801, "env", 1004571, "", 1 ]
sql>select 1 from tables;
single_value name
int type
1 tuplecount
10 id
[ 1 ]
## Comment 1068
Date: 2004-04-22 12:47:02 +0200
From: @njnes
Logged In: YES
user_id=43556
Duplicate of
939397 'SELECT DISTINCT 1 from tables' fails
## Comment 1069
Date: 2004-04-23 10:55:32 +0200
From: @peterboncz
Logged In: YES
user_id=591107
If I recall correctly, 'SELECT DISTINCT 1 from tables'
produces an MIL error with a wrong signature given to a
CTgroup, whereas 'SELECT1 from tables' just gives a wrong
answer.
but maybe internally in SQL it is the same bug
## Comment 1070
Date: 2005-10-06 18:49:00 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / DUPLICATE
## Comment 1071
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 939402 at http://sourceforge.net/support/tracker.php?aid=939402
This bug was marked DUPLICATE in the database it was moved from.
Changing resolution to "MOVED"
| No Cartesian Product from Constants | https://api.github.com/repos/MonetDB/MonetDB/issues/197/comments | 0 | 2020-11-30T08:17:33Z | 2024-06-28T13:36:53Z | https://github.com/MonetDB/MonetDB/issues/197 | 753,261,216 | 197 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-21 06:16:41 +0200
From: Jan Rittinger <<jan.rittinger>>
To: @njnes
Version: -- development
Last updated: 2004-04-22 12:50:32 +0200
## Comment 1063
Date: 2004-04-21 18:16:41 +0200
From: Jan Rittinger <<jan.rittinger>>
QUERY = SELECT DISTINCT 1 from tables;
ERROR = !ERROR: interpret: no matching MIL operator to
'CTgroup(sht)'.
!MAYBE YOU MEAN:
! CTgroup(BAT[oid,any], BAT[oid,any]) : BAT[oid,oid]
! CTgroup(BAT[oid,any]) : BAT[oid,oid]
## Comment 1064
Date: 2004-04-22 12:50:32 +0200
From: @njnes
Logged In: YES
user_id=43556
The with from/where query without groupby/aggr but with single
value selection part wasn't handled.
## Comment 1065
Date: 2005-10-06 18:47:18 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / ALREADY IN TESTWEB
## Comment 1066
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 939397 at http://sourceforge.net/support/tracker.php?aid=939397
| 'SELECT DISTINCT 1 from tables' fails | https://api.github.com/repos/MonetDB/MonetDB/issues/196/comments | 0 | 2020-11-30T08:17:30Z | 2024-06-27T10:56:22Z | https://github.com/MonetDB/MonetDB/issues/196 | 753,261,188 | 196 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-21 05:22:15 +0200
From: @sjoerdmullender
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-04-21 07:27:41 +0200
## Comment 1059
Date: 2004-04-21 17:22:15 +0200
From: @sjoerdmullender
make html in SQL build directory fails with
make: *** No rule to make target
`/ufs/sjoerd/src/MonetDB/sql/jdbcmanual.html',
needed by `/ufs/sjoerd/Monet/doc/jdbcmanual.html'. Stop.
## Comment 1060
Date: 2004-04-21 19:27:41 +0200
From: @drstmane
Logged In: YES
user_id=572415
fixed html.mk
## Comment 1061
Date: 2005-10-06 18:42:58 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / COMPILATION
## Comment 1062
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 939373 at http://sourceforge.net/support/tracker.php?aid=939373
| make html in SQL fails | https://api.github.com/repos/MonetDB/MonetDB/issues/195/comments | 0 | 2020-11-30T08:17:27Z | 2024-06-27T10:56:21Z | https://github.com/MonetDB/MonetDB/issues/195 | 753,261,158 | 195 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-21 05:14:46 +0200
From: Jan Rittinger <<jan.rittinger>>
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2004-04-22 12:43:45 +0200
## Comment 1053
Date: 2004-04-21 17:14:46 +0200
From: Jan Rittinger <<jan.rittinger>>
http://monetdb.cwi.nl/TechDocs/Additions/SQL/index.html
## Comment 1054
Date: 2004-04-21 17:43:49 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
This is most likely because "make html" in the SQL build
tree fails. See http://sourceforge.net/support/tracker.php?aid=939373.
I just checked in a fix to make the SQL html before the
(failing) jdbc html. We'll see tomorrow if that fixes this
problem.
## Comment 1055
Date: 2004-04-22 10:05:04 +0200
From: @grobian
Logged In: YES
user_id=963970
JDBC manual doesn't generate errors anymore, but it has,
just like the SQL-Guide the same problem, that all links
don't work.
for the JDBC docs it seems that the only files placed on the
webserver are:
content.shtml* sidebar-b.shtml* subsubsubmenu.shtml*
index.html* submenu.shtml* title.shtml*
link.shtml* subsubmenu.shtml*
which feels like no JDBC documentation at all. Do the files
get actually copied?
## Comment 1056
Date: 2004-04-22 12:43:45 +0200
From: @njnes
Logged In: YES
user_id=43556
guide seems back!
## Comment 1057
Date: 2005-10-06 18:42:26 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / DOCUMENTATION
## Comment 1058
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 939366 at http://sourceforge.net/support/tracker.php?aid=939366
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.
| SQL-Guide is not linked | https://api.github.com/repos/MonetDB/MonetDB/issues/194/comments | 0 | 2020-11-30T08:17:24Z | 2024-06-27T10:56:20Z | https://github.com/MonetDB/MonetDB/issues/194 | 753,261,133 | 194 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-21 09:23:53 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-22 03:50:18 +0200
## Comment 1047
Date: 2004-04-21 09:23:53 +0200
From: @grobian
SELECT null AS TABLE_CAT, schemas.name AS TABLE_SCHEM,
tables.name AS TABLE_NAME, columns.name AS COLUMN_NAME,
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, "null" AS
nullable, null 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, null AS SCOPE_CATALOG, null AS
SCOPE_SCHEMA, null AS SCOPE_TABLE FROM columns, tables,
schemas WHERE columns.table_id = tables.id AND
tables.schema_id = schemas.id AND schemas.name LIKE
'sys' AND tables.name LIKE 'tables' ORDER BY
TABLE_SCHEM, TABLE_NAME, ORDINAL_POSITION;
triggers the following error:
!Column: columns.mediumint unknown
The only place I can think of that this happens is in
columns.type, but taking that one apart doesn't
generate the error.
## Comment 1048
Date: 2004-04-21 09:37:38 +0200
From: @grobian
Logged In: YES
user_id=963970
The smallest script that causes this bug:
SELECT columns.number
FROM columns
Looks like columns."number" does work, so number is a
reserved keyword or something else by now?
## Comment 1049
Date: 2004-04-21 21:46:20 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Yes, this error seems to occurs at.
select columns.number from columns;
we have to check all important lexicals.
## Comment 1050
Date: 2004-04-22 15:50:18 +0200
From: @njnes
Logged In: YES
user_id=43556
aliases where converted to types in the parser. This is to
early as
sometimes the 'alias' names are also used as names of
columns, ...
Fixed by moving the alias2type conversion happen in the parser
## Comment 1051
Date: 2005-10-06 18:41:54 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=939112 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/reserved_words.SF-939112.*
## Comment 1052
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 939112 at http://sourceforge.net/support/tracker.php?aid=939112
| SQL: !Column: columns.mediumint unknown | https://api.github.com/repos/MonetDB/MonetDB/issues/193/comments | 0 | 2020-11-30T08:17:21Z | 2024-06-27T10:56:19Z | https://github.com/MonetDB/MonetDB/issues/193 | 753,261,105 | 193 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-20 07:49:55 +0200
From: @drstmane
To: @njnes
Version: -- development
Last updated: 2004-04-20 08:58:34 +0200
## Comment 1043
Date: 2004-04-20 19:49:55 +0200
From: @drstmane
For primary keys both of the following syntax
alternatives seem to work:
========
create table pk1 (a1 integer, a2 integer, primary key
(a1));
create table pk2 (b1 integer primary key, b2 integer);
========
For foreign keys, only the first one seems to work, the
second crashes the server instantly:
========
create table fk1 (c1 integer, c2 integer, foreign key
(c1) references pk1 (a1));
create table fk2 (d1 integer references pk2 (b1), d2
integer);
========
I'm actually not quite sure, whether the second is
correct SQL-99, but at least, it should not crash the
server.
Here's an exhaustive sample session:
========
sql>select * from keys;
id, table_id, type, name, rkey name
mediumint, mediumint, mediumint,
mediumint, mediumint type
0, , , , tuplecount
0, , , , id
sql>create table pk1 (a1 integer, a2 integer, primary
key (a1));
sql>select * from keys;
id, table_id, type, name, rkey name
mediumint, mediumint, mediumint,
mediumint, mediumint type
1, , , , tuplecount
1, , , , id
[ 1005591, 1005549, 0, "c1005590",
-1 ]
sql>create table pk2 (b1 integer primary key, b2 integer);
sql>select * from keys;
id, table_id, type, name, rkey name
mediumint, mediumint, mediumint,
mediumint, mediumint type
2, , , , tuplecount
2, , , , id
[ 1005591, 1005549, 0, "c1005590",
-1 ]
[ 1005715, 1005706, 0, "c1005714",
-1 ]
sql>create table fk1 (c1 integer, c2 integer, foreign
key (c1) references pk1 (a1));
sql>select * from keys;
id, table_id, type, name, rkey name
mediumint, mediumint, mediumint,
mediumint, mediumint type
3, , , , tuplecount
3, , , , id
[ 1005591, 1005549, 0, "c1005590",
-1 ]
[ 1005715, 1005706, 0, "c1005714",
-1 ]
[ 1005811, 1005797, 2, "c1005810",
1005591 ]
sql>create table fk2 (d1 integer foreign key references
pk2 (b1), d2 integer);
MAPI = monetdb@localhost:60002
ACTION= mapi_get_row
QUERY = create table fk2 (d1 integer foreign key
references pk2 (b1), d2 integer);
ERROR = !ERROR parse error at token (374) in statement:
create table fk2 (d1 integer foreign
sql>rollback;
sql>create table fk2 (d1 integer references pk2 (b1),
d2 integer);
MAPI = monetdb@localhost:60002
ACTION= mapi_fetch_line
QUERY = create table fk2 (d1 integer references pk2
(b1), d2 integer);
ERROR = Connection terminated
monetdb manegold@draco:/var/tmp/MonetDB $
========
The behaviour is identical in the released and in the
latest version.
## Comment 1044
Date: 2004-04-20 20:58:34 +0200
From: @njnes
Logged In: YES
user_id=43556
Couldn't reproduce the crash!
Correct syntax for column references is
sql>create table pk2 (b1 integer primary key, b2 integer);
sql>create table fk2 (d1 integer references pk2 (b1), d2
integer);
sql>\q
But there was an other bug in this code (table name was
treated as column name)
## Comment 1045
Date: 2005-10-06 18:33:22 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: ALREADY IN TESTWEB
sql/src/test/bugs/Tests/bug-sf-938733.sql
## Comment 1046
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 938733 at http://sourceforge.net/support/tracker.php?aid=938733
| wrong(?) alternative foreign key syntax crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/192/comments | 0 | 2020-11-30T08:17:18Z | 2024-06-27T10:56:18Z | https://github.com/MonetDB/MonetDB/issues/192 | 753,261,068 | 192 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-20 07:27:42 +0200
From: @drstmane
To: @njnes
Version: -- development
Last updated: 2004-04-20 07:47:40 +0200
## Comment 1039
Date: 2004-04-20 19:27:42 +0200
From: @drstmane
In the latest (Tuesday, April 20 2004) SQL +
MonetDB-HEAD (4.3.17) and SQL + MonetDB_4-3-16-HEAD,
"create schema" seems to be broken; well, it does not
give an error, but a subsequent "select * from
schemas;" does:
========
sql>select * from schemas;
id, name, authorization name
mediumint, varchar, varchar type
2, , tuplecount
0, , id
[ 1004594, "sys", "sqladmin" ]
[ 1004785, "default-schema", "sqladmin" ]
sql>create schema abc;
sql>select * from schemas;
MAPI = monetdb@localhost:50002
ACTION= mapi_get_row
QUERY = select * from schemas;
ERROR = !Unknown table schemas
sql>select * from schemas;
MAPI = monetdb@localhost:50002
ACTION= mapi_get_row
QUERY = select * from schemas;
ERROR = !current transaction is aborted (please ROLLBACK)
sql>rollback;
sql>select * from schemas;
id, name, authorization name
mediumint, varchar, varchar type
2, , tuplecount
1, , id
[ 1004594, "sys", "sqladmin" ]
[ 1004785, "default-schema", "sqladmin" ]
sql>
========
With the released MonetDB_4-3-16[_release] plus the
respective SQL, the problem does not occur:
========
sql>select * from schemas;
id, name, authorization name
mediumint, varchar, varchar type
2, , tuplecount
0, , id
[ 1004610, "sys", "sqladmin" ]
[ 1004806, "default-schema", "sqladmin" ]
sql>create schema abc;
sql>select * from schemas;
id, name, authorization name
mediumint, varchar, varchar type
3, , tuplecount
1, , id
[ 1004610, "sys", "sqladmin" ]
[ 1004806, "default-schema", "sqladmin" ]
[ 1006118, "abc", "sqladmin" ]
sql>
========
## Comment 1040
Date: 2004-04-20 19:47:40 +0200
From: @njnes
Logged In: YES
user_id=43556
create schema abc; Now also changes the schema, ie moved from
sys or default-schema to abc. Try select * from sys.schemas;
## Comment 1041
Date: 2005-10-06 18:23:51 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 1042
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 938718 at http://sourceforge.net/support/tracker.php?aid=938718
| "create schema" seems to be broken | https://api.github.com/repos/MonetDB/MonetDB/issues/191/comments | 0 | 2020-11-30T08:17:14Z | 2024-06-28T13:36:52Z | https://github.com/MonetDB/MonetDB/issues/191 | 753,261,034 | 191 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-20 04:40:29 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-20 05:25:29 +0200
## Comment 1035
Date: 2004-04-20 16:40:29 +0200
From: @arjan
sql>select * from test;
a name
varchar type
2 tuplecount
1 id
[ "Aap" ]
[ "Noot" ]
sql>select concat(a, 'x') from test;
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = select concat(a, 'x') from test;
ERROR = Connection terminated
Error message is:
monet>Mserver: sql_scope.mx:218: find_pivot: Assertion
`0' failed.
## Comment 1036
Date: 2004-04-20 17:25:29 +0200
From: @njnes
Logged In: YES
user_id=43556
was allready fixed in head
## Comment 1037
Date: 2005-10-06 18:23:26 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 1038
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 938606 at http://sourceforge.net/support/tracker.php?aid=938606
| Constants + row variables in functions crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/190/comments | 0 | 2020-11-30T08:17:10Z | 2024-06-28T13:37:49Z | https://github.com/MonetDB/MonetDB/issues/190 | 753,260,993 | 190 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-20 03:57:02 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-22 12:42:54 +0200
## Comment 1031
Date: 2004-04-20 15:57:02 +0200
From: @arjan
sql>create table t (i int);
sql>create view t2 as select * from t;
sql>drop table t;
sql>select * from t2;
After the last statement, the server crashes.
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1160997680 (LWP 6323)]
0x4186cfc0 in scope_add_table_columns (sql=0x84388a0,
scp=0x8606870,
t=0x8604ab0, tname=0x84cfa38 "t2") at
sql_semantic.mx:210
210 for ( n = t->columns.set->h, m
= sq->op1.lval->h;
Current language: auto; currently c
(gdb) bt
0 0x4186cfc0 in scope_add_table_columns
(sql=0x84388a0, scp=0x8606870,
t=0x8604ab0, tname=0x84cfa38 "t2") at
sql_semantic.mx:210
1 0x41868143 in table_ref (sql=0x84388a0,
scp=0x8606870, tableref=0x8606528)
at sql_select.mx:98
2 0x418681fb in sql_subquery (sql=0x84388a0,
scp=0x8606870, sq=0x84cffe8)
at sql_select.mx:139
## Comment 1032
Date: 2004-04-22 12:42:54 +0200
From: @njnes
Logged In: YES
user_id=43556
drop table now checks if views depend on it, drop will fail
in that case
## Comment 1033
Date: 2005-10-06 18:22:51 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=938583 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/view_on_deleted_table.SF-938583.*
## Comment 1034
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 938583 at http://sourceforge.net/support/tracker.php?aid=938583
| View on non-existent table crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/189/comments | 0 | 2020-11-30T08:17:06Z | 2024-06-27T10:56:15Z | https://github.com/MonetDB/MonetDB/issues/189 | 753,260,952 | 189 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-20 01:39:55 +0200
From: @arjan
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2004-04-20 04:02:06 +0200
## Comment 1027
Date: 2004-04-20 13:39:55 +0200
From: @arjan
The latest CVS checkout gives a wrong cartesian product:
sql>select count(*) from test a;
count_id name
mediumint type
1 tuplecount
8 id
[ 4 ]
sql>select count(*) from test a, test;
count_id name
mediumint type
1 tuplecount
10 id
[ 32 ]
This should yield 4*4 = 16. It seems it adds undefined
tuples in the second relation (with null values).
By the way; on a view the cartesian product is correct:
sql>select count(*) from test2 a, test2 b;
count_id name
mediumint type
1 tuplecount
13 id
[ 16 ]
## Comment 1028
Date: 2004-04-20 16:02:06 +0200
From: @arjan
Logged In: YES
user_id=20087
Re-creating the table fixed the problem, somehow.
## Comment 1029
Date: 2005-10-06 18:17:36 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 1030
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 938504 at http://sourceforge.net/support/tracker.php?aid=938504
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.
| Invalid cartesian product | https://api.github.com/repos/MonetDB/MonetDB/issues/188/comments | 0 | 2020-11-30T08:17:03Z | 2024-06-28T07:36:32Z | https://github.com/MonetDB/MonetDB/issues/188 | 753,260,913 | 188 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-20 01:10:21 +0200
From: @swingbit
To: @njnes
Version: -- development
Last updated: 2004-04-20 02:53:03 +0200
## Comment 1023
Date: 2004-04-20 13:10:21 +0200
From: @swingbit
CREATE TABLE t1 (
id INT,
text1 VARCHAR(32)
);
INSERT INTO t1 (id) VALUES(1);
INSERT INTO t1 (text1) VALUES('test1');
The first INSERT assumes an empty string for the
missing value of text1 field.
The second INSERT says:
ERROR = !column id has no valid default value
It seems that default values exist for strings but not
for numeric types.
Anyway, I think this behaviour is incorrect, even for
strings. As far as I know default values should work as
follows:
When a value is not provided by the INSERT statement,
- if a DEFAULT *clause* was not specified in the column
definition, then a NULL value should be assumed
- if a DEFAULT *value* was not specified in the column
definition (.i.e. DEFAULT clause with no value), then a
default value should be assumed depending on the column
type (empty string for strings, 0 for numeric, etc..)
- if a complete DEFAULT clause was specified, then that
value should be assumed.
The current implementation always tries to fall in the
second case. In the example provided above should
instead assume NULL values.
## Comment 1024
Date: 2004-04-20 14:53:03 +0200
From: @njnes
Logged In: YES
user_id=43556
we now correctly use NULL if now default clause is given.
We do not support DEFAULTs with out values!
## Comment 1025
Date: 2005-10-06 18:17:00 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / OBSOLETE
the DEFAULT is now supported.
## Comment 1026
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 938490 at http://sourceforge.net/support/tracker.php?aid=938490
| DEFAULT values | https://api.github.com/repos/MonetDB/MonetDB/issues/187/comments | 0 | 2020-11-30T08:16:59Z | 2024-06-27T10:56:11Z | https://github.com/MonetDB/MonetDB/issues/187 | 753,260,840 | 187 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-20 10:37:12 +0200
From: @sjoerdmullender
To: @njnes
Version: -- development
Last updated: 2004-04-20 02:58:17 +0200
## Comment 1019
Date: 2004-04-20 10:37:12 +0200
From: @sjoerdmullender
Start Mserver and then from MapiClient try this:
select cast(0 as integer);
select cast(0 as integer);
The second select causes the server to crash with an
assertion failure.
Mserver:
/ufs/sjoerd/src/MonetDB/sql/src/server/sql_symbol.mx:474:
dnode_cmp: Assertion `0' failed.
(The casts can be to anything legal and don't have to
be the same.)
## Comment 1020
Date: 2004-04-20 14:58:17 +0200
From: @njnes
Logged In: YES
user_id=43556
dnode_cmp now also handles data value 'type' (ie does a
subtype_cmp)
## Comment 1021
Date: 2005-10-06 18:13:53 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=938419 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/multiple_select_cast.SF-938419.*
## Comment 1022
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 938419 at http://sourceforge.net/support/tracker.php?aid=938419
| crash after cast | https://api.github.com/repos/MonetDB/MonetDB/issues/186/comments | 0 | 2020-11-30T08:16:56Z | 2024-06-27T10:56:10Z | https://github.com/MonetDB/MonetDB/issues/186 | 753,260,796 | 186 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-19 05:09:59 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-19 07:50:11 +0200
## Comment 1015
Date: 2004-04-19 17:09:59 +0200
From: @arjan
The following (imho correct) query yields a parse error:
sql>select usr, count(cnt) from test group by usr order
by count(cnt);
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select usr, count(cnt) from test group by usr
order by count(cnt);
ERROR = !ERROR parse error at token (40) in statement:
select usr, count(cnt) from test group by usr order by
count(
Am I doing something wrong here?
## Comment 1016
Date: 2004-04-19 18:52:29 +0200
From: @grobian
Logged In: YES
user_id=963970
yep sure you do :)
select usr, count(cnt) as cnt from test group by usr order
by cnt;
the order by only knows the final columns, and for the count
you don't know how it's called (is different on each
database). Giving it an alias makes it sure you know the
right name.
## Comment 1017
Date: 2005-10-06 18:07:12 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 1018
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 937951 at http://sourceforge.net/support/tracker.php?aid=937951
| ordering on aggregate function | https://api.github.com/repos/MonetDB/MonetDB/issues/185/comments | 0 | 2020-11-30T08:16:53Z | 2024-06-28T07:36:32Z | https://github.com/MonetDB/MonetDB/issues/185 | 753,260,762 | 185 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-19 02:47:54 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-20 02:47:41 +0200
## Comment 1010
Date: 2004-04-19 14:47:54 +0200
From: @arjan
sql>select a.name, b.name, fstrcmp(a.name, b.name) from
britney a, britney b where a.id > b.id and
fstrcmp(a.name, b.name)>0.9 limit 10;
name, name, fstrcmp_name name
varchar, varchar, double type
10, , tuplecount
46, , id
[ "Britney, Nsync, Nelly, Aerosmith", "Britney",
0.35897435897435898 ]
[ "Britney, Nsync, Nelly, Aerosmith", "Spritney
Bears", 0.43478260869565216 ]
[ "Britney, Nsync, Nelly, Aerosmith", "Metallica and
Britney Spears", 0.36666666666666664 ]
[ "Britney, Nsync, Nelly, Aerosmith", "Eminem and
Britney Spears", 0.42105263157894735 ]
[ "Britney, Nsync, Nelly, Aerosmith", "Britney
Cleary", 0.47826086956521741 ]
The third column should be > 0.9 always, as specified.
It is not, as can be seen.
The custom fstrcmp function just takes two strings as
input, and a dbl as output.
By the way, i have not been able to reproduce these
results by using another function, yet.
## Comment 1011
Date: 2004-04-19 14:57:37 +0200
From: @arjan
Logged In: YES
user_id=20087
Okay the same problem seems to exist with the query below.
sql>select * from test;
id, s name
int, varchar type
4, tuplecount
74, id
[ 1, "a" ]
[ 2, "b" ]
[ 3, "c" ]
[ 4, "aap" ]
sql>select char_length(a.s)-char_length(b.s) from test a,
test b where a.id < b.id and
char_length(a.s)-char_length(b.s)=0;
sql_sub_char_length_s name
mediumint type
6 tuplecount
75 id
[ 0 ]
[ 0 ]
[ -2 ]
[ 0 ]
[ -2 ]
[ -2 ]
## Comment 1012
Date: 2004-04-20 14:47:41 +0200
From: @njnes
Logged In: YES
user_id=43556
added implicit join
## Comment 1013
Date: 2005-10-06 18:06:22 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=937859 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/custom_function_in_where.SF-937859.*
## Comment 1014
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 937859 at http://sourceforge.net/support/tracker.php?aid=937859
| custom function in where clause | https://api.github.com/repos/MonetDB/MonetDB/issues/184/comments | 0 | 2020-11-30T08:16:50Z | 2024-06-27T10:56:08Z | https://github.com/MonetDB/MonetDB/issues/184 | 753,260,736 | 184 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-19 02:07:55 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-20 05:12:15 +0200
## Comment 1005
Date: 2004-04-19 14:07:55 +0200
From: @arjan
sql>create table test (id int, s varchar(255));
sql>insert into test values(1, 'a');
sql>insert into test values(2, 'b');
sql>insert into test values(3, 'c');
sql>commit;
sql>select a.id, b.id from test a,test b where a.id < b.id;
id, id name
int, int type
3, tuplecount
20, id
[ 1, 2 ]
[ 1, 3 ]
[ 2, 3 ]
** So far, so good. But now the above query applied on
a view:
sql>create view test2 as select * from test;
sql>commit;
sql>select a.id, b.id from test2 a,test2 b where a.id <
b.id;
id, id name
int, int type
0, tuplecount
21, id
sql>
This should yield the same result as the previous
query, but it does not... strange.
## Comment 1006
Date: 2004-04-20 10:20:25 +0200
From: @njnes
Logged In: YES
user_id=43556
In the view case the test2 a and b where seen as the same
table, ie
a.id < b.id was evaluated as a.id < a.id. Fixed but checking
peding
## Comment 1007
Date: 2004-04-20 13:41:14 +0200
From: @arjan
Logged In: YES
user_id=20087
With the latest CVS checkout, this bug still occurs.
Maybe this apparent 'fix' is the cause of http://sourceforge.net/support/tracker.php?aid=938504
(cartesian product)
## Comment 1008
Date: 2005-10-06 17:46:54 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: ALREADY IN TESTWEB
MonetDB/tests/bugs/view_selfjoin-bug-sf-937835.*
## Comment 1009
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 937835 at http://sourceforge.net/support/tracker.php?aid=937835
| Weird view behaviour | https://api.github.com/repos/MonetDB/MonetDB/issues/183/comments | 0 | 2020-11-30T08:16:46Z | 2024-06-27T10:56:06Z | https://github.com/MonetDB/MonetDB/issues/183 | 753,260,692 | 183 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-16 11:05:38 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-04-17 07:37:05 +0200
## Comment 1001
Date: 2004-04-16 23:05:38 +0200
From: @mlkersten
Start the server with the sql module
and enter <cntrl-d>
module Monet Database Server V4.3.16
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.
(monet>sql_server);
[New Thread 1102748976 (LWP 12049)]
[New Thread 1104850224 (LWP 12050)]
[New Thread 1106951472 (LWP 12051)]
monet>
Program received signal SIGSEGV, Segmentation fault.
[Switching to Thread 1102748976 (LWP 12049)]
sql_ref_dec (r=0x0) at sql_mem.mx:94
94 return (--r->refcnt);
Current language: auto; currently c
## Comment 1002
Date: 2004-04-17 07:37:05 +0200
From: @njnes
Logged In: YES
user_id=43556
As some types moved from the types table to the alias table,
this
made old db's corrupt as they still see the aliases as full
types.
Solution recreate the db, ie start with an empty db.
## Comment 1003
Date: 2005-10-06 17:42:16 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 1004
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 936597 at http://sourceforge.net/support/tracker.php?aid=936597
| SQL segfault during cntrl-d | https://api.github.com/repos/MonetDB/MonetDB/issues/182/comments | 0 | 2020-11-30T08:16:43Z | 2024-06-28T07:36:31Z | https://github.com/MonetDB/MonetDB/issues/182 | 753,260,651 | 182 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-16 06:54:35 +0200
From: @MarcinZukowski
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-04-23 10:46:36 +0200
## Comment 997
Date: 2004-04-16 18:54:35 +0200
From: @MarcinZukowski
The following script (attached):
<<<
x:=new(void,int);
x.insert(nil,1);
x.insert(nil,2);
x.insert(nil,3);
x.seqbase(1@0);
y:=new(oid,int);
y.insert(1@0,14);
y.insert(3@0,34);
warning here
x.mirror.outerjoin(y);
and here as well
x.reverse.outerjoin(y);
>>>
produces warnings when run with --debug=10 (default in Mtest):
<<<
!WARNING: BATpropcheck: BAT tmp_88 was incorrectly marked
sorted!
>>>
Error lies in improper propagation of 'sorted' flag in mergejoin -
NILs are not taken into account.
Peter's fix is on the way :)
## Comment 998
Date: 2004-04-23 10:46:36 +0200
From: @peterboncz
Logged In: YES
user_id=591107
the propagation code was fixed (both Stable and Current)
## Comment 999
Date: 2005-10-06 17:41:31 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=936469 as
MonetDB/tests/BugDay_2005-10-06_4.9.3/Tests/BATpropcheck_after_outerjoin.SF-936469.*
## Comment 1000
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 936469 at http://sourceforge.net/support/tracker.php?aid=936469
| BATpropcheck incorrect after outerjoin | https://api.github.com/repos/MonetDB/MonetDB/issues/181/comments | 0 | 2020-11-30T08:16:40Z | 2024-06-28T07:36:31Z | https://github.com/MonetDB/MonetDB/issues/181 | 753,260,617 | 181 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-16 04:44:27 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-16 10:31:53 +0200
## Comment 992
Date: 2004-04-16 16:44:27 +0200
From: @grobian
When issuing the JDBC method getTables as the
JdbcClient's function \d does, Mserver outputs the
message ! TODO: stmt_reset(2=var) infinite times.
I'll tear down the query to a minimum later.
## Comment 993
Date: 2004-04-16 20:14:23 +0200
From: @grobian
Logged In: YES
user_id=963970
On a faster computer it came clear that if you wait long
enough the results come back... Mserver just get's slowed
down by printing millions TODO lines on the terminal it is.
The TODO's get exponentially more when more UNIONs are
involved and a WHERE clause with 1=1 is present.
a little example which produces a little amount of TODO's
(the larger the statements, the more the TODO's) is attached
## Comment 994
Date: 2004-04-16 22:31:53 +0200
From: @njnes
Logged In: YES
user_id=43556
added missing case to switch
## Comment 995
Date: 2005-10-06 17:32:51 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: ALREADY IN TESTWEB
sql/src/test/bugs/union_all_where_1=1-bug-sf-936386
## Comment 996
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 936386 at http://sourceforge.net/support/tracker.php?aid=936386
| SQL: infinite repeating message ! TODO: stmt_reset(2=var) | https://api.github.com/repos/MonetDB/MonetDB/issues/180/comments | 0 | 2020-11-30T08:16:36Z | 2024-06-27T10:56:03Z | https://github.com/MonetDB/MonetDB/issues/180 | 753,260,581 | 180 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-16 04:40:48 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-16 10:32:54 +0200
## Comment 988
Date: 2004-04-16 16:40:48 +0200
From: @grobian
select * from (
select * from tables
union all
select * from tables
) as a;
gives only once the records from table 'tables' which
is incorrect because I use UNION ALL to specify I want
duplicates.
## Comment 989
Date: 2004-04-16 22:32:54 +0200
From: @njnes
Logged In: YES
user_id=43556
distinct vs all handling was switched.
## Comment 990
Date: 2005-10-06 17:26:49 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: AREADY IN TESTWEB
sql/src/test/BUGS/except-union-intersect-bug-sf-1146079
## Comment 991
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 936384 at http://sourceforge.net/support/tracker.php?aid=936384
| SQL: UNION ALL acts like UNION DISTINCT | https://api.github.com/repos/MonetDB/MonetDB/issues/179/comments | 0 | 2020-11-30T08:16:34Z | 2024-06-27T10:56:02Z | https://github.com/MonetDB/MonetDB/issues/179 | 753,260,554 | 179 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-16 02:56:07 +0200
From: Jens Teubner <<teubner>>
To: Jens Teubner <<teubner>>
Version: -- development
Last updated: 2004-04-18 10:47:39 +0200
## Comment 983
Date: 2004-04-16 14:56:07 +0200
From: Jens Teubner <<teubner>>
The current build setup sets NDEBUG to disable
debugging stuff (even with --enable-debug). Some code,
however, depends on assert() stuff being executed, in
particular parser.l for the initialization of the
lexical_stack.
## Comment 984
Date: 2004-04-16 15:25:35 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
NDEBUG is only set by default in the 4.3.16 release branch.
You can turn it off (and enable assertions) by using
--enable-assert.
However, one should *never* use side effect in asserts since
it should be possible to turn them off.
## Comment 985
Date: 2004-04-18 10:47:39 +0200
From: Jens Teubner <<teubner>>
Logged In: YES
user_id=731390
Thanks, Sjoerd, you don't have to tell me ;-).
I just checked in the fixes.
## Comment 986
Date: 2005-10-06 17:18:06 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: RELEASED
PathFinder not installed
## Comment 987
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 936290 at http://sourceforge.net/support/tracker.php?aid=936290
| Pathfinder does not function properly with NDEBUG | https://api.github.com/repos/MonetDB/MonetDB/issues/178/comments | 0 | 2020-11-30T08:16:31Z | 2024-06-27T10:56:01Z | https://github.com/MonetDB/MonetDB/issues/178 | 753,260,525 | 178 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-16 02:38:14 +0200
From: @sjoerdmullender
To: @njnes
Version: -- development
Last updated: 2004-04-16 03:27:38 +0200
## Comment 978
Date: 2004-04-16 14:38:14 +0200
From: @sjoerdmullender
Start Mserver .../sql_server.mil
In another window start MapiClient -lsql -umonetdb
-Pmonetdb
Result:
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
ERROR = Connection terminated
from MapiClient and Mserver has exited.
Using gdb reveals that Mserver gets a SEGFAULT when
checking the assertion in sql_ref_inc (sql_mem.mx)
since r == NULL.
## Comment 979
Date: 2004-04-16 15:58:05 +0200
From: @grobian
Logged In: YES
user_id=963970
Confirming!!!
A freshly checked out head branch crashes on connect!!!
## Comment 980
Date: 2004-04-16 16:35:46 +0200
From: @grobian
Logged In: YES
user_id=963970
solution = throw away your old database(s)
## Comment 981
Date: 2005-10-06 17:17:12 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 982
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 936281 at http://sourceforge.net/support/tracker.php?aid=936281
| crash at initial connect | https://api.github.com/repos/MonetDB/MonetDB/issues/177/comments | 0 | 2020-11-30T08:16:28Z | 2024-06-28T07:36:31Z | https://github.com/MonetDB/MonetDB/issues/177 | 753,260,493 | 177 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-16 01:52:16 +0200
From: @sjoerdmullender
To: @njnes
Version: -- development
Last updated: 2004-04-16 02:17:48 +0200
## Comment 973
Date: 2004-04-16 13:52:16 +0200
From: @sjoerdmullender
create table crash_me (a integer not null,b char(10)
not null);
insert into crash_me (a,b) values (1,'a');
select a from crash_me order by 1;
This causes an assertion failed:
Mserver:
/ufs/sjoerd/src/MonetDB/sql/src/server/sql_semantic.mx:2403:
orderby_column_ref: Assertion `column_r->token ==
SQL_COLUMN && column_r->type == type_list' failed.
Aborted
## Comment 974
Date: 2004-04-16 14:09:48 +0200
From: @grobian
Logged In: YES
user_id=963970
For Niels:
This seems valid sql, and it means that the order by applies
to the first column (1) of the result. 2 would mean the
second, and so on.
## Comment 975
Date: 2004-04-16 14:17:48 +0200
From: @njnes
Logged In: YES
user_id=43556
simply droped support for order by 'number', as its no longer
part of sql 99 and up and ugly..
## Comment 976
Date: 2005-10-06 17:16:19 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / OBSOLETE
## Comment 977
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 936255 at http://sourceforge.net/support/tracker.php?aid=936255
| assertion failed on | https://api.github.com/repos/MonetDB/MonetDB/issues/176/comments | 0 | 2020-11-30T08:16:25Z | 2024-06-27T10:55:59Z | https://github.com/MonetDB/MonetDB/issues/176 | 753,260,461 | 176 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-16 09:26:18 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-16 01:09:08 +0200
## Comment 969
Date: 2004-04-16 09:26:18 +0200
From: @grobian
(pictor:~) % MapiClient --help
Usage: MapiClient [ options ]
Options are:
-b t/f | --blocked=true/false /* blocked mode */
-d | --debug=[level] /* set debug level */
-D | --dump /* dump sql database */
-e | --error /* exit on error */
-h hostname | --host=hostname /* host to connect to */
-l language | --language=lang /* {mal,sql,mil} */
-P passwd | --passwd=passwd /* password */
-p portnr | --port=portnr /* port to connect to */
-q | --quiet /* don't print welcome
message */
-s stmt | --statement=stmt /* run single statement */
-t | --trace /* trace Monet
interaction */
-T | --time /* time commands */
-u user | --user=user /* user id */
-? | --help /* show this usage
message */
So I try:
(pictor:~) fabian% MapiClient -b t
MapiClient: invalid option -- b
ehrm, yeah right, so I try again:
(pictor:~) % MapiClient -lsql --blocked=true
sql>select * from tables;
more>;
more>
more>
more>MapiClient: MapiClient.mx:486: doFileByLines:
Assertion `hdl == ((void *)0)' failed.
Abort
The blocked mode doesn't make sense to me, since it
seems to always ask for more...
Perhaps remove it from the help message?
## Comment 970
Date: 2004-04-16 13:09:08 +0200
From: @njnes
Logged In: YES
user_id=43556
added -b option
## Comment 971
Date: 2005-10-06 17:10:47 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / NOT APPLICABLE
## Comment 972
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 936135 at http://sourceforge.net/support/tracker.php?aid=936135
| MapiClient && blocked != true? | https://api.github.com/repos/MonetDB/MonetDB/issues/175/comments | 0 | 2020-11-30T08:16:22Z | 2024-06-27T10:55:58Z | https://github.com/MonetDB/MonetDB/issues/175 | 753,260,426 | 175 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-15 04:09:41 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-16 12:46:08 +0200
## Comment 965
Date: 2004-04-15 16:09:41 +0200
From: @arjan
Scenario:
sql>create table a (b integer);
sql>insert into a values(1);
sql>insert into a values(2);
sql>commit;
sql>create view c as select * from a;
sql>select count(*) from c;
count_b name
mediumint type
1 tuplecount
11 id
[ 2 ]
sql>select count(*) from c;
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select count(*) from c;
ERROR = !WARNING: interpret: convert identifier 's9' to
string.
!ERROR: interpret: no matching MIL operator to
'mark(str, oid)'.
!MAYBE YOU MEAN:
! mark(BAT[any::1,any], oid) :
BAT[any::1,oid]
! mark(BAT[any::1,any]) : BAT[any::1,oid]
The second time a query is done on a view, the above
error occurs.
## Comment 966
Date: 2004-04-16 12:46:08 +0200
From: @njnes
Logged In: YES
user_id=43556
reuse of statement tree wasn't correctly done. Tree is now
properly
reset. This bug however also showed the statement caching
was broken, this was because the scanners statement hash
key value wasn't properly reset after each parsed statement.
## Comment 967
Date: 2005-10-06 17:07:43 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=935639 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/multiple_select_on_view.SF-935639.*
## Comment 968
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 935639 at http://sourceforge.net/support/tracker.php?aid=935639
| SELECT on view only works once | https://api.github.com/repos/MonetDB/MonetDB/issues/174/comments | 0 | 2020-11-30T08:16:18Z | 2024-06-27T10:55:57Z | https://github.com/MonetDB/MonetDB/issues/174 | 753,260,398 | 174 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-15 03:34:25 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-19 02:05:48 +0200
## Comment 958
Date: 2004-04-15 15:34:25 +0200
From: @arjan
sql>select min(artist) from profile_track_history;
[ 1 ]
sql>select max(artist) from profile_track_history;
[ 1306810 ]
sql>select count(*) from profile_track_history;
[ 4516232 ]
sql>select count(*) from profile_track_history where
artist > 0;
[ 4516232 ]
sql>select sum(artist) from profile_track_history;
[ 2803793763916 ]
sql>select avg(artist) from profile_track_history;
[ -181.54080038403697 ]
This does not seem a right average, given that the rest
of the queries above give correct answers.
## Comment 959
Date: 2004-04-15 16:30:58 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
I wouldn't be surprised if this is due to overflow. The
implementation of avg is (module algebra):
proc avg(BAT[any, @1] b) : dbl {
var cnt := count(b);
if (cnt = 0)
return dbl(nil);
return dbl(sum(b))/dbl(cnt);
}
It seems to me that sum(b) can easily overflow if the BAT
type (@1 in the above) is sht or int, and then the result
will be nonsense.
## Comment 960
Date: 2004-04-15 17:44:35 +0200
From: @arjan
Logged In: YES
user_id=20087
This is indeed the case. Who and how should we fix this?
A workaround (in SQL) could be to quietly cast the column to
double first...
## Comment 961
Date: 2004-04-16 23:02:55 +0200
From: @njnes
Logged In: YES
user_id=43556
please check if current 'MonetDB' version works
## Comment 962
Date: 2004-04-19 14:05:48 +0200
From: @arjan
Logged In: YES
user_id=20087
Current version indeed gives the correct results. Thanks.
## Comment 963
Date: 2005-10-06 17:03:51 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 964
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 935617 at http://sourceforge.net/support/tracker.php?aid=935617
| Incorrect AVG aggregate? | https://api.github.com/repos/MonetDB/MonetDB/issues/173/comments | 0 | 2020-11-30T08:16:15Z | 2024-06-28T07:36:31Z | https://github.com/MonetDB/MonetDB/issues/173 | 753,260,357 | 173 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-15 03:09:59 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-16 12:48:41 +0200
## Comment 953
Date: 2004-04-15 15:09:59 +0200
From: @arjan
The output of some operators like isnull and = cannot
be directly compared to the values true and false,
since these are of the 'bool' type but the outcome of
the expressions are of type 'boolean'.
This is probably easy to fix by making 'true' and
'false' of the 'boolean' type.
sql>select isnull(1);
isnull_single_value name
boolean type
1 tuplecount
0 id
[ false ]
sql>select isnull(1)=false;
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select isnull(1)=false;
ERROR = !Types boolean(0,0) (bit) and bool(0,0) (bit)
are not equal
sql>select (1=1)=true;
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select (1=1)=true;
ERROR = !Types boolean(0,0) (bit) and bool(0,0) (bit)
are not equal
## Comment 954
Date: 2004-04-15 15:22:12 +0200
From: @arjan
Logged In: YES
user_id=20087
I already fixed it in my checkout.
Lines 1706-1710 of sql_parser.mx should read:
| BOOL_FALSE
{ sql_subtype *t = sql_bind_subtype("BOOLEAN", 0, 0 );
$$ = _newAtomNode( atom_general(t, _strdup("false"))); }
| BOOL_TRUE
{ sql_subtype *t = sql_bind_subtype("BOOLEAN", 0, 0 );
$$ = _newAtomNode( atom_general(t, _strdup("true"))); }
## Comment 955
Date: 2004-04-16 12:48:41 +0200
From: @njnes
Logged In: YES
user_id=43556
BOOL and BOOLEAN are just aliases. So a new alias list
is added which will convert the alias to the real type, ie.
incase of a table create x (bla bool) it will result in
a table create x (bla boolean). reduces the number of types
and function table.
## Comment 956
Date: 2005-10-06 17:02:05 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=935601 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/bool_boolean.SF-935601.*
## Comment 957
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 935601 at http://sourceforge.net/support/tracker.php?aid=935601
| Types boolean(0,0) (bit) and bool(0,0) (bit) are not equal | https://api.github.com/repos/MonetDB/MonetDB/issues/172/comments | 0 | 2020-11-30T08:16:12Z | 2024-06-27T10:55:55Z | https://github.com/MonetDB/MonetDB/issues/172 | 753,260,329 | 172 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-15 02:06:09 +0200
From: @arjenpdevries
To: @njnes
Version: -- development
Last updated: 2004-04-16 12:51:05 +0200
## Comment 949
Date: 2004-04-15 14:06:09 +0200
From: @arjenpdevries
zie attached.
## Comment 950
Date: 2004-04-16 12:51:05 +0200
From: @njnes
Logged In: YES
user_id=43556
missed some tail(column/type), which crashed the server.
Updated the st_stmt2string to easily detect which st_* value
isn't handled.
## Comment 951
Date: 2005-10-06 16:53:37 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=935578 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/foreign_key_in_frontend_crashes_backend.SF-935578.*
## Comment 952
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 935578 at http://sourceforge.net/support/tracker.php?aid=935578
| foreign key constraint in frontend crashes backend | https://api.github.com/repos/MonetDB/MonetDB/issues/171/comments | 0 | 2020-11-30T08:16:08Z | 2024-06-27T10:55:54Z | https://github.com/MonetDB/MonetDB/issues/171 | 753,260,296 | 171 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-15 10:10:24 +0200
From: @mlkersten
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-04-15 01:41:51 +0200
## Comment 945
Date: 2004-04-15 10:10:24 +0200
From: @mlkersten
A lengthly, buggy single MIL line sent from MapiClient
is sufficient to hang Mserver
Monet Database Server V4.3.16
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.
monet>module(sql_server);
monet>fatal flex scanner internal error--end of buffer
missed
Notice, the fatal doesn;t shut down the system either
source file in ~mk/crackers/mockup/err
## Comment 946
Date: 2004-04-15 13:41:51 +0200
From: @peterboncz
Logged In: YES
user_id=591107
was fixed in both stable and current
apparently, YY_FLUSH_OUTPUT may not be called during
parsing? we do now so afterwards
## Comment 947
Date: 2005-10-06 16:34:20 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / NOT NEEDED
The test would be simply "module(sql_server);", which is of
course included in many other tests.
## Comment 948
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 935481 at http://sourceforge.net/support/tracker.php?aid=935481
| Flex internal error | https://api.github.com/repos/MonetDB/MonetDB/issues/170/comments | 0 | 2020-11-30T08:16:05Z | 2024-06-28T07:36:30Z | https://github.com/MonetDB/MonetDB/issues/170 | 753,260,261 | 170 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-14 09:52:59 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-15 03:50:22 +0200
## Comment 941
Date: 2004-04-14 21:52:59 +0200
From: @grobian
$ java -cp ../jars/MonetJDBC.jar:. Test_Csavepoints
true true
savepoint...failed :)
false false
savepoint...passed :)
savepoint...passed :)
table 0 items passed :)
savepoint...passed :)
table 3 items, 1, 2, 3 passed :)
release...failed :( java.io.IOException: Connection to
server lost!
ABORTING TEST!!!
Mserver says (among other things) at the end:
= TODO st_sets filter not handled
= TODO st_sets filter not handled
= TODO st_sets filter not handled
Killed
$
I'm trying to release a savepoint I 'created' earlier
using RELEASE SAVEPOINT.
sql way of doin' it:
savepoint sp1;
release savepoint sp1;
## Comment 942
Date: 2005-10-06 16:32:32 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 943
Date: 2005-10-06 18:40:00 +0200
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-10-06: ALREADY IN TESTWEB
!!!!!
invalid... pffffffft...
## Comment 944
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 935162 at http://sourceforge.net/support/tracker.php?aid=935162
| SQL: RELEASE SAVEPOINT crashes server? | https://api.github.com/repos/MonetDB/MonetDB/issues/169/comments | 0 | 2020-11-30T08:16:02Z | 2024-06-28T07:36:30Z | https://github.com/MonetDB/MonetDB/issues/169 | 753,260,227 | 169 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-14 03:45:31 +0200
From: @sjoerdmullender
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2004-04-15 12:39:23 +0200
## Comment 937
Date: 2004-04-14 15:45:31 +0200
From: @sjoerdmullender
Due to a bug in the SQL server (leaking of BATs--Niels
is working on this), BBPtrim gets called. Mserver
hangs during this call while trying to obtain the
GDKcacheLock. The relevant (for this bug) part of the
stack trace is:
0 0xffffe002 in ?? ()
1 0x401cc8aa in MT_set_lock (l=0x404dab38,
name=0x40394824 "BBPtrim")
at gdk_system.mx:425
2 0x4010f6cd in BBPtrim (memtarget=199563,
vmtarget=0) at gdk_bbp.mx:1814
3 0x401370c6 in GDKmemchk (memchk=1, vmchk=1) at
gdk_utils.mx:573
4 0x40138102 in GDKvmrealloc (pold=0x42134000,
oldsize=729600,
newsize=875472, oldmax=50331648, newmax=0x4552d328)
at gdk_utils.mx:1126
5 0x40107fb5 in BBPextend (factor=1.2, buildhash=1)
at gdk_bbp.mx:373
6 0x40112bb1 in BBPgetentry () at gdk_bbp.mx:801
7 0x40109bc6 in BBPinsert (b=0xc6c9dd8) at gdk_bbp.mx:818
8 0x4016fb25 in BATcreatedesc (ht=6, tt=11,
heapnames=1) at gdk_bat.mx:112
9 0x4017014f in BATnewstorage (ht=6, tt=11, cap=8194)
at gdk_bat.mx:181
10 0x40170433 in BATnew (ht=6, tt=11, cap=8194) at
gdk_bat.mx:218
In BBPinsert, the lock is acquired (locked_by == 0, pid
!= 0), then BBPgetentry is called. This call to
BBPgetentry eventually leads to BBPtrim being called
which also tries to acquire the lock.
## Comment 938
Date: 2004-04-15 12:39:23 +0200
From: @peterboncz
Logged In: YES
user_id=591107
fixed in both stable and current
deadlock becauce BBPextend has a cachelock but causes a
BBPtrim (which needs all locks).
So, BBPextend should be protected with a BBP_notrim.
BBPinsert actually does not need BBP_notrim, once
BBPextend has protection.
## Comment 939
Date: 2005-10-06 16:31:55 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: REASSIGNED boncz
It's not very clear how to reproduce this scenario...
## Comment 940
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 934944 at http://sourceforge.net/support/tracker.php?aid=934944
| deadlock in BBPtrim | https://api.github.com/repos/MonetDB/MonetDB/issues/168/comments | 0 | 2020-11-30T08:15:59Z | 2024-06-28T07:36:30Z | https://github.com/MonetDB/MonetDB/issues/168 | 753,260,198 | 168 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-14 02:46:57 +0200
From: @menzowindhouwer
To: @njnes
Version: -- development
Last updated: 2004-04-14 09:09:05 +0200
## Comment 933
Date: 2004-04-14 14:46:57 +0200
From: @menzowindhouwer
To use the current_timestamp() and related MIL PROCs
the alarm module should be loaded, i.e. monettime needs
a MEL USE statement or an explicit module load in
monettime.mil.
Monet Database Server V4.3.16_rc06
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.
monet>module(monettime);
monet>current_timestamp().print;
!ERROR: interpret: no matching MIL operator to 'epoch()'.
!MAYBE YOU MEAN:
! epoch(timestamp) : int
!ERROR: timestamp(param 1): evaluation error.
!ERROR: print(param 1): evaluation error.
monet>module(alarm);
monet>current_timestamp().print;
[ 2004-04-14 14:46:43.000 ]
## Comment 934
Date: 2004-04-14 21:09:05 +0200
From: @njnes
Logged In: YES
user_id=43556
Fixed: moved module(alarm) call to start of mil file, not
only for monettime test script
## Comment 935
Date: 2005-10-06 16:26:04 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=934899 as
MonetDB/tests/BugDay_2005-10-06_4.9.3/Tests/monettime_needs_alarm.SF-934899.*
## Comment 936
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 934899 at http://sourceforge.net/support/tracker.php?aid=934899
| monettime needs alarm module | https://api.github.com/repos/MonetDB/MonetDB/issues/167/comments | 0 | 2020-11-30T08:15:56Z | 2024-06-27T10:55:50Z | https://github.com/MonetDB/MonetDB/issues/167 | 753,260,168 | 167 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-14 02:43:50 +0200
From: @menzowindhouwer
To: @sjoerdmullender
Version: -- development
Last updated: 2004-04-14 04:31:26 +0200
## Comment 929
Date: 2004-04-14 14:43:50 +0200
From: @menzowindhouwer
The MIL PROCs in monettime contain several times the
use of the same variable name, i.e. 'm', for different
arguments:
rule creation synonyms
proc "rule" (int m, int d, int w, int h, int m)
: rule
return rule(m, d, w, (h * 60) + m);
proc "rule" (int m, str d, int w, int h, int m)
: rule
return rule(m, daynum(d), w, (h* 60) + m);
proc "rule" (int m, str d, int w, int m) : rule
return rule(m, daynum(d), w, m);
proc "[rule]" (bat[any::1,int] m,
bat[any::1,int] d, bat[any::1,int] w,
bat[any::1,int] h, bat[any::1,int]
m) : bat[any::1,rule]
return [rule](m, d, w, [+]([*](h,60),m));
proc "[rule]" (bat[any::1,int] m,
bat[any::1,str] d, bat[any::1,int] w,
bat[any::1,int] h, bat[any::1,int]
m) : bat[any::1,rule]
return [rule](m, [daynum](d), w,
[+]([*](h,60),m));
proc "[rule]" (bat[any::1,int] m,
bat[any::1,str] d, bat[any::1,int] w,
bat[any::1,int] m) : bat[any::1,rule]
return [rule](m, [daynum](d), w, m);
## Comment 930
Date: 2004-04-14 16:31:26 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Fixe don the main branch.
## Comment 931
Date: 2005-10-06 16:08:22 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / NOT APPLICABLE
## Comment 932
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 934895 at http://sourceforge.net/support/tracker.php?aid=934895
| m means minute AND millisecond | https://api.github.com/repos/MonetDB/MonetDB/issues/166/comments | 0 | 2020-11-30T08:15:53Z | 2024-06-27T10:55:49Z | https://github.com/MonetDB/MonetDB/issues/166 | 753,260,132 | 166 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-12 11:25:33 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-04-13 11:30:00 +0200
## Comment 926
Date: 2004-04-12 11:25:33 +0200
From: @mlkersten
sql>create view m as select count(*) from tables;
sql>select * from m;
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select * from m;
ERROR = !ERROR: interpret: no matching MIL operator to
'mark(int, oid)'.
!MAYBE YOU MEAN:
! mark(BAT[any::1,any], oid) :
BAT[any::1,oid]
! mark(BAT[any::1,any]) : BAT[any::1,oid]
## Comment 927
Date: 2005-10-06 16:03:00 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=933573 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/singleton_view.SF-933573.*
## Comment 928
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 933573 at http://sourceforge.net/support/tracker.php?aid=933573
| Singleton (view) table problem | https://api.github.com/repos/MonetDB/MonetDB/issues/165/comments | 0 | 2020-11-30T08:15:50Z | 2024-06-27T10:55:48Z | https://github.com/MonetDB/MonetDB/issues/165 | 753,260,104 | 165 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-12 09:16:01 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2006-02-07 11:58:32 +0100
## Comment 917
Date: 2004-04-12 09:16:01 +0200
From: @mlkersten
The session below was sufficient to terminate the server
sql>create schema mydummy;
sql>drop schema mydummy;
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = drop schema mydummy;
ERROR = Connection terminated
## Comment 918
Date: 2004-04-12 11:22:39 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Another one:
sql>create view m as select * from tables with check option;
sql>select * from m;
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = select * from m;
ERROR = Connection terminated
## Comment 919
Date: 2005-10-06 15:43:32 +0200
From: @swingbit
Logged In: YES
user_id=883451
*** This bug reports contains 2 different bugs, but only the
first one was addressed. Therefore, the second bug is
re-opened as a new bug
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=933546 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/create_drop_schema.SF-931619.*
## Comment 920
Date: 2006-01-15 14:12:46 +0100
From: @drstmane
Logged In: YES
user_id=572415
re-opened, as the respective test (still?) failes with
"!FATAL: Your disk seems to be full, exiting ..."
at least on
http://monetdb.cwi.nl/testing/projects/monetdb/Current/sql/.mTests103/GNU.64.32.d-Fedora4/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Current/sql/.mTests103/GNU.64.64.d-Fedora4/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Current/sql/.mTests103/GNU.32.32.d-Gentoo1.6.13/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Current/sql/.mTests103/GNU.32.32.d-SuSE9.3/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Current/sql/.mTests103/GNU.32.32.d-SuSE9ES/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
Further, also http://sourceforge.net/support/tracker.php?aid=1406591 "several tests cause segfault on the same subset of
platforms"
https://sourceforge.net/tracker/index.php?func=detail&aid=1406591&group_id=56967&atid=482468
might be related...
## Comment 921
Date: 2006-01-22 21:07:45 +0100
From: @mlkersten
Logged In: YES
user_id=490798
After the recent changes to SQL catalogue manager this test
works correctly.
## Comment 922
Date: 2006-01-22 21:15:33 +0100
From: @drstmane
Logged In: YES
user_id=572415
????????
were the changes after 13:13 today the fix these crashes:
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.64.64.d-Fedora4/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.64.64.d-Debian3.1/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.64.32.d-Fedora4/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.32.32.d-Gentoo1.6.14/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.32.32.d-SuSE9.3/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.64.64.d-SuSE9.3/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
http://monetdb.cwi.nl/testing/projects/monetdb/Stable/sql/.mTests103/GNU.32.32.d-SuSE9ES/src_test_BugDay_2005-10-06_2.9.3/create_drop_schema.SF-933546.err.00.html
????????
## Comment 923
Date: 2006-01-22 21:19:33 +0100
From: @mlkersten
Logged In: YES
user_id=490798
You are right. Should have waited for the testweb at large.
To eager to get rid of bugs.
## Comment 924
Date: 2006-01-22 21:23:14 +0100
From: @drstmane
Logged In: YES
user_id=572415
re-opened, and rephrasing my question:
were there changes after 13:13 today that *are supposed to*
fix these crashes?
## Comment 925
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 933546 at http://sourceforge.net/support/tracker.php?aid=933546
| drop schema crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/164/comments | 0 | 2020-11-30T08:15:47Z | 2024-06-27T10:55:47Z | https://github.com/MonetDB/MonetDB/issues/164 | 753,260,072 | 164 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-11 08:30:15 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-04-11 09:02:44 +0200
## Comment 911
Date: 2004-04-11 20:30:15 +0200
From: @mlkersten
Type error should have been catched by the parser
sql>set autocommit=true;
sql>create table tmp(i int);
sql>insert into tmp values(1);
sql>insert into tmp values(null);
sql>select i in (1) from tmp;
i name
int type
1 tuplecount
0 id
[ 1 ]
sql>select i in (1,'a') from tmp;
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select i in (1,'a') from tmp;
ERROR = !ERROR: cast2lng: operation failed.
## Comment 912
Date: 2004-04-11 21:02:44 +0200
From: @njnes
Logged In: YES
user_id=43556
Its correct to mix int and string in a single set as
they are casted to int correctly. Problem is the casts are
dynamic as
the same mil proc can be used with an other set of values
(same type).
## Comment 913
Date: 2004-04-11 21:08:47 +0200
From: @njnes
Logged In: YES
user_id=43556
Its correct to mix int and string in a single set as
they are casted to int correctly. Problem is the casts are
dynamic as
the same mil proc can be used with an other set of values
(same type).
## Comment 914
Date: 2004-04-13 08:58:17 +0200
From: @njnes
Logged In: YES
user_id=43556
Its correct to mix int and string in a single set as
they are casted to int correctly. Problem is the casts are
dynamic as
the same mil proc can be used with an other set of values
(same type).
## Comment 915
Date: 2005-10-06 14:52:25 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 916
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 933295 at http://sourceforge.net/support/tracker.php?aid=933295
| Incomplete semantic checks | https://api.github.com/repos/MonetDB/MonetDB/issues/163/comments | 0 | 2020-11-30T08:15:43Z | 2024-06-28T07:36:29Z | https://github.com/MonetDB/MonetDB/issues/163 | 753,260,040 | 163 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-11 06:18:22 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2010-05-04 11:17:30 +0200
## Comment 906
Date: 2004-04-11 18:18:22 +0200
From: @mlkersten
Script below illustrates an experiment with the IN
predicate. A side-effect is that the table created
disappeers
sql>create table tmp(i int);
sql>insert into tmp values(1);
sql>insert into tmp values(null);
sql>select * from tmp where i in (1);
i name
int type
1 tuplecount
1 id
[ 1 ]
sql>select i in (1) from tmp;
i name
int type
1 tuplecount
2 id
[ 1 ]
sql>select i in (null,1) from tmp;
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select i in (null,1) from tmp;
ERROR = !ERROR parse error at token (341) in statement:
select i in (null
sql>rollback;
sql>select * from tmp;
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select * from tmp;
ERROR = !Unknown table tmp
sql>
## Comment 907
Date: 2004-04-11 19:35:13 +0200
From: @grobian
Logged In: YES
user_id=963970
The table disappears because autocommit is not on, and a
rollback is given. Thus the table is gone as well.
If a commit was given after the inserts, the table would
have persisted.
## Comment 908
Date: 2004-04-11 19:43:08 +0200
From: @grobian
Logged In: YES
user_id=963970
The real error right here is that the parser complains about
the query. It's resultset is not correct as well for the in
(1) query.
The IN in the select clause represents an evaluation which
returns either true or false. (boolean type) In the first
SELECT i IN case this boolean is not returned, on the second
case (with a list) the parser just fails.
Someone who can should invalidate this bug and open a new
one which issues the parser complaint and result behaviour.
## Comment 909
Date: 2005-10-06 14:51:50 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 910
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 933248 at http://sourceforge.net/support/tracker.php?aid=933248
Bug has invalid status, setting status to "NEW".
Previous status was "DELETED".
| Disappearing SQL table after IN test | https://api.github.com/repos/MonetDB/MonetDB/issues/162/comments | 0 | 2020-11-30T08:15:40Z | 2024-06-28T07:36:29Z | https://github.com/MonetDB/MonetDB/issues/162 | 753,260,006 | 162 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-11 03:26:55 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-11 09:23:01 +0200
## Comment 900
Date: 2004-04-11 15:26:55 +0200
From: @grobian
When starting the server with emptied database:
[orion:~/local/program/_GNU_32_d__] monet% bin/Mserver
!WARNING: GDKlockHome: created directory
/Users/monet/local/program/_GNU_32_d__/var/MonetDB/dbfarm/demo/
!WARNING: GDKlockHome: ignoring empty or invalid .gdk_lock.
!WARNING: BBPdir: initializing BBP.
Monet Database Server V4.3.16
Copyright (c) 1993-2004, CWI. All rights reserved.
Compiled for powerpc-apple-darwin6.8/32bit;
dynamically linked.
Visit http://monetdb.cwi.nl for further information.
monet>module(sql_server);
monet>= TODO: sql_rel2bin: shrink_select_ranges: handle
st_reverse(st_diff)!
= TODO: sql_rel2bin: shrink_select_ranges: handle
st_reverse(st_diff)!
Then a small test with and without nulls on that database:
monetdb-> create table test (
monetdb-( id1 int,
monetdb-( id2 int not null,
monetdb-( id3 int null
monetdb-( );
affected rows
-------------
-1
results in:
monetdb-> \d test
CREATE TABLE sys.test (
id1 int(19) NOT NULL,
id2 int(19) NOT NULL,
id3 int(19) NOT NULL
);
monetdb->
while on Linux the same results in what one would
expect (null, not null, null)
?!?!?
I'm completely lost here, how this can happen.
## Comment 901
Date: 2004-04-11 15:33:35 +0200
From: @grobian
Logged In: YES
user_id=963970
I just found out that rc_05 suffers the same problem...
## Comment 902
Date: 2004-04-11 21:21:54 +0200
From: @grobian
Logged In: YES
user_id=963970
The just checked in fix by Niels fixed this problem.
confirming; works now -> bug resolved/fixed
## Comment 903
Date: 2005-10-06 14:50:58 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: REASSIGNED metldown
MAC OS specific bug
## Comment 904
Date: 2005-10-06 15:07:12 +0200
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-10-06: CLAIMED BY meltdown
BugDay_2005-10-06: TEST ADDED / SUCCESS
src/test/BugDay_2005-10-06_2.9.3/Tests/not_null.SF-933194*
we have a darwin test machine, so I just added the test
## Comment 905
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 933194 at http://sourceforge.net/support/tracker.php?aid=933194
| SQL: each column NOT NULL on Mac OS X 10.2.8? | https://api.github.com/repos/MonetDB/MonetDB/issues/161/comments | 0 | 2020-11-30T08:15:36Z | 2024-06-27T10:55:44Z | https://github.com/MonetDB/MonetDB/issues/161 | 753,259,965 | 161 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-09 06:49:30 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-13 11:29:28 +0200
## Comment 896
Date: 2004-04-09 18:49:30 +0200
From: @grobian
It seems the server doesn't care much about the
savepoint name, it doesn't seem to require the
savepoint to be set.
The shortest and most descriptive example I can come up
with it simply this:
ROLLBACK TO SAVEPOINT blablahfdsjkheiorjed;
or
RELEASE SAVEPOINT scared_yet;
I would expect an error to be thrown, as that would
make life in a JDBC driver much more easier (full
dependancy on the server for savepoints management)
As desert, a Melt-Down scenario: (guess I have reached
my bugs quota for now)
start transaction;
create table savepointtest (
id int,
primary key(id)
);
select * from savepointtest;
savepoint name1;
insert into savepointtest values(1);
insert into savepointtest values(2);
insert into savepointtest values(3);
select * from savepointtest;
savepoint name2;
insert into savepointtest values(4);
insert into savepointtest values(5);
insert into savepointtest values(6);
select * from savepointtest;
insert into savepointtest values(7);
insert into savepointtest values(8);
insert into savepointtest values(9);
savepoint name3;
release savepoint name1;
rollback to savepoint name1;
commit;
it crashes on commit, while I would expect an error on
the rollback for name1.
## Comment 897
Date: 2004-04-13 23:29:28 +0200
From: @njnes
Logged In: YES
user_id=43556
name checks improved
## Comment 898
Date: 2005-10-06 14:47:23 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: ALREADY IN TESTWEB
src/test/savepoints.*
src/test/savepoints2.*
## Comment 899
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 932388 at http://sourceforge.net/support/tracker.php?aid=932388
| SQL: ROLLBACK and RELEASE don't care about the savepoint?!? | https://api.github.com/repos/MonetDB/MonetDB/issues/160/comments | 0 | 2020-11-30T08:15:33Z | 2024-06-27T10:55:43Z | https://github.com/MonetDB/MonetDB/issues/160 | 753,259,930 | 160 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-09 05:27:28 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-13 11:28:48 +0200
## Comment 890
Date: 2004-04-09 17:27:28 +0200
From: @grobian
The parser currently accepts the statement
SET reply_size = true;
and returns 1 row
SET reply_size = false;
seems to return everything
Apart from the fact that true and false are not clearly
defined as integers this is syntactically incorrect,
since the size should require an integer.
Actually I feel this set command should not have
existed, for it can completely mess up my JDBC client :(
SET auto_commit = 60;
is recognized as a completely legal statement, but what
does it do?!?
Should I assume it is 'true' here? Or does it only
auto_commit 60 statements?
## Comment 891
Date: 2004-04-09 18:12:06 +0200
From: @grobian
Logged In: YES
user_id=963970
with autocommit instead of auto_commit it seems 60 means
autocommit is off.
Why is it autocommit and not auto_commit or replysize?!?
## Comment 892
Date: 2004-04-09 18:16:48 +0200
From: @grobian
Logged In: YES
user_id=963970
Additionally it seems that SET autocommit is only recognized
when using 'S' as language prefix, not when using 's'. Could
this behaviour be applied to reply_size (and other set
commands) as well, please? I would be very pleased with that.
## Comment 893
Date: 2004-04-13 23:28:48 +0200
From: @njnes
Logged In: YES
user_id=43556
new table session holds these variables, types are now checked
## Comment 894
Date: 2005-10-06 14:36:56 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=932360 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/set_types.SF-931619.*
## Comment 895
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 932360 at http://sourceforge.net/support/tracker.php?aid=932360
| SET reply_size = true is allowed, SET auto_commit = 20 ? | https://api.github.com/repos/MonetDB/MonetDB/issues/159/comments | 0 | 2020-11-30T08:15:30Z | 2024-06-27T10:55:42Z | https://github.com/MonetDB/MonetDB/issues/159 | 753,259,904 | 159 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-08 02:55:16 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2004-04-08 03:21:17 +0200
## Comment 883
Date: 2004-04-08 14:55:16 +0200
From: @arjan
The use of a column alias (x as y) does not work in a
HAVING clause.
select item.title, avg(rating) as a, count(item.title)
AS B from rating inner join item using (item_id) group
by item.title HAVING B >10 order by a desc limit 10;
Whereas the following equivalent query does work:
select item.title, avg(rating) as a, count(item.title)
as b from rating inner join item using (item_id) group
by item.title having count(item.title)>10 order by a
desc limit 10;
## Comment 884
Date: 2004-04-08 15:03:54 +0200
From: @arjan
Logged In: YES
user_id=20087
by the way, the error is the following
ERROR = !Identifier: b unknown or ambigious
## Comment 885
Date: 2004-04-08 15:04:44 +0200
From: @grobian
Logged In: YES
user_id=963970
I think this is correct behaviour...
order is a post operator and only knows the final columns
(therefore order by on columns not in select is not allowed)
Having on the other hand works on the grouping, which
(someone correct me if I'm wrong here) is some sort of inner
query which doesn't know your alias of your select.
## Comment 886
Date: 2004-04-08 15:21:17 +0200
From: @njnes
Logged In: YES
user_id=43556
The alias is not jet defined (selection fase is after the
having fase) ie the
alias is not known jet! So not a bug ;-)
## Comment 887
Date: 2004-04-08 15:54:49 +0200
From: @njnes
Logged In: YES
user_id=43556
The alias is not jet defined (selection fase is after the
having fase) ie the
alias is not known jet! So not a bug ;-)
## Comment 888
Date: 2005-10-06 13:59:29 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 889
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 931723 at http://sourceforge.net/support/tracker.php?aid=931723
| Use of column alias in HAVING clause | https://api.github.com/repos/MonetDB/MonetDB/issues/158/comments | 0 | 2020-11-30T08:15:28Z | 2024-06-28T07:36:29Z | https://github.com/MonetDB/MonetDB/issues/158 | 753,259,865 | 158 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-08 11:22:40 +0200
From: @arjan
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2004-04-08 11:37:47 +0200
## Comment 876
Date: 2004-04-08 11:22:40 +0200
From: @arjan
When in a .sql file, the last line consists of a
comment, this results in a server crash. The string
gets sent to the server, which should not be the case,
i guess.
Test scenario:
echo "-- comment" | MapiClient -l sql -u monetdb -P monetdb
results in :
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = -- comment
ERROR = Connection terminated
and on the server :
monet>Mserver: sql_scan.mx:464:
scanner_query_processed: Assertion `s->rs->pos <=
s->rs->len' failed.
Aborted
It just should ignore the input.
## Comment 877
Date: 2004-04-08 11:26:32 +0200
From: @grobian
Logged In: YES
user_id=963970
this bug should have been fixed yesterday.
Did you update and recompile?
## Comment 878
Date: 2004-04-08 11:34:38 +0200
From: @drstmane
Logged In: YES
user_id=572415
does indeed crash with _rc06,
but seems to work fine with the latest MonetDB_4-3-16/SQL_2-0-16
## Comment 879
Date: 2004-04-08 11:37:47 +0200
From: @njnes
Logged In: YES
user_id=43556
was fixed by a change to MonetDB stream library
## Comment 880
Date: 2004-04-08 15:19:40 +0200
From: @njnes
Logged In: YES
user_id=43556
was fixed by a change to MonetDB stream library
## Comment 881
Date: 2005-10-06 13:54:09 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=931619 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/comment_only.SF-931619.*
## Comment 882
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 931619 at http://sourceforge.net/support/tracker.php?aid=931619
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.
This bug was marked DUPLICATE in the database it was moved from.
Changing resolution to "MOVED"
| Comment-only statement crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/157/comments | 0 | 2020-11-30T08:15:25Z | 2024-06-28T07:36:28Z | https://github.com/MonetDB/MonetDB/issues/157 | 753,259,842 | 157 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-08 10:11:25 +0200
From: @drstmane
To: @njnes
Version: -- development
Last updated: 2004-04-08 11:36:15 +0200
## Comment 870
Date: 2004-04-08 10:11:25 +0200
From: @drstmane
I'm not quite sure, whether this is correct SQL, but at
least the server should not crash with an assertion:
========
10:07:50 manegold@draco:/tmp/mallocbugscript $
MapiClient -lsql
sql>CREATE TABLE t1 ( a1 INTEGER DEFAULT 0 );
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = CREATE TABLE t1 ( a1 INTEGER DEFAULT 0 );
ERROR = Connection terminated
10:09:37 manegold@draco:/tmp/mallocbugscript $
========
10:07:50 manegold@draco:~/Monet/4.3/sql $ Mserver --set
monet_mod_path=$MONET_PREFIX/lib/MonetDB:$SQL_PREFIX/lib/MonetDB
Monet Database Server V4.3.16_rc06
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.
monet>module(sql_server);
monet>Mserver:
/ufs/manegold/Monet/4.3/sql/src/server/sql_schema.mx:883:
column_option: Assertion `0' failed.
Aborted (core dumped)
10:09:37 manegold@draco:~/Monet/4.3/sql $
========
## Comment 871
Date: 2004-04-08 11:24:44 +0200
From: @arjan
Logged In: YES
user_id=20087
This does work for me (using both server / sql rc06)
## Comment 872
Date: 2004-04-08 11:32:10 +0200
From: @drstmane
Logged In: YES
user_id=572415
Indeed, it does work with MonetDB_4-3-16_rc06/SQL_2-0-6_rc06,
but it crashes with the latest MonetDB_4-3-16/SQL_2-0-6 ...
## Comment 873
Date: 2004-04-08 11:36:15 +0200
From: @njnes
Logged In: YES
user_id=43556
Added missing code for defaults other then string
## Comment 874
Date: 2005-10-06 13:46:40 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=931573 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/create_table_default.SF-931573.*
## Comment 875
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 931573 at http://sourceforge.net/support/tracker.php?aid=931573
| CREATE TABLE t1 ( a1 INTEGER DEFAULT 0 ); crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/156/comments | 0 | 2020-11-30T08:15:22Z | 2024-06-27T10:55:39Z | https://github.com/MonetDB/MonetDB/issues/156 | 753,259,801 | 156 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-07 09:38:25 +0200
From: @grobian
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2004-04-07 08:13:04 +0200
## Comment 867
Date: 2004-04-07 09:38:25 +0200
From: @grobian
monetdb-> \d maps
CREATE TABLE sys.maps (
id int(19) NOT NULL,
parent int(19),
world int(19) NOT NULL,
imagefile varchar(255) NOT NULL,
top_left_x real(51),
top_left_y real(51),
bot_right_x real(51),
bot_right_y real(51),
PRIMARY KEY (id), -- c1012117
FOREIGN KEY (parent) REFERENCES sys.maps(id),
-- c1012388
FOREIGN KEY (world) REFERENCES sys.worlds(id),
-- c1012118
);
Notice the foreign key relation c1012388, which
references the maps table itself on a different column.
Assume in the table world exists an entry with id 1 as in:
monetdb-> insert into worlds values (1, 'earth', 'Our
world; Earth');
affected rows
-------------
-1
Then when inserting values into the database:
monetdb-> insert into maps values (1, 0, 1,
'world-map-1600-1700.jpg', 0, 0, 0, 0);
affected rows
-------------
-1
monetdb->
monetdb-> insert into maps values (2, 10, 1,
'world-map-1600-1700.jpg', 0, 0, 0, 0);
affected rows
-------------
-1
The foreign key contraint on the table itself doesn't
seem to be checked at all!!!
(no the records with those id's are not sneakily
somewhere in the database:)
monetdb-> select * from maps;
+----------
| id parent world imagefile top_left_x
top_left_y bot_right_x bot_right_y
+----------
| 1 0 1 world-map-1600-1700.jpg 0
0 0 0
| 2 10 1 world-map-1600-1700.jpg 0
0 0 0
+----------
2 rows
monetdb->
## Comment 868
Date: 2005-10-06 13:38:53 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / INVALID
## Comment 869
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 930906 at http://sourceforge.net/support/tracker.php?aid=930906
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: Foreign key constraints which reference the table itsel | https://api.github.com/repos/MonetDB/MonetDB/issues/155/comments | 0 | 2020-11-30T08:15:18Z | 2024-06-28T07:36:28Z | https://github.com/MonetDB/MonetDB/issues/155 | 753,259,762 | 155 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-06 01:17:01 +0200
From: @arjan
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-04-08 10:14:13 +0200
## Comment 860
Date: 2004-04-06 13:17:01 +0200
From: @arjan
When performing a COPY INTO a table for a mere 5000
records, the database crashes with above message.
Attached is an archive with load script and data file.
Note, this seems data-dependent; COPY INTO's on other
data are working like they should.
## Comment 861
Date: 2004-04-07 17:11:32 +0200
From: @arjan
Logged In: YES
user_id=20087
This same (apparently) bug appears when using the data file
+ load script in plain MIL, see the attached file.
## Comment 862
Date: 2004-04-07 22:19:41 +0200
From: @drstmane
Logged In: YES
user_id=572415
fixed ascii_io to properly handle non-quoted space-only strings:
it now inserts and empty string ("")
instead of calling GDKmalloc with a negative value,
which of course resulted in a crash.
this fixes the MIL variant of BUG #930364 "GDKvmalloc failed
using COPY INTO", and most probably also the SQL variant;
yes, Arjan, you had a space-only value in your data file ---
did cost me some time to detect this ... |-(
Arjan, could you please test, whether this indeed fixes also
the SQL version?
Added some test scripts for space handling of asii_io.
OPEN:
Sould ascii_io really stop silently on an empty line with a
single-column load ?
## Comment 863
Date: 2004-04-08 10:14:13 +0200
From: @drstmane
Logged In: YES
user_id=572415
I just checked it, and it works fine, now --- after removing
the "DEFAULT 0" from the CREATE TABLE statement, which
crashes the server (cf. bug report 931573).
## Comment 864
Date: 2004-04-08 10:50:52 +0200
From: @drstmane
Logged In: YES
user_id=572415
I just checked it, and it works fine, now --- after removing
the "DEFAULT 0" from the CREATE TABLE statement, which
crashes the server (cf. bug report 931573).
## Comment 865
Date: 2005-10-06 13:33:48 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: REASSIGN stmane
Stefan wrote in the comments the he added some tests, but I
wasn't able to find them in the repository.
## Comment 866
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 930364 at http://sourceforge.net/support/tracker.php?aid=930364
| GDKvmalloc failed using COPY INTO | https://api.github.com/repos/MonetDB/MonetDB/issues/154/comments | 0 | 2020-11-30T08:15:15Z | 2024-06-27T10:55:37Z | https://github.com/MonetDB/MonetDB/issues/154 | 753,259,727 | 154 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-04 07:46:42 +0200
From: @mlkersten
To: @njnes
Version: -- development
Duplicates: #152
Last updated: 2010-05-04 11:17:01 +0200
## Comment 857
Date: 2004-04-04 19:46:42 +0200
From: @mlkersten
sql>create table t(i int);
sql>create index myindex on t(i);
sql>drop index myindex;
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = drop index myindex;
ERROR = Connection terminated
says it all. The server just stopped without message
## Comment 858
Date: 2005-12-20 01:11:33 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay: stmane: NO TEST / DUPLICATE
## Comment 859
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 929300 at http://sourceforge.net/support/tracker.php?aid=929300
Bug has invalid status, setting status to "NEW".
Previous status was "DELETED".
## Comment 13984
Date: 2010-05-04 11:17:01 +0200
From: @grobian
*** This bug has been marked as a duplicate of bug #152 ***
| Drop indices crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/153/comments | 0 | 2020-11-30T08:15:12Z | 2024-06-28T07:36:28Z | https://github.com/MonetDB/MonetDB/issues/153 | 753,259,699 | 153 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-04 05:01:41 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2010-05-04 11:17:01 +0200
## Comment 852
Date: 2004-04-04 17:01:41 +0200
From: @mlkersten
sql>create table t(i int);
sql>create index myindex on t(i);
sql>drop index myindex;
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = drop index myindex;
ERROR = Connection terminated
says it all. The server just stopped without message
## Comment 853
Date: 2004-04-05 11:30:16 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
It seems that in sql_list.mx:list_remove_node, the node n is
not a member of list l which causes p to become NULL, and
since p != n, the second if in the body takes the else
branch which redirects through p (which is NULL) causing a
segmentation violation.
I don't know whether it is an error that n is not in l, or
whether that is legitimate, so I don't know where to fix things.
list_remove_node is called from cs_del, which is calle from
mvc_drop_idx.
## Comment 854
Date: 2004-04-07 20:14:02 +0200
From: @njnes
Logged In: YES
user_id=43556
fixed bug in removing the index node from the two lists it
is used in.
## Comment 855
Date: 2005-10-06 13:10:52 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: TEST ADDED / SUCCESS
Test added for http://sourceforge.net/support/tracker.php?aid=929243 as
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/drop_index.SF-929243.*
## Comment 856
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 929243 at http://sourceforge.net/support/tracker.php?aid=929243
## Comment 13985
Date: 2010-05-04 11:17:01 +0200
From: @grobian
*** Bug #153 has been marked as a duplicate of this bug. ***
| Drop indices crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/152/comments | 0 | 2020-11-30T08:15:09Z | 2024-06-27T10:55:35Z | https://github.com/MonetDB/MonetDB/issues/152 | 753,259,669 | 152 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-04 11:00:57 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2010-05-04 11:16:20 +0200
## Comment 848
Date: 2004-04-04 11:00:57 +0200
From: @mlkersten
The users table in the SQL catalog lacks the
attribute 'id', which makes it impossible to join
the user role's relation with the user table
## Comment 849
Date: 2004-04-04 11:50:53 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Link works through the authorization table
## Comment 850
Date: 2005-10-06 12:48:26 +0200
From: @swingbit
Logged In: YES
user_id=883451
BugDay_2005-10-06: CLAIMED BY cornuz
BugDay_2005-10-06: NO TEST / DELETED
## Comment 851
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 929158 at http://sourceforge.net/support/tracker.php?aid=929158
Bug has invalid status, setting status to "NEW".
Previous status was "DELETED".
| Link users <-> user_role | https://api.github.com/repos/MonetDB/MonetDB/issues/151/comments | 0 | 2020-11-30T08:15:07Z | 2024-06-28T07:36:28Z | https://github.com/MonetDB/MonetDB/issues/151 | 753,259,637 | 151 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-04 10:31:57 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-04-07 08:14:54 +0200
## Comment 844
Date: 2004-04-04 10:31:57 +0200
From: @mlkersten
create table t(i int);
insert into t values(0);
insert into t values(1);
insert into t values(2);
insert into t values(3);
insert into t values(4);
select case i
when 0 then 'base table'
when 1 then 'system table'
when 2 then 'view'
when 3 then 'session temporary table'
when 4 then 'temporary table'
from t;
Such expressions are needed to create a user friendly
output while displaying the SQL catalog information
(e.g. convert table type back into something
understandable)
ps. the error token number should be converted back
into its string equivalent to improve localization of
the error in the original query text
## Comment 845
Date: 2004-04-07 20:14:54 +0200
From: @njnes
Logged In: YES
user_id=43556
case is supported, need to add an end at the end of the case!
## Comment 846
Date: 2005-10-06 23:53:20 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/case.SF-929151.*
## Comment 847
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 929151 at http://sourceforge.net/support/tracker.php?aid=929151
| CASE operator missing | https://api.github.com/repos/MonetDB/MonetDB/issues/150/comments | 0 | 2020-11-30T08:15:04Z | 2024-06-28T07:36:27Z | https://github.com/MonetDB/MonetDB/issues/150 | 753,259,594 | 150 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-03 03:55:51 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-04-07 08:15:30 +0200
## Comment 840
Date: 2004-04-03 15:55:51 +0200
From: @mlkersten
The following statements are valid in MySQL and appears
to be valid SQL'99 as well.
CREATE TABLE services (
sid tinyint(4) NOT NULL default '0',
sname varchar(255) NOT NULL default '',
sfee float(6,2) NOT NULL default '0.00',
PRIMARY KEY (sid)
);
INSERT INTO services (sid, sname, sfee) VALUES (1,
'Accounting', '1500.00');
MonetDB/SQL complaints as follows:
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = INSERT INTO services (sid, sname, sfee) VALUES
(1, 'Accounting', '1500.
0');
ERROR = !Types char(7,0) (str) and float(6,2) (flt) are
not equal
But the string is a legal numeric value and fits the
column type
## Comment 841
Date: 2004-04-07 20:15:30 +0200
From: @njnes
Logged In: YES
user_id=43556
Added special conversion for string 2 number(d,s).
## Comment 842
Date: 2005-10-06 23:49:13 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/string2float_coercions.SF-928850.*
## Comment 843
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 928850 at http://sourceforge.net/support/tracker.php?aid=928850
| Handling string-> float coercions | https://api.github.com/repos/MonetDB/MonetDB/issues/149/comments | 0 | 2020-11-30T08:15:00Z | 2024-06-27T10:55:32Z | https://github.com/MonetDB/MonetDB/issues/149 | 753,259,549 | 149 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-02 03:52:50 +0200
From: @MarcinZukowski
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-04-04 08:22:55 +0200
## Comment 834
Date: 2004-04-02 15:52:50 +0200
From: @MarcinZukowski
The following code doesn't produce expected result:
x:=new(int,void);
x.insert(1,nil);
x.insert(1,nil);
x.insert(2,nil);
x.insert(3,nil);
x.insert(3,nil);
x:=x.mark(0@0);
x.print;
{min}(x).print;
shold be:
[ 1, 0@0 ]
[ 2, 2@0 ]
[ 3, 3@0 ]
and is:
[ 1, nil ]
[ 2, nil ]
[ 3, nil ]
Attached script that shows this.
## Comment 835
Date: 2004-04-03 14:55:38 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Seems to belong to the aggr* module group, which by itself
may require a 'reshuffling' of functionality
## Comment 836
Date: 2004-04-03 15:08:10 +0200
From: @drstmane
Logged In: YES
user_id=572415
Well, the original bug report does not say anything about
modules, hence, I assume it refers to the standard min
function (resp. its multiplexed version)
as defined in modules algebra ... right?
Anyway, I'll have a look ... ;-)
(actually, I already checked that the standard {min}
(indeed) does not work on void, while it does on oid ...)
## Comment 837
Date: 2004-04-04 20:22:54 +0200
From: @drstmane
Logged In: YES
user_id=572415
Fixed handling of void-tailed BATs in pump (i.e.,
interpret_setaggr):
If we do "clever" (dirty?) tricks with "fake-views" on
sorted BATs, we should take proper care of void columns,
i.e., set the proper seqbase for the "fake-views" ...
Apparently, Marcin was the first to use a grouped-aggregate
on head-sorted void-tailed BATs since this code has been
introduced (actually "optimized") on August 10 1998(!)...
## Comment 838
Date: 2005-10-06 23:34:48 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: NO TEST / REDUNDANT
test script exists as
/home/manegold/Monet/Testing/Current/MonetDB/tests/BugsViaSourgeforce/Tests/ID.928253.*
## Comment 839
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 928253 at http://sourceforge.net/support/tracker.php?aid=928253
| aggregates do not work on void | https://api.github.com/repos/MonetDB/MonetDB/issues/148/comments | 0 | 2020-11-30T08:14:58Z | 2024-06-27T10:55:31Z | https://github.com/MonetDB/MonetDB/issues/148 | 753,259,524 | 148 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-01 09:10:46 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-07 08:16:06 +0200
## Comment 828
Date: 2004-04-01 21:10:46 +0200
From: @grobian
monetdb-> \d idxs
CREATE SYSTEM TABLE sys.idxs (
id mediumint(9),
table_id mediumint(9),
type mediumint(9),
name mediumint(9)
);
monetdb-> select * from idxs;
+----------
| id table_id type name
+----------
| 1014477 1005783 0 c1014463
| 1020838 1020777 0 c1020764
| 1021147 1021075 0 c1021061
| 1021175 1021075 1 c1021062
| 1021346 1021075 1 c1021332
| 1021587 1021525 0 c1021511
| 1021615 1021525 1 c1021512
| 1021870 1021828 0 c1021814
| 1021898 1021828 0 c1021815
| 1022151 1022104 0 c1022089
| 1022183 1022104 1 c1022090
| 1022189 1022104 1 c1022091
| 1022442 1022400 1 c1022387
| 1059822 1005783 0 test_var_idx
+----------
14 rows
monetdb->
Although the type of column name is mediumint, it
produces a varchar, making it difficult to show the
data for a law abiding JDBC driver in a generic
environment.
Why is this column typed mediumint? Shouldn't it be
more logical if it were varchar?
## Comment 829
Date: 2004-04-02 10:52:51 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
I assume that in sql_mvc.mx, the call to
bootstrap_create_column on behalf of the "name" column of
the "idxs" table should have a 4th argument "VARCHAR"
instead of "MEDIUMINT". I'll leave it to Niels to decide
whether this is a correct analysis.
You do need to recreate your databases after this change.
## Comment 830
Date: 2004-04-07 20:16:06 +0200
From: @njnes
Logged In: YES
user_id=43556
cut/paste bug
## Comment 831
Date: 2005-10-06 23:55:53 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: NO TEST / REASSIGNED to mr-meltdown
Fabian,
adding a simple "select * from idxs;" doesn't seem
reasonable to me.
Would you have a better idea?
Otherwise, we should just say, that there is no reasonable
test script...
## Comment 832
Date: 2005-10-07 09:36:39 +0200
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-10-06: NO TEST / IMPLICIT
this is implicitly tested by the JDBC tests, because they
bark when types don't match with output.
## Comment 833
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 927743 at http://sourceforge.net/support/tracker.php?aid=927743
| SQL: column types from system tables weird | https://api.github.com/repos/MonetDB/MonetDB/issues/147/comments | 0 | 2020-11-30T08:14:54Z | 2024-06-27T10:55:30Z | https://github.com/MonetDB/MonetDB/issues/147 | 753,259,478 | 147 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-01 03:43:52 +0200
From: @arjan
To: @njnes
Version: -- development
Last updated: 2006-12-04 08:38:10 +0100
## Comment 816
Date: 2004-04-01 15:43:52 +0200
From: @arjan
When loading a 100+ column tab-separated file with MIL,
the server crashes with a GDKvmalloc failed message.
The following archive contains the data file / format
file + load script.
http://homepages.cwi.nl/~scherpen/columnbug.tar.bz2
BTW: This is using rc06.
## Comment 817
Date: 2004-04-02 11:18:43 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
You are trying to create 100 BATs (100 columns, 1 BAT per
column) of 10e6 rows each (the MAXDATA in your test script).
That's 10e9 values (assuming they're all void BATs) of
probably at least 4 bytes per value, giving a grand total of
4*10e9 bytes, i.e. 4 GB. That's the size of your address
space, leaving no space for the actual program or any other
data.
No wonder the server complains.
However, it should probably not give a fatal error,
therefore assigning to Peter.
## Comment 818
Date: 2004-04-02 15:44:14 +0200
From: @peterboncz
Logged In: YES
user_id=591107
Given the fact that proper BAT descriptor checking in many
routines is missing, it is practically impossible to generate an
error message and continue (the NULL BAT will cause a
crash).
this is a known but grave engineering error in MonetDB that
is very hard to correct.
the bug will not occur on a 64-bits platform (such as
spin.cwi.nl or titan.ins.cwi.nl).
the other issue is that load_data should actually not crash. In
many sensible uses of MonetDB, indeed very large tables are
read in. MonetDB can still query these tables, as long as not
all columns are accessed at the same time from a single MIL
command.
For instance, the datadistilleries verion of bulk-load would
load chunks of 100K tuples into temporary 100K BATs for
each column. Then, it would append these columns to
memory mapped result bats, one at-a-time. That works
pretty well (rather effiecient), and is highly scalable.
We could consider adding this functionality to ascii_io.
one must realise that these problems become gradually less
important as more people switch to a 64-bits platform.
## Comment 819
Date: 2004-04-24 23:33:30 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Fixed by making GDK more rebust against memory claims on BATs
and keeping track of errors reported to a client.
## Comment 820
Date: 2005-10-06 23:38:44 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: NO TEST / POSTPONED?
- test requires huge amount of data and is hence not done (yet?)
- set priority to 9 as reminder
## Comment 821
Date: 2005-11-06 10:54:31 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 822
Date: 2005-11-21 04:20:19 +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 823
Date: 2006-11-22 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 824
Date: 2006-11-29 10:05:15 +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 825
Date: 2006-12-04 20:38:10 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
We discarded writing a test for this as it indeed requires huge amounts of data, which for nightly testing isn't a good idea. Ie closing...
## Comment 826
Date: 2006-12-04 20:40:37 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
We discarded writing a test for this as it indeed requires huge amounts of data, which for nightly testing isn't a good idea. Ie closing...
## Comment 827
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 927516 at http://sourceforge.net/support/tracker.php?aid=927516
| GDKvmalloc failed on load_data | https://api.github.com/repos/MonetDB/MonetDB/issues/146/comments | 0 | 2020-11-30T08:14:51Z | 2024-06-27T10:55:29Z | https://github.com/MonetDB/MonetDB/issues/146 | 753,259,434 | 146 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-01 11:24:39 +0200
From: @drstmane
To: @njnes
Version: -- development
Last updated: 2004-04-07 08:17:23 +0200
## Comment 812
Date: 2004-04-01 11:24:39 +0200
From: @drstmane
When loading module sql_server, Mserver immediately exits,
in case sql_server cannot create (access?) the sql logdir:
========
11:23:30 manegold@draco:~ $ x Mserver
-rwxr-xr-x 1 monetdb monetdb 14331 Mar 31 11:19
/usr/bin/Mserver*
11:23:33 manegold@draco:~ $ Mserver
--dbfarm=/tmp/MonetDB/dbfarm
Monet Database Server V4.3.16_rc06
Copyright (c) 1993-2004, CWI. All rights reserved.
compiled for i686-pc-linux-gnu/32bit; dynamically linked.
monet>module(sql_server);
logger could not create log directory
/var/lib/MonetDB/log/demo/
11:23:40 manegold@draco:~ $
========
Wouldn't it be nicer to just cancel loading sql_server,
but keep Mserver alive?
## Comment 813
Date: 2004-04-07 20:17:23 +0200
From: @njnes
Logged In: YES
user_id=43556
when mvc_init fails the sql_server module is now gracefully
dropped
## Comment 814
Date: 2005-10-06 23:31:17 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/bad_sql_logdir.SF-927371*
## Comment 815
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 927371 at http://sourceforge.net/support/tracker.php?aid=927371
| Mserver exits if loading sql_server fails on logdir | https://api.github.com/repos/MonetDB/MonetDB/issues/145/comments | 0 | 2020-11-30T08:14:48Z | 2024-06-27T10:55:28Z | https://github.com/MonetDB/MonetDB/issues/145 | 753,259,406 | 145 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-01 10:05:59 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-07 08:18:23 +0200
## Comment 805
Date: 2004-04-01 10:05:59 +0200
From: @grobian
It seems like the timestamp parsing returns null if the
given string could not be parsed as a date.
ie. SELECT timestamp '01-04-2004 10:03'; returns SQL
NULL. From an SQL point of view this is incorrect and a
parse error should be thrown stating that the given
string could not be parsed into a date. If I (the user)
wanted to have SQL NULL as value, I should simply enter
SELECT NULL;
## Comment 806
Date: 2004-04-01 11:17:51 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
I imagine this is a more general problem. Generally,
MonetDB returns nil when asked to convert incorrect strings
to atom types. Perhaps the SQL parser should note that the
input is some string and the output is nil and then complain
with a syntax error.
## Comment 807
Date: 2004-04-01 16:28:49 +0200
From: @grobian
Logged In: YES
user_id=963970
Melt-Down:
select timestamp '+-06-1783 0:00';
## Comment 808
Date: 2004-04-02 09:59:32 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
The problem is in the parser.
In the rule literal: ... | TIMESTAMP string, atom_general()
returns NULL, a legitmate result for an incorrect string.
This NULL is stored by _newAtomNode in a symbol structure.
Then in the rule for atom: literal, the symbol pointer is
stored in "an" and passed to atom_type() which dereferences
the NULL pointer.
This same problem will occur in other literals.
Assigning to Niels for fixing.
## Comment 809
Date: 2004-04-07 20:18:23 +0200
From: @njnes
Logged In: YES
user_id=43556
added checks for the return nil form atomfromstr. And handle
null
returns from atom_general
## Comment 810
Date: 2005-10-06 23:05:06 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: NO TEST / REDUNDANT
bug is covered by test script
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/timestamp_parsing.SF-927300.*
## Comment 811
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 927333 at http://sourceforge.net/support/tracker.php?aid=927333
| SQL: timestamp crashes monet or returns null on parse error | https://api.github.com/repos/MonetDB/MonetDB/issues/144/comments | 0 | 2020-11-30T08:14:45Z | 2024-06-27T10:55:27Z | https://github.com/MonetDB/MonetDB/issues/144 | 753,259,367 | 144 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-01 09:49:19 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-07 08:25:27 +0200
## Comment 801
Date: 2004-04-01 09:49:19 +0200
From: @grobian
creating a table, and defining a unique works fine, it
gets stored in the keys table, and the constraint seems
to hold.
Going into indexes (JDBC seems to look at UNIQUEs not
as a key, but as an (unique) index) a few bugs apply:
it is possible to create an index:
CREATE INDEX my_index ON my_table(my_column);
this records it in the idxs table, next to auto
generated indexes which is fine.
If I execute the above query again, NO ERROR comes up
notifying me that (apart from the fact there already is
an index on that column) the name (my_index) is already
in use. The index simply gets recorded into the idxs
table with the same name.
like:
monetdb-> select * from idxs where table_id = 1044619;
+----------
| id table_id type name
+----------
| 1044661 1044619 0 c1044605
| 1044693 1044619 0 c1044606
| 1046564 1044619 0 test_id_idx
| 1046726 1044619 0 test_id_idx
+----------
4 rows
If I create another index using:
CREATE UNIQUE INDEX my_index ON my_table(my_column);
the UNIQUE gets accepted, but not enforced, ie. I can
insert duplicate values on the unique indexed column.
dropping an index seems not to work either
monetdb-> drop index test_id_idx
monetdb-> ;
Error on line 39: Drop Index, index test_id_idx unknown
using fully specified name doesn't work either
monetdb-> drop index sys.test_id_idx;
Error on line 47: Drop Index, index test_id_idx unknown
## Comment 802
Date: 2004-04-07 20:25:27 +0200
From: @njnes
Logged In: YES
user_id=43556
drop now works correctly. Also the same key more than once
now fails.
Creating a unique index works as before (ie no key
constraint is added).
Problem is that the name 'unique index' is sort of wrong, it
should
have been 'hash', which is a good index to support 'unique
constraints', its
main purpose.
## Comment 803
Date: 2005-10-06 23:03:04 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/index_management.SF-927326.*
## Comment 804
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 927326 at http://sourceforge.net/support/tracker.php?aid=927326
| SQL: CREATE INDEX / UNIQUE | https://api.github.com/repos/MonetDB/MonetDB/issues/143/comments | 0 | 2020-11-30T08:14:42Z | 2024-06-27T10:55:26Z | https://github.com/MonetDB/MonetDB/issues/143 | 753,259,332 | 143 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-04-01 09:02:50 +0200
From: @grobian
To: @sjoerdmullender
Version: -- development
Last updated: 2004-04-01 09:39:07 +0200
## Comment 794
Date: 2004-04-01 09:02:50 +0200
From: @grobian
The timestamp interpretation seems to be broken (it
used to work)
For instance
select timestamp '01-04-2004 08:57';
returns SQL NULL,
the same happens in an insert statement.
## Comment 795
Date: 2004-04-01 09:39:07 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Try select timestamp '2004-04-01 08:57';
In other words, year-month-day instead of day-month-year.
## Comment 796
Date: 2004-04-01 09:50:09 +0200
From: @grobian
Logged In: YES
user_id=963970
there is no "date-parse error"? just null?!? That's a bug.
## Comment 797
Date: 2004-04-01 09:59:39 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Fair enough. Open a bug saying that.
## Comment 798
Date: 2004-04-01 10:02:02 +0200
From: @grobian
Logged In: YES
user_id=963970
ok, I admit I was wrong :) (it was only 8:57)
## Comment 799
Date: 2005-10-06 22:22:23 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/timestamp_parsing.SF-927300.*
## Comment 800
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 927300 at http://sourceforge.net/support/tracker.php?aid=927300
| SQL: timestamps? | https://api.github.com/repos/MonetDB/MonetDB/issues/142/comments | 0 | 2020-11-30T08:14:39Z | 2024-06-28T07:36:27Z | https://github.com/MonetDB/MonetDB/issues/142 | 753,259,307 | 142 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-31 04:36:44 +0200
From: @arjan
To: GDK devs <<bugs-common>>
Version: -- development
Last updated: 2004-04-03 03:02:10 +0200
## Comment 787
Date: 2004-03-31 16:36:44 +0200
From: @arjan
monet>t := str("André");
monet>t.print;
[ nil ]
And, through SQL:
DROP TABLE bugtest;
CREATE TABLE bugtest (
id INTEGER NOT NULL,
name VARCHAR(255)
);
INSERT INTO bugtest VALUES (1, "André");
INSERT INTO bugtest VALUES (1, "test");
COMMIT;
this crashes the server with a segfault.
## Comment 788
Date: 2004-03-31 16:58:50 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Are you using UTF-8 or Latin-1 (ISO-8859-1)?
MonetDB only supports UTF-8, and if you feed it a Latin-1
string, the result will be nil.
Still, it shouldn't crash, so that certainly is a bug.
## Comment 789
Date: 2004-04-01 14:56:04 +0200
From: @arjan
Logged In: YES
user_id=20087
I was using Latin-1, so that should be the problem.
## Comment 790
Date: 2004-04-01 15:14:22 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Actually, there was also a problem with reading UTF-8
strings which I fixed today.
I haven't been able to reproduce the crash. Does that still
happen with a freshly checked-out version?
## Comment 791
Date: 2004-04-03 15:02:10 +0200
From: @mlkersten
Logged In: YES
user_id=490798
I haven;t been able to reproduce the crash.
Consider the bug closed for the time being.
## Comment 792
Date: 2005-10-06 18:45:17 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-10-06_4.9.3/Tests/accents_in_strings.SF-926709.*
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/accents_in_strings.SF-926709.*
## Comment 793
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 926709 at http://sourceforge.net/support/tracker.php?aid=926709
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.
| accents in strings | https://api.github.com/repos/MonetDB/MonetDB/issues/141/comments | 0 | 2020-11-30T08:14:37Z | 2024-06-28T07:36:27Z | https://github.com/MonetDB/MonetDB/issues/141 | 753,259,270 | 141 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-31 07:47:23 +0200
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-04-07 08:19:47 +0200
## Comment 781
Date: 2004-03-31 07:47:23 +0200
From: @mlkersten
After a typo in the query the system does not seem to reset
the transaction properties.
>select count(*) from tapestry where attr2<0;
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = select count(*) from tapestry where attr2<0;
ERROR = !Identifier: attr2 unknown or ambigious
>select count(*) from tapestry where attr1<0;
current transaction is aborted (please ROLLBACK)
Timer 32359 usec
>select count(*) from tapestry where attr1<0;
current transaction is aborted (please ROLLBACK)
Timer 34466 usec
>select ;
current transaction is aborted (please ROLLBACK)
Timer 39433 usec
>select 1;
current transaction is aborted (please ROLLBACK)
Timer 39273 usec
>
## Comment 782
Date: 2004-03-31 08:06:29 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Possitive point. After the ROLLBACK you can issue a request.
Question is what is the correct semantics according to the
standard. Such errors do not affect the database and as such
need not be rolled back.
## Comment 783
Date: 2004-03-31 09:00:58 +0200
From: @grobian
Logged In: YES
user_id=963970
Although it feels wrong, I think this IS correct behaviour
though.
Because mapiclient has no autocommit mode, the whole session
is seen as one transaction. In such transaction, upon every
error the transaction is (and should be) invalidated.
Suppose the typo was intended, then it seems like the
current database does not not fit for the transaction. One
thing that is for sure is that all following statements are
deemed to be dirty/phantom reads which are completely
unpredictable. Because the column doesn't exist the current
transaction MUST be invalidated.
If you want to query in a more relaxed environment, consider
the use of a tool which has auto-commit support, so
errornous statements get directly rolled back.
(For instance the JDBC suite has such command line tool :) )
## Comment 784
Date: 2004-04-07 20:19:47 +0200
From: @njnes
Logged In: YES
user_id=43556
Not a bug. But server now has a autocommit mode, which
start with 'set autocommit=true;' and then these rollback
requests
are gone.
## Comment 785
Date: 2005-10-06 18:36:30 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/rollback_and_autocommit.SF-926453.*
... to make others profit from the time I spent on this:
it must be
"set auto_commit=true;",
not
"set autocommit=true;"
...
## Comment 786
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 926453 at http://sourceforge.net/support/tracker.php?aid=926453
| Errorneous Rollback request | https://api.github.com/repos/MonetDB/MonetDB/issues/140/comments | 0 | 2020-11-30T08:14:34Z | 2024-06-28T07:36:26Z | https://github.com/MonetDB/MonetDB/issues/140 | 753,259,242 | 140 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-31 07:41:43 +0200
From: @mlkersten
To: @sjoerdmullender
Version: -- development
Last updated: 2007-02-25 06:27:58 +0100
## Comment 771
Date: 2004-03-31 07:41:43 +0200
From: @mlkersten
The assertion in streams is fired when you kill
MapiClient while it is receiving a large answer. It
seems that this is too strict.
monet>Mserver:
/ufs/mk/opensource/MonetDB/src/common/stream.mx:1734:
bs_write: Assertion `s->nr < sizeof(s->buf)' failed.
Aborted
## Comment 772
Date: 2004-03-31 10:45:14 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
In three different places it could happen that the server
tried writing to a stream on which an error had occurred.
These have been fixed.
## Comment 773
Date: 2005-10-06 18:22:51 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: NO TEST / POSTPONED?
- postponed test of server behaviour when client is killed
while receiving data
- set priority to 9 as reminder
## Comment 774
Date: 2005-11-06 10:53:56 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 775
Date: 2005-11-21 04:20:19 +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 776
Date: 2006-11-22 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 777
Date: 2006-11-29 10:05:07 +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 778
Date: 2006-12-04 19:01:14 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
No urgency. accept its close.
## Comment 779
Date: 2007-02-25 18:27:58 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
This situation is difficult to capture in a test,
because such errors are 1) depend on not handling
the socket close or 2) protocol error
A manual check against CURRENT with TPCH using
'select * from lineitem' did not reproduce the
situation.
Consider the bug once again closed. NO TEST FILE REQUIRED
## Comment 780
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 926450 at http://sourceforge.net/support/tracker.php?aid=926450
| Crash server with MapiClient | https://api.github.com/repos/MonetDB/MonetDB/issues/139/comments | 0 | 2020-11-30T08:14:32Z | 2024-06-27T10:55:22Z | https://github.com/MonetDB/MonetDB/issues/139 | 753,259,219 | 139 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-30 11:00:06 +0200
From: @mlkersten
To: @sjoerdmullender
Version: -- development
Last updated: 2004-03-31 12:38:03 +0200
## Comment 763
Date: 2004-03-30 23:00:06 +0200
From: @mlkersten
Experiment runSQLSelectivityPrint performs a selection
over a 1M bat from 0 to 100% and prints the result.
From an earlier experiment it is known that selection and
saving into a new table is consistently faster then the
other
systems.
However, in MonetDB/SQL the cost of printing large
tables explodes. On small results it is about 80 times
faster than PostgreSQL and on 100% selectivity about 4
times slower.
Figure is available in crackers directory
SelPrint-29March.ps
We have to look into IO resource management.
Are we allocating space without releaseing it???
## Comment 764
Date: 2004-03-30 23:29:54 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Also included the information for Mysql. Things look pretty
bad for MonetDB. It can;t get the data fast enough out of
the system using print.
see ~mk/crackers/SelPrint-29March.ps
## Comment 765
Date: 2004-03-31 01:14:14 +0200
From: @drstmane
Logged In: YES
user_id=572415
On draco (Athlon, 1.4 GHz), *print*ing a 1M tuple [oid,int]
*BAT* in *MIL* takes just 3(!) seconds(!) --- no matter
whether a 100% (scan)select is involved or not.
3 seconds is about the performance that your plot shows for
MySQL and Postgres, though I don't know, on which machine
you ran it (orion?).
"ssrun"-profiling on medusa reveals that MIL "print()"
spends about 60% of it's time in libstream's "file_write",
these 60% are evenly made-up by libc's "fwrite",
"___libc_lockfile", & "___libc_unlockfile".
IMHO, the performance problem that you observed is not in
(server-side) *printing* but in the communication between
server and client, including the parsing and re-printing in
the client.
Sorry, I had no time, yet(?), to also profile the SQL print
(i.e., the client-server interaction/communication and the
client-side parsing/re-printing...
## Comment 766
Date: 2004-03-31 08:00:51 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Indeed a bug in MapiClient. The intend was not to cache the
results obtained from the server, but dump them on the screen
immediately.
However, it called the routine to fully cache the result first.
## Comment 767
Date: 2004-03-31 11:12:42 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Using mapi_quick_query instead of the combo
mapi_query/mapi_quick_response in MapiClient (when using the
-s option) the time to print a 102400-tuple result to
/dev/null went down from 5.35 seconds to 0.31 seconds.
## Comment 768
Date: 2004-03-31 12:38:03 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
With the change in MapiClient, and when using the -s option
to perform the tests, the tests performed faster than the
competition, so the bug can be closed.
## Comment 769
Date: 2005-10-06 18:20:48 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: NO TEST / PERFORMANCE
## Comment 770
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 926254 at http://sourceforge.net/support/tracker.php?aid=926254
| Exploding print cost | https://api.github.com/repos/MonetDB/MonetDB/issues/138/comments | 0 | 2020-11-30T08:14:28Z | 2024-06-27T10:55:21Z | https://github.com/MonetDB/MonetDB/issues/138 | 753,259,180 | 138 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-30 08:35:06 +0200
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-04-07 08:27:14 +0200
## Comment 757
Date: 2004-03-30 20:35:06 +0200
From: @grobian
it looks like savepoint <name> is supported, but
rollback <name> and commit <name> are not accepted by
the parser.
Attached my little testscript for savepoints.
Could not find a database which supports savepoints.
## Comment 758
Date: 2004-03-30 20:36:38 +0200
From: @grobian
Logged In: YES
user_id=963970
seems like you fist have to check the checkbox that you want
to upload a file... browsing for a file seems not enough...
(sigh)
## Comment 759
Date: 2004-04-07 20:26:57 +0200
From: @njnes
Logged In: YES
user_id=43556
now correctly implemented (parsing was correct allready, ie
there is no commit <name>) but
only:
rollback to savepoint <name>
release savepoint <name>
commit
## Comment 760
Date: 2005-10-06 18:17:17 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/savepoints.SF-926177.*
BugDay_2005-10-06: REASSIGNED TO mr-meltdown
Fabian, could you please verify the stable output?
## Comment 761
Date: 2005-10-06 20:46:45 +0200
From: @grobian
Logged In: YES
user_id=963970
BugDay_2005-10-06: HANDLED BY meltdown
BugDay_2005-10-06: ALREADY IN TESTWEB
sql/src/test/savepoints.sql
removed duplicate bug in bugday
## Comment 762
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 926177 at http://sourceforge.net/support/tracker.php?aid=926177
| Savepoints only partly implemented? | https://api.github.com/repos/MonetDB/MonetDB/issues/137/comments | 0 | 2020-11-30T08:14:26Z | 2024-06-27T10:55:20Z | https://github.com/MonetDB/MonetDB/issues/137 | 753,259,141 | 137 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-29 07:52:45 +0200
From: @drstmane
To: @sjoerdmullender
Version: -- development
Last updated: 2004-03-30 03:55:35 +0200
## Comment 751
Date: 2004-03-29 19:52:45 +0200
From: @drstmane
Test tests/by_Peter/Tests/tst_writestr show quite a
significant performance difference between 32-bit and
64-bit systems:
On 32-bit systems, it takes only 18 secs (P4, 2.2 Ghz)
to 100 secs (PIII, 450 MHz);
on 64-bit systems, it takes 190 secs (Itanium2, 1.3GHz)
to 15 minutes(!) (R12k, 250MHz).
The test does a lot of str processing, hence, I suspect
this phenomenon is the same as reported by Alex and
Johan; their initial analysis suggested, that the
problem might be related to double-elimination in the
str heaps.
Analyzing the nightly testing logs, I found that the
performance drop of test
tests/by_Peter/Tests/tst_writestr occurs since April
18, 2003:
========
I646.5 I646.5 I646.5 I646.5 L2.4.18 L2.4.18
L2.4.18 L2.4.18 S5.8 S5.8
GNU32d GNU64d ntv32d ntv64d GNU32d GNU64d ntv32d
ntv64d GNU32d GNU64d
========
2003.04.17:
--------
108.575 87.153 73.234 68.175 18.340 48.716 50.973
44.509 82.117 80.579
========
2003.04.18:
--------
96.941 240,100 70.561 240,114 18.117 240,211 50.459
240,214 86.794 240,322
^^^^^^^ ^^^^^^^ ^^^^^^^
^^^^^^^ ^^^^^^^
========
GNU32d GNU64d ntv32d ntv64d GNU32d GNU64d ntv32d
ntv64d GNU32d GNU64d
I646.5 I646.5 I646.5 I646.5 L2.4.18 L2.4.18
L2.4.18 L2.4.18 S5.8 S5.8
========
(240 sec = 4 min was the timeout for this test back then)
The only changes around these days were the following:
========
19:02:01 manegold@draco:~/Monet/4.3.16/MonetDB $ cvs
log -N -d'2003/04/16<2003/04/19' | cvslog
--------
2003/04/17 - sjoerd: src/gdk/gdk.mx
src/gdk/gdk_align.mx src/gdk/gdk_atoms.mx
src/gdk/gdk_bat.mx src/gdk/gdk_batop.mx
src/gdk/gdk_bbp.mx
src/gdk/gdk_delta.mx src/gdk/gdk_desc.mx
src/gdk/gdk_heap.mx
src/gdk/gdk_qsort.mx src/gdk/gdk_relop.mx
src/gdk/gdk_scanselect.mx
src/gdk/gdk_search.mx src/gdk/gdk_storage.mx
src/mel/glue.mx
src/mel/proto.mx src/modules/contrib/mprof.mx
src/modules/contrib/tpcd.mx
src/modules/contrib/vector.mx
src/modules/plain/algebra.mx
src/modules/plain/bat.mx
src/modules/plain/blob.mx
src/modules/plain/ddbench.mx
src/modules/plain/decimal.mx
src/modules/plain/enum.mx
src/modules/plain/radix.mx
src/modules/plain/streams.mx
src/modules/plain/sys.mx
src/modules/plain/xtables.mx
src/monet/monet_interpreter.mx
src/monet/monet_multiplex.mx
src/monet/searchpath.mx
The before tag is Before64BitStage3, the after tag is
After64BitStage3.
This is step 3 towards the support of 64 bits.
I introduced a new type var_t for the pointers into the
heap for
var-sized BATs. There is also a symbol TYPE_var which
gets expanded
to TYPE_int or TYPE_lng, depending on the size of a var_t.
Everywhere where we now point into the heap, we use
var_t, and var_t
is 64 bits on a 64-bit machine.
The upshot is: heaps can be bigger than 4 GB on a
64-bit machine.
Lots more testing to do, but things seem to work pretty
well.
===================================================================
19:02:23 manegold@draco:~/Monet/4.3.16/MonetDB $
========
Hence, it seems to be obvious, that the 64-bit
performance problems are related to these changes, and
hence indeed related to var-sized atom (str) heaps.
Further investigations are necessary to finally locate
and eventually solve the problem.
## Comment 752
Date: 2004-03-30 00:18:27 +0200
From: @drstmane
Logged In: YES
user_id=572415
Profiling on medusa revealed that the hash-table lookup for
double-elimination in str heaps is indeed the performance
bottleneck with 64-bit (see attached profiling extract).
The fact that on the same data, strPut calls strCmpNoNil
(via GDK_STRCMP)
1,053,521,553 times with 64-bit while only
77,831,229 times with 32-bit, i.e., about 13.5 times more
often(!)
indicates that the hash table (or the bucket size/chain
length) get far too big!
Still more analysis required...
## Comment 753
Date: 2004-03-30 07:36:52 +0200
From: @drstmane
Logged In: YES
user_id=572415
Let us consider this one a high priority (performance) bug.
## Comment 754
Date: 2004-03-30 15:55:34 +0200
From: @drstmane
Logged In: YES
user_id=572415
Another debugging session of mine as been successfully ended
by Sjoerd:
Replacing a ">" by a ">=" corrected the check in str-heap
double-elimination (strPut),
whether a chunk-boundary has been exceeded, and hence the
hash table has to be cleared
(the incorrect check appeared to happen only with 64-bits...).
Also fixed GDK_STRHASHSIZE, thus that the whole hash atble
gets cleared also with 64 bits.
Thanks, Sjoerd!
## Comment 755
Date: 2005-10-06 18:06:32 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED By stmane
BugDay_2005-10-06: NO TEST / REDUNDANT
refers to test tests/by_Peter/Tests/tst_writestr
## Comment 756
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 925473 at http://sourceforge.net/support/tracker.php?aid=925473
| bad var-sized atom (str) heap performance on 64-bit | https://api.github.com/repos/MonetDB/MonetDB/issues/136/comments | 0 | 2020-11-30T08:14:22Z | 2024-06-27T10:55:19Z | https://github.com/MonetDB/MonetDB/issues/136 | 753,259,085 | 136 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-29 11:04:26 +0200
From: @drstmane
To: @sjoerdmullender
Version: -- development
Last updated: 2004-03-29 05:11:41 +0200
## Comment 747
Date: 2004-03-29 11:04:26 +0200
From: @drstmane
Since the simulated "quit();" on Ctrl-D has been added
to monet_client, many (though not all, and not always
the same)
Mapi client tests on SunOS complain about
"!WARNING: readClient: unexpected end of file;
discarding partial input"
either once Mshutdown.py is called to shutdown the
Mapi-enabled Mserver,
or once the Mserver's console (stdin) is closed thereafter.
I have not idea, yet, whether this is related to/caused
by the very changes in monet_client.mx, or by
Mshutdown, or by Mapi, or by Mtest.py, or by python, ...
## Comment 748
Date: 2004-03-29 17:11:41 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
The problem had to do with the fact that calling Mshutdown
doesn't exit the server immediately. The thread that was
reading from an already existing client wasn't killed and
was able to read more data (in particular, the comment that
Mtest.py sends starting " All but") before it received an
end-of-file.
Fixed by using the same trick on exit as on Linux: execl
either /bin/true or /bin/false instead of calling exit().
## Comment 749
Date: 2005-10-06 17:40:26 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / REDUNDANT
MapiClient is tested in many tests
## Comment 750
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 925200 at http://sourceforge.net/support/tracker.php?aid=925200
| monet_client/Mapi error on SunOS | https://api.github.com/repos/MonetDB/MonetDB/issues/135/comments | 0 | 2020-11-30T08:14:18Z | 2024-06-27T10:55:18Z | https://github.com/MonetDB/MonetDB/issues/135 | 753,259,042 | 135 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-29 08:10:03 +0200
From: @mlkersten
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2006-12-04 08:32:59 +0100
## Comment 738
Date: 2004-03-29 08:10:03 +0200
From: @mlkersten
If you remove the definition of gdk_dbname from the
monet.conf file, starting the server will produce a
segmentation fault.
Perhaps there are more potential problems in this area.
## Comment 739
Date: 2004-03-29 08:53:18 +0200
From: @drstmane
Logged In: YES
user_id=572415
To be more precise, the segfault did (at least in my case)
not occur, if the definition of gdk_dbname was removed from
the monet.conf file, but if any variable in the monet.conf
file was defined with a value that consisted of unquoted
white space, only.
My last changes in src/common/monet_options.mx.in should fix
this bug.
## Comment 740
Date: 2005-10-06 17:37:58 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO_TEST / POSTPONED
- testing the MonetDB.conf parser is posponed
- set priority to 9 as reminder
## Comment 741
Date: 2005-11-06 10:53:28 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 742
Date: 2005-11-21 04:20:18 +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 743
Date: 2006-11-22 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 744
Date: 2006-11-29 10:06:28 +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 745
Date: 2006-12-04 20:32:59 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
added simple test with (close to) empty config file
no_monetconfig.SF-925140
## Comment 746
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 925140 at http://sourceforge.net/support/tracker.php?aid=925140
| Crash on erroneous monet.conf | https://api.github.com/repos/MonetDB/MonetDB/issues/134/comments | 0 | 2020-11-30T08:14:16Z | 2024-06-27T10:55:17Z | https://github.com/MonetDB/MonetDB/issues/134 | 753,259,009 | 134 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-28 05:11:30 +0200
From: @mlkersten
To: Peter Boncz <<boncz>>
Version: -- development
Last updated: 2007-02-20 05:53:57 +0100
## Comment 727
Date: 2004-03-28 17:11:30 +0200
From: @mlkersten
Running the DBtapestry programs repeatedly show the
following behaviour on
~/crackers/experiments/run{MIL,SQL}Tapestry
During the run after 30M/20M both client and server hang.
The client is waiting for output.
The server is still active, i.e. I can perform at least
2 successive 'print(1);' commands and get a response.
[I can not connect with another MapiClient though]
However, if I start the command 'clients()' or
'help("sigs")' the server hangs too. Forcefully
quiting the server tells me that 7 threads had to be
killed.
Looking into the issue further, the error is reported
when I am executing the command:
MapiClient -h $HOST -s 'persists(BAT("R1"),false);
commit;' </dev/null
which is about to remove a bat prepared in the previous
experiment.
Note, this behaviour has not been seen before on these
test runs. Running the experiments in isolation does
not cause these problems, which seems to suggest a race
condition or queuing issue.
Moreover, both experimental runs stop at different
places, which indicate an administration related bug.
## Comment 728
Date: 2004-03-31 17:47:25 +0200
From: @peterboncz
Logged In: YES
user_id=591107
BUGFIX 924909
this is a race condition. BBPfree saves a BAT, and then
removes it
from the BBPcache (BBPclear, requiring a GDKCachelock
and/or BBPuncacheit,
requiring a GDKswapLock). After BBPfree, unloadBarrier is
released.
(unloadBarrier is taken as long as at least one BAT is being
unloaded --
this can take a long time as the BAT may need to be saved
to disk).
TMcommit first gets all the GDKcacheLock, then all
GDKswapLocks
and then waits for unloadBarrier
The unloadBarrier exists just because in the commit we want
to
make sure that descriptors do not disappear during the
commit because
a BAT finishes saving. BBPcyns (ie commit) gets all locks,
blocks BBPtrim,
and assumes that the BBP loaded state is stable while it runs.
the deadlock is as follows:
TMcommit: BBPtrim(b):
set BBPUNLOADING bit in
BBP_status(b)
get unloadBarrier
BBPfree(b)
+BBPsave
get cachelock,swaplocks |
get unloadbarrier ==> BLOCKS! |
+BBPclear(b)
get cacheLock ==>BLOCKS!
the resolution is to release unloadBarrier from within BBPfree
after
the save has been done but before the BBPclear
However, this would mean that BBPclear would remove the
bat from BBP_cache
during the commit. That defeats the purpose of unloadbarrier.
Therefore a BBPuncacheit in BBPfree must happen before
giving
back the barrier (after giving back the barrier, this BBP_cache
() may not
be modified anymore).
But, BBPuncachit took a BBPswapLock! This would have
caused another deadlock,
as TMcommit, blocked on unloadbarrier, already has all
swaplocks.
The way out is to realize that BBPuncacheit does not need
the swapLock at all.
You may read from BBP_cache() without locking, but *only*
if you have got
a BBPfix on the BAT. You can never have a BBPfix while a
BAT is unloading.
So we know that nobody is reading BBP_cache(bid) anyway
during a BBPuncacheit.
summary of modifications:
- give back unloadBarrier from within BBPfree instead of after
it
- remove locking from BBPuncacheit
## Comment 729
Date: 2005-10-06 17:33:00 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / POSTPONED?
- race condition tests are not (easily) possible in Mtest (yet?)
- set priority to 9 as reminder
## Comment 730
Date: 2005-11-06 10:52:34 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 731
Date: 2005-11-21 04:20:18 +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 732
Date: 2006-11-22 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 733
Date: 2006-11-29 10:06:20 +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 734
Date: 2006-12-04 18:53:48 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
accepted the explanation and resolution
## Comment 735
Date: 2007-02-20 17:26:43 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
Originator: NO
Nice that you accepted the explanation and resolution, but what about a test?
## Comment 736
Date: 2007-02-20 17:53:57 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
The DBtapestry code is dormant for too long to
revive it and make a (stress) test out of it.
Also, too many changes in the concurrent handling
seems this issue (if still existent) will have been
tested.
So, closed with no test file left behind.
## Comment 737
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 924909 at http://sourceforge.net/support/tracker.php?aid=924909
| Race/Queuing/coordination issues | https://api.github.com/repos/MonetDB/MonetDB/issues/133/comments | 0 | 2020-11-30T08:14:12Z | 2024-06-28T07:36:26Z | https://github.com/MonetDB/MonetDB/issues/133 | 753,258,972 | 133 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-27 09:30:24 +0100
From: @mlkersten
To: @mlkersten
Version: -- development
Last updated: 2004-03-28 12:52:08 +0200
## Comment 721
Date: 2004-03-27 21:30:24 +0100
From: @mlkersten
Recent changes to the SQL-mapi protocol has broken
the interaction with Mknife on SQL
## Comment 722
Date: 2004-03-27 22:45:41 +0100
From: @grobian
Logged In: YES
user_id=963970
Any specific details?
My best guess is that the newly introduced transaction
support screws things up. This for short means that whenever
there is an error by the parser, the current transaction is
invalidated, resulting in errors upon every action which is
not commit or rollback.
I consider it a bug that (syntactical) parser errors are
considered to break the transaction as well.
## Comment 723
Date: 2004-03-27 22:52:39 +0100
From: @grobian
Logged In: YES
user_id=963970
I take back the last sentence. In a transaction each error
breaks the transaction, since the effect of all following
statements are undefined if some earlier statement was not
able to run.
It's simply a matter of using some sort of auto-commit
strategy in order to continue after errors.
## Comment 724
Date: 2004-03-28 12:52:08 +0200
From: @mlkersten
Logged In: YES
user_id=490798
The authorization check in sql_server was tigthened.
Mknife also performs a better check on session initialization.
## Comment 725
Date: 2005-10-06 17:28:12 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / INTERACTIVE
## Comment 726
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 924594 at http://sourceforge.net/support/tracker.php?aid=924594
| Mknife broken on SQL | https://api.github.com/repos/MonetDB/MonetDB/issues/132/comments | 0 | 2020-11-30T08:14:09Z | 2024-06-27T10:55:15Z | https://github.com/MonetDB/MonetDB/issues/132 | 753,258,939 | 132 |
[
"MonetDB",
"MonetDB"
] | # Deleted Bugzilla Bug
Date: 2020-11-30T08:13:31Z
This bug was created during the import from Bugzilla.
GitHub Issues Tracker allows to assign consecutive
ids only. So this dummy issue was created in place of
a deleted Bugzilla Bug.
Please ignore this report.
| Deleted Bugzilla Bug | https://api.github.com/repos/MonetDB/MonetDB/issues/131/comments | 0 | 2020-11-30T08:13:31Z | 2020-11-30T08:13:33Z | https://github.com/MonetDB/MonetDB/issues/131 | 753,258,542 | 131 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-26 08:21:20 +0100
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2006-12-04 07:00:09 +0100
## Comment 704
Date: 2004-03-26 08:21:20 +0100
From: @mlkersten
set explain='SQL';
select count(*) from tapestry;
does not update the history table,
but dumps information on the Mserver console.
insert into history values(now,"select count(*) from
tapestry;",82,67804,6971553,7039441,user);
Here 'user' should also be replaced by the actual
user id, i.e. monetdb
## Comment 705
Date: 2004-03-26 09:16:59 +0100
From: @njnes
Logged In: YES
user_id=43556
The query needs to be single quote after recent changes
## Comment 706
Date: 2005-10-06 17:14:06 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / POSTPONED?
- DBtapestry specific test; no simple test possible
- set priority to 9 as reminder
## Comment 707
Date: 2005-11-06 10:52:05 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 708
Date: 2005-11-21 04:20:18 +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 709
Date: 2006-11-22 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 710
Date: 2006-11-29 10:06:13 +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 711
Date: 2006-12-04 18:54:48 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
Accepted resolution
## Comment 712
Date: 2006-12-04 19:00:09 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
Closing as some tests for this allready exists
for example
./src/test/bugs/Tests/profile_bug-SF.1314649.sql
## Comment 713
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 923710 at http://sourceforge.net/support/tracker.php?aid=923710
| History table not updated | https://api.github.com/repos/MonetDB/MonetDB/issues/130/comments | 0 | 2020-11-30T08:12:58Z | 2024-06-27T10:55:14Z | https://github.com/MonetDB/MonetDB/issues/130 | 753,258,175 | 130 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-26 08:19:10 +0100
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2006-12-04 04:01:52 +0100
## Comment 689
Date: 2004-03-26 08:19:10 +0100
From: @mlkersten
It appears that the query
select count(*) from tapestry
over a 30M table takes about 10 seconds.
If I am correct, this count function does not exclude
null values, which means the execution time should be
in the order of a millisecond.
## Comment 690
Date: 2004-03-26 10:18:05 +0100
From: @drstmane
Logged In: YES
user_id=572415
set explain="mil";
select count(*) from tapestry;
will tell you, what it actually does ;-)
## Comment 691
Date: 2004-03-26 11:48:32 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
+ MapiClient -lsql -umonetdb -Pmonetdb
>set explain="mil";
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = set explain="mil";
ERROR = !Identifier: mil unknown or ambigious
## Comment 692
Date: 2004-03-27 08:03:12 +0100
From: @njnes
Logged In: YES
user_id=43556
set explain='mil'; should work just fine.
7secs are spent in kdiff(30M, 0)! As kdiff with small right
operants
is used offten in sql I reassign this to peter.
3secs are spent in a not needed join (optimizer?)
## Comment 693
Date: 2004-03-27 23:27:28 +0100
From: @peterboncz
Logged In: YES
user_id=591107
probably loading of the table
is this a serious bug, or just limiting the damage to 99 euros?
## Comment 694
Date: 2004-03-27 23:54:17 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Here some snippets from the SQL solution
>print(count(tapestry_attr0_14));
[ 31457280 ]
Timer 39311 usec
>z:= new(oid,int);
>a:= kdiff(tapestry_attr0_14,z);
Timer 4569149 usec
>b:= tapestry_attr0_14.copy();
Timer 3039275 usec
>a:= kdiff(tapestry_attr0_14,z);
Timer 14681513 usec
I think the kdiff could at least be replaced by a fast(er)
copy()
Perhaps even a BATview because the right operand won;t
change the operand.
## Comment 695
Date: 2004-04-08 10:37:36 +0200
From: @peterboncz
Logged In: YES
user_id=591107
it would be better in sql to switch from using kdiff to using
replace, that is, modify the table fragments yielded by the
query to reflect the new values.
this involves a positional join between the deltas and a void-
oid remapping, and then positional replaces.
this saves an entire pass and materializaion of the
intermediate results.
of course the kdiff could be optimized for one (more) special
case, but that I do not consider a bug.
## Comment 696
Date: 2005-10-06 17:10:27 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / PERFORMANCE TEST POSTPONED
- set priority to 9 as a reminder
## Comment 697
Date: 2005-11-06 10:51:40 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 698
Date: 2005-11-21 04:20:17 +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 699
Date: 2006-11-22 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 700
Date: 2006-11-29 10:06:08 +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 701
Date: 2006-12-04 16:01:52 +0100
From: @njnes
Logged In: YES
user_id=43556
Originator: NO
added tests slow_count_1,2. First generates big table, second does the count. The bug it self is fixed because big tables are now mmapped.
## Comment 702
Date: 2006-12-04 17:56:11 +0100
From: @mlkersten
Logged In: YES
user_id=490798
Originator: YES
This error report is obsolete and covered by a separate
bug. The experiments with scaleable TPC-H already covers
this space sufficiently.
## Comment 703
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 923709 at http://sourceforge.net/support/tracker.php?aid=923709
| Slow counting | https://api.github.com/repos/MonetDB/MonetDB/issues/129/comments | 0 | 2020-11-30T08:12:55Z | 2024-06-27T10:55:13Z | https://github.com/MonetDB/MonetDB/issues/129 | 753,258,139 | 129 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-25 06:02:57 +0100
From: @MarcinZukowski
To: @sjoerdmullender
Version: -- development
Last updated: 2007-11-30 04:20:04 +0100
## Comment 678
Date: 2004-03-25 18:02:57 +0100
From: @MarcinZukowski
mapi_query returns MapiHdl object.
I would expect it to return NULL if the query resulted in an error, but
it seems it always returns a not-null pointer (again, documentation
on the web is incorrect).
Hence, the only way to check if the query succeeded is to try
mapi_error() but it's behaviour is incorrect.
In the following code:
h1 = mapi_query(conn, "ls;");
/* this call was successfull, mapi_error(conn)==0 */
h2 = mapi_query(conn, "foo?;");
/* this is an errror, mapi_error(conn)==-1 */
h3 = mapi_query(conn, "ls;");
/* this call was successfull, BUT mapi_error(conn)==-1 */
A check on mapi_error_str after h3 returns the same error string as
after h2.
## Comment 679
Date: 2004-03-25 21:34:13 +0100
From: @drstmane
Logged In: YES
user_id=572415
Set Category to "Mapi".
Group is still open;
which version of MonetDB do you refer to, Marcin?
## Comment 680
Date: 2004-03-26 14:32:38 +0100
From: @MarcinZukowski
Logged In: YES
user_id=607094
I refer to last rc branch, as usual.
Thanks for creating Mapi group :)
Increased the priority since it's crucial to make PHP extension work.
## Comment 681
Date: 2004-03-26 15:09:52 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
The bug is fixed.
If you want the documentation fixed, please submit a new
report with information on where *exactly* you found the
erroneous information.
## Comment 682
Date: 2005-10-06 17:08:07 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / C-code / POSTPONED
C-code test is postponed
set priority to 9 as reminder
## Comment 683
Date: 2005-11-06 10:51:12 +0100
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: "left-overs" are now marked as "pending"
## Comment 684
Date: 2005-11-21 04:20:17 +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 685
Date: 2006-11-22 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 686
Date: 2006-11-29 10:06:01 +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 687
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 688
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 923257 at http://sourceforge.net/support/tracker.php?aid=923257
| mapi_query() incorrect behaviour | https://api.github.com/repos/MonetDB/MonetDB/issues/128/comments | 0 | 2020-11-30T08:12:52Z | 2024-06-27T10:55:12Z | https://github.com/MonetDB/MonetDB/issues/128 | 753,258,111 | 128 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-25 04:52:03 +0100
From: @MarcinZukowski
To: @sjoerdmullender
Version: -- development
Last updated: 2004-03-30 12:47:40 +0200
## Comment 669
Date: 2004-03-25 16:52:03 +0100
From: @MarcinZukowski
Just an example:
--cut--
MapiHdl mapi_query(Mapi mid, char *Command)
: This routine sends the Command to the database server
represented by mid. It is one of the most common operations. If
Command is zero it takes the last query string kept around. It
returns zero upon success, otherwise it indicates a failure of the
request. The command response is buffered for consumption, e.g.
mapi_fetch_row().
MapiMsg mapi_query_handle(MapiHdl hdl, char *Command)
: This routine sends the Command to the database server
represented by mid. It is one of the most common operations. If
Command is zero it takes the last query string kept around. It
returns zero upon success, otherwise it indicates a failure of the
request. The command response is buffered for consumption, e.g.
mapi_fetch_row()
--cut--
Something's fishy here.
Somebody that really knows what is doing what should take a look.
## Comment 670
Date: 2004-03-25 21:35:14 +0100
From: @drstmane
Logged In: YES
user_id=572415
Set Category to "Mapi".
Group is still open;
which version of MonetDB do you refer to, Marcin?
## Comment 671
Date: 2004-03-26 14:33:16 +0100
From: @MarcinZukowski
Logged In: YES
user_id=607094
I refer to last rc branch, as usual.
Thanks for creating Mapi group :)
Slightly decreased priority.
## Comment 672
Date: 2004-03-26 15:30:52 +0100
From: @MarcinZukowski
Logged In: YES
user_id=607094
The incorrect documentation is at official Monet website:
http://monetdb.cwi.nl/TechDocs/APIs/MapiC/index.html
And its source (I guess) can be found in
src/mapi/clients/C/Mapi.mx.
## Comment 673
Date: 2004-03-28 11:33:09 +0200
From: @mlkersten
Logged In: YES
user_id=490798
Indeed, we have to look into it and decide on the behavior/doc.
mapi_query creates the query handle and mapi_query_handle
re-uses a handle
It is unclear what happens when an error occurs.
In mapi_query you still have the handle, while documentation
tells you a zero is returned.
In mapi_query_handle too.
Since it is heavily used in the ODBC interface, the
hot-potato is passed on.
## Comment 674
Date: 2004-03-29 13:28:13 +0200
From: @sjoerdmullender
Logged In: YES
user_id=43607
Marcin, please check out the latest version of the Mapi
documentation and see if the changes I made are sufficient.
## Comment 675
Date: 2004-03-30 12:47:40 +0200
From: @MarcinZukowski
Logged In: YES
user_id=607094
Documentation seems much better and accurate now, thanks to Sjoerd.
## Comment 676
Date: 2005-10-06 17:06:03 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / DOCUMENTAION
## Comment 677
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 923203 at http://sourceforge.net/support/tracker.php?aid=923203
| Incorrect Mapi documentation | https://api.github.com/repos/MonetDB/MonetDB/issues/127/comments | 0 | 2020-11-30T08:12:49Z | 2024-06-28T07:36:26Z | https://github.com/MonetDB/MonetDB/issues/127 | 753,258,066 | 127 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-25 11:47:43 +0100
From: @grobian
To: @njnes
Version: -- development
Last updated: 2004-03-25 04:45:16 +0100
## Comment 665
Date: 2004-03-25 11:47:43 +0100
From: @grobian
Monet Database Server V4.3.16_rc05
Copyright (c) 1993-2004, CWI. All rights reserved.
compiled for i686-pc-linux-gnu/32bit; dynamically linked.
monet>module(sql_server);
monet>Mserver:
../../../../src-4.3.16/sql/src/server/sql_mvc.mx:391:
bun_inplace: Assertion `BATcount(b) > id' failed.
Abort
Melt down takes place on the updates file of the VOC
data set.
files:
DIR = ~fabian/files/voc/sql
${DIR}/create_voc_schema.sql
${DIR}/insert_into_invoices.sql
${DIR}/insert_into_voyages.sql
${DIR}/insert_into_crewtables.sql
${DIR}/update_voyages.sql
The only necessary parts are create_voc_schema,
insert_into_voyages and the crashing update_voyages.
## Comment 666
Date: 2004-03-25 16:45:16 +0100
From: @njnes
Logged In: YES
user_id=43556
fixed transaction support, at first error stop processing
queries until commit (which will rollback) or rollback
## Comment 667
Date: 2005-10-06 17:04:54 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED By stmane
BugDay_2005-10-06: NO TEST:
VOC update files are not available.
## Comment 668
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 923035 at http://sourceforge.net/support/tracker.php?aid=923035
| Melt down: VOC update statements | https://api.github.com/repos/MonetDB/MonetDB/issues/126/comments | 0 | 2020-11-30T08:12:45Z | 2024-06-27T10:55:10Z | https://github.com/MonetDB/MonetDB/issues/126 | 753,258,025 | 126 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 11:07:08 +0100
From: @mlkersten
To: @sjoerdmullender
Version: -- development
Last updated: 2004-03-25 10:54:21 +0100
## Comment 658
Date: 2004-03-24 23:07:08 +0100
From: @mlkersten
Na deze maandag het tentamen ADB te hebben gemaakt, ben
ik weer aan de
slag gegaan met het practicum.
Helaas gaat het niet zo soepel als gehoopt. Ik had al
eerder een keer de
monet server (4.3.14c) geinstalleerd. Ik probeerde nu
de sql module(?hoe
noem je het eigenlijk?) te installeren, maar dat lukte
niet.
Ik merkte toen dat er een nieuwe versie is (4.3.16) ik
heb zowel de
rpm's geprobeerd als de source tar, maar beide kreeg ik
niet aan de
praat.
De rpm gaf:
error: Failed dependencies:
libmutils.so.0 is needed by
MonetDB-server-4.3.16_rc04-1
libstream.so.0 is needed by
MonetDB-server-4.3.16_rc04-1
En ik kon geen informatie vinden over de betreffende
libraries.
De tar file (MonetDB-4.3.16_rc04.tar.gz) compileerde
niet. (fout tijdens
'make') In server_tar_4.3.16.tar.gz heb ik de
config.log en de outputs
van bootstrap, configure en make bijgevoegd.
Vervolgens heb ik de cvs versie gedownload (4.3.17) -
deze compileerde
wel - en heb ik nogmaals de sql module gedownload en
geprobeert te
installeren. Helaas lukte dit weer niet: Tijdens make
gebeurt vrij vlug
het volgende:
lockfile -r 2 waiting
bison -y -d -p sql -v sql_parser.y || { rm -f -f
waiting ; exit 1 ; }
sql_parser.y:1217.14-1228.16: type clash (`sym' `sval')
on default
action
sql_parser.y:1228.17: parse error, unexpected ":",
expecting ";" or "|"
sql_parser.y:2054.10: parse error, unexpected ":",
expecting ";" or "|"
sql_parser.y:2062.4-11: invalid $ value
sql_parser.y:2062.4-11: $5 of `nonzero' has no declared
type
sql_parser.y:2062.4-2065.86: invalid $ value
sql_parser.y:2062.4-2065.86: $5 of `nonzero' has no
declared type
sql_parser.y:2062.4-2069.50: invalid $ value
...
...
sql_parser.y:2103.4-2110.42: $5 of `nonzero' has no
declared type
make[3]: *** [sql_parser.tab.c] Error 1
Ook hiervan heb ik de config.log en output van
bootstrap, configure en
make bijgevoegd. (sql_cvs.tar.gz)
Ik draai suse 9.0 op een athlon 750, asus av7, 128mb
frans@Frenzy:~/Documents/UvA/adb/install_trouble> gcc -v
Reading specs from
/usr/lib/gcc-lib/i586-suse-linux/3.3.1/specs
Configured with: ../configure --enable-threads=posix
--prefix=/usr
--with-local-prefix=/usr/local --infodir=/usr/share/info
--mandir=/usr/share/man --libdir=/usr/lib
--enable-languages=c,c++,f77,objc,java,ada
--disable-checking
--enable-libgcj --with-gxx-include-dir=/usr/include/g++
--with-slibdir=/lib --with-system-zlib --enable-shared
--enable-__cxa_atexit i586-suse-linux
Thread model: posix
gcc version 3.3.1 (SuSE Linux)
Ik hoop dat u hier iets aan heeft. Om verder te komen
moet ik iig de sql
module aan het werk krijgen dus als u suggesties heeft,
hoor ik dat
graag.
Met vriendelijke groet,
Frans Oliehoek
## Comment 659
Date: 2004-03-24 23:15:40 +0100
From: @njnes
Logged In: YES
user_id=43556
De rpm gaf:
error: Failed dependencies:
libmutils.so.0 is needed by
MonetDB-server-4.3.16_rc04-1
libstream.so.0 is needed by
MonetDB-server-4.3.16_rc04-1
This remark is correct. This simply means that the server
depends
on the client package. Try installing all rpms at once!
## Comment 660
Date: 2004-03-25 10:18:41 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
Niels has adequately explained the first issue in this report.
As to the problem with sql_parser.y, what is the version of
bison you are using?
I can't reproduce the problems with the version I have.
I have added some missing semicolons to the grammar which
should fix some of the error messages, but the first one
(sym vs. sval) is still a mystery to me.
So, I need the bison version number.
## Comment 661
Date: 2004-03-25 10:29:57 +0100
From: @grobian
Logged In: YES
user_id=963970
According to the SuSe website (it's german):
http://www.suse.de/de/private/products/suse_linux/i386/packages_professional/index_all.html
bison 1.75 (The GNU Parser Generator)
## Comment 662
Date: 2004-03-25 10:54:21 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
I downloaded and compiled bison 1.75 and tried it on the
grammar, both before and after my semicolon checkin.
The version before the checkin gave lots or errors (among
others, the errors in the bug report), but the version after
the checkin only gave the usual benign message about
shift/reduce and reduce/reduce conflicts.
I conclude that my change fixed the bug, so I close the report.
## Comment 663
Date: 2005-10-06 17:01:48 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / COMPILATION
## Comment 664
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922760 at http://sourceforge.net/support/tracker.php?aid=922760
| Compilation on Suse | https://api.github.com/repos/MonetDB/MonetDB/issues/125/comments | 0 | 2020-11-30T08:12:42Z | 2024-06-27T10:55:09Z | https://github.com/MonetDB/MonetDB/issues/125 | 753,257,987 | 125 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 09:26:32 +0100
From: @grobian
To: Stefan Manegold <<Stefan.Manegold>>
Version: -- development
Last updated: 2004-03-26 05:32:20 +0100
## Comment 650
Date: 2004-03-24 21:26:32 +0100
From: @grobian
% ./Mserver
!WARNING: GDKlockHome: created directory
/Volumes/OrionUNIX/monet/program/var/MonetDB/dbfarm/demo/
!WARNING: GDKlockHome: ignoring empty or invalid .gdk_lock.
!WARNING: BBPdir: initializing BBP.
Monet Database Server V4.3.16_rc05
Copyright (c) 1993-2004, CWI. All rights reserved.
compiled for powerpc-apple-darwin6.8/32bit;
dynamically linked.
!ERROR: TBL_M_load: _kernel not a module.
!ERROR: module(_kernel): load error.
!ERROR: interpret: no matching MIL operator to
'module_push(str)'.
!ERROR: TBL_M_load: _arith not a module.
!ERROR: module(_arith): load error.
!ERROR: " sht1.print;"
! ^
! can't help you here, sorry.
!ERROR: TBL_M_load: _str not a module.
!ERROR: module(_str): load error.
!ERROR: " (("concat an int "+1)+" with a string
"+i1+(" with an dbl "+dbl(1.0))).print; "
!
^
! can't help you here, sorry.
!ERROR: TBL_M_load: _uchr not a module.
!ERROR: module(_uchr): load error.
!ERROR: interpret: no matching MIL operator to
'module_push(str)'.
!ERROR: TBL_M_load: _bat not a module.
!ERROR: module(_bat): load error.
!ERROR: " save(b.bbpname);"
! ^
! can't help you here, sorry.
!ERROR: TBL_M_load: _algebra not a module.
!ERROR: module(_algebra): load error.
!ERROR: " VAR fid :=
monet_fcn_nme.select(cmd).kdiff(monet_fcn_aut).reverse.min;"
!
^
! can't help you here, sorry.
!ERROR: TBL_M_load: _sys not a module.
!ERROR: module(_sys): load error.
!ERROR: " IF (tpe = TOK_PROC) RETURN "PROC: ";"
! ^
! can't help you here, sorry.
!ERROR: TBL_M_load: _trans not a module.
!ERROR: module(_trans): load error.
!ERROR: interpret: no matching MIL operator to
'module_push(str)'.
monet>
I believe this problem is reproducable on Linux as well
by removing the var directory and restarting monet.
## Comment 651
Date: 2004-03-24 21:29:48 +0100
From: @grobian
Logged In: YES
user_id=963970
I think (hope) this is related, but there is a little problem:
monet>quit;
!WARNING: interpret: convert identifier 'quit' to string.
monet>quit();
!ERROR: interpret: no matching MIL operator to 'quit()'.
monet>^C!FATAL: signal 2 caught by thread a0000ded
pressing ctrl-D lets monet go in an infinite loop of the
last error (without crashing or quitting -> not leading to
the desired effect of stopping the program)
## Comment 652
Date: 2004-03-24 22:52:47 +0100
From: @drstmane
Logged In: YES
user_id=572415
These module-loading problems are (most probably) MacOS X
10.2 aka Darwin 6.8 aka Jaguar specific.
When our major (and till recently only ;-) MacOS X "client"
upgraded from Jaguar to Panther aka Darwin 7.2 aka MacOS X
10.3, I needed change a lot of things in MonetDB, to get it
compiled and working again (see below).
Lacking a second Mac (running Jaguar) I could not take care
of "backward compatibility", yet.
The challenge is now, to the MonetDB code (and the
respective configure tests) flexible enough, to work on
both, Jaguar and Panther...
About the "quit" problem:
With no modules loaded, there is no "quit;" command available.
Since Sjoerd's fix of the hanging Ctrl-D problem, also
Ctrl-D now requires "quit", and hence does not work either
in case no modules can be loaded. (Well, since this is not a
"frequent situation", we simply must live with it ... ;-)
## Comment 653
Date: 2004-03-24 22:53:59 +0100
From: @drstmane
Logged In: YES
user_id=572415
Oops, forgot the promised cvs log extract:
===================================================================
2004/02/14 - stmane: bootstrap sysdefs.h.in conf/conf.bash
conf/monet.m4
src/mapi/Makefile.ag src/modules/calibrator/Makefile.ag
src/modules/contrib/Makefile.ag
src/modules/plain/Makefile.ag
src/modules/plain/decimal.mx
With the new "Darwin 7.2" (aka "MacOS X 10.3" aka "Panther")
and the respective versions of the configure tools
(glibtool 1.5, automake 1.6.3, autoconf 2.57),
the Darwin-specific patches in bootstrap as well as
several other Darwin-specific settings
(adding /usr/local/lib to DYLD_LIBRARY_PATH in conf/conf.bash,
adding -Ddlsym=dlsym_prepend_underscore to CFLAGS in
conf/monet.m4,
adding -module to LDFLAGS in Makefile.ag's)
are not required any more --- in contrary,
they even seem to make compilation fail, now.
Hence, we remove all these Darwin-specific settings.
For convenience, these changes a wrapped in CVS tags
"Darwin_6-8" & "Darwin_7-2".
===================================================================
## Comment 654
Date: 2004-03-25 11:49:16 +0100
From: @grobian
Logged In: YES
user_id=963970
Cannot reproduce it on linux, will try to reproduce it on
the Mac.
## Comment 655
Date: 2004-03-26 17:32:20 +0100
From: @drstmane
Logged In: YES
user_id=572415
Jaguar (Darwin 6.8) requires
CFLAGS="$CFLAGS -Ddlsym=dlsym_prepend_underscore"
fixed conf/monet.m4 accordingly.
## Comment 656
Date: 2005-10-06 17:00:09 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED By stmane
BugDay_2005-10-06: NO TEST / COMPILATION
obsolete Darwin 6.8 ("Jaguar") compilation/linking problem
## Comment 657
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922715 at http://sourceforge.net/support/tracker.php?aid=922715
| Monet gives lots of errors on startup of fresh install | https://api.github.com/repos/MonetDB/MonetDB/issues/124/comments | 0 | 2020-11-30T08:12:39Z | 2024-06-27T10:55:08Z | https://github.com/MonetDB/MonetDB/issues/124 | 753,257,957 | 124 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 07:40:12 +0100
From: @mlkersten
To: @njnes
Version: -- development
Duplicates: #122
Last updated: 2010-05-04 11:15:45 +0200
## Comment 647
Date: 2004-03-24 19:40:12 +0100
From: @mlkersten
>create table t(i int);
>insert into t values(1);
>insert into t values(2);
>select * from t where i= (select max(i) from t);
i name
int type
1 tuplecount
0 id
[ 2 ]
>select * from t having i= max(i);
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = select * from t having i= max(i);
ERROR = Connection terminated
## Comment 648
Date: 2005-10-06 16:56:46 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / REDUNDANT
duplicate bug report
## Comment 649
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922639 at http://sourceforge.net/support/tracker.php?aid=922639
Bug has invalid status, setting status to "NEW".
Previous status was "DELETED".
## Comment 13982
Date: 2010-05-04 11:15:45 +0200
From: @grobian
*** This bug has been marked as a duplicate of bug #122 ***
| Having syntax crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/123/comments | 0 | 2020-11-30T08:12:36Z | 2024-06-28T07:36:25Z | https://github.com/MonetDB/MonetDB/issues/123 | 753,257,926 | 123 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 06:57:27 +0100
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2010-05-04 11:15:45 +0200
## Comment 643
Date: 2004-03-24 18:57:27 +0100
From: @mlkersten
>create table t(i int);
>insert into t values(1);
>insert into t values(2);
>select * from t where i= (select max(i) from t);
i name
int type
1 tuplecount
0 id
[ 2 ]
>select * from t having i= max(i);
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = select * from t having i= max(i);
ERROR = Connection terminated
## Comment 644
Date: 2004-03-25 00:01:28 +0100
From: @njnes
Logged In: YES
user_id=43556
Fixed crashed, now correctly create the implicite group!
Afcourse * together with having is group by is not possible!
This error message is also added
## Comment 645
Date: 2005-10-06 16:55:35 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/having.SF-922614.*
## Comment 646
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922614 at http://sourceforge.net/support/tracker.php?aid=922614
## Comment 13983
Date: 2010-05-04 11:15:45 +0200
From: @grobian
*** Bug #123 has been marked as a duplicate of this bug. ***
| Having syntax crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/122/comments | 0 | 2020-11-30T08:12:33Z | 2024-06-27T10:55:06Z | https://github.com/MonetDB/MonetDB/issues/122 | 753,257,898 | 122 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 01:55:40 +0100
From: @menzowindhouwer
To: Menzo Windhouwer <<windhouw>>
Version: -- development
Last updated: 2004-03-24 02:18:19 +0100
## Comment 638
Date: 2004-03-24 13:55:40 +0100
From: @menzowindhouwer
The tail role of a BAT is ignored, and replaced by the
BAT name, in the print of the BAT. This only happens in
single BAT prints.
monet>VAR b := new(oid,oid).roles("A","B");
monet>b.print;
-----------------
A tmp_68 name
oid oid type
-----------------
monet>VAR c := new(oid,oid).roles("A","C");
monet>print(b,c);
-------------------------
A B C name
oid oid oid type
-------------------------
## Comment 639
Date: 2004-03-24 14:18:18 +0100
From: @menzowindhouwer
Logged In: YES
user_id=572977
fixed by also using the generic check for default head and
tail roles in the case of a print of only one BAT.
## Comment 640
Date: 2004-03-24 15:57:13 +0100
From: @menzowindhouwer
Logged In: YES
user_id=572977
Some bookkeeping ...
## Comment 641
Date: 2005-10-06 16:52:17 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
MonetDB/tests/BugDay_2005-10-06_4.9.3/Tests/tail_role_printing.SF-922417.*
## Comment 642
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922417 at http://sourceforge.net/support/tracker.php?aid=922417
| tail role ignored by print(bat) | https://api.github.com/repos/MonetDB/MonetDB/issues/121/comments | 0 | 2020-11-30T08:12:30Z | 2024-06-27T10:55:05Z | https://github.com/MonetDB/MonetDB/issues/121 | 753,257,871 | 121 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 01:43:34 +0100
From: @grobian
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2005-10-13 09:05:30 +0200
## Comment 632
Date: 2004-03-24 13:43:34 +0100
From: @grobian
An update statement which tries to set a column value
to NULL fails:
ERROR parse error at token (328) in statement: update
voyages set particulars = null
## Comment 633
Date: 2004-03-24 16:43:34 +0100
From: @njnes
Logged In: YES
user_id=43556
somewhere in the NULL fixes the correct set = NULL code got lost
## Comment 634
Date: 2004-03-25 10:44:59 +0100
From: @njnes
Logged In: YES
user_id=43556
somewhere in the NULL fixes the correct set = NULL code got lost
## Comment 635
Date: 2005-10-06 16:45:07 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / FAILS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/update_column_to_null.SF-922406.*
(for now without .stable.out!)
BugDay_2005-10-06: RE-OPEND bug report
Since there is no test scipt given, I invented the following:
create table up ( att int );
insert into up values (1);
commit;
update up set att = null;
but that results in:
MAPI = monetdb@localhost:44775
QUERY = update up set att = null;
ERROR = !ERROR: BUNreplace: access denied to tmp_551, aborting.
!ERROR: bat_replace_oid_bat_wrap: operation failed.
This look like a bug to me...
## Comment 636
Date: 2005-10-13 21:05:30 +0200
From: @drstmane
Logged In: YES
user_id=572415
Seems to be fixed in CVS, now.
## Comment 637
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922406 at http://sourceforge.net/support/tracker.php?aid=922406
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: UPDATE table SET column = null fails | https://api.github.com/repos/MonetDB/MonetDB/issues/120/comments | 0 | 2020-11-30T08:12:27Z | 2024-06-27T10:55:04Z | https://github.com/MonetDB/MonetDB/issues/120 | 753,257,835 | 120 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 01:14:10 +0100
From: @grobian
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2004-03-24 11:02:57 +0100
## Comment 628
Date: 2004-03-24 13:14:10 +0100
From: @grobian
the following doesn't work
CREATE TABLE position (
harbour varchar(23) NOT NULL,
x double NOT NULL,
y double NOT NULL
);
ALTER TABLE position ADD PRIMARY KEY (harbour);
ALTER TABLE position ADD UNIQUE (x, y);
COPY 8 RECORDS INTO position FROM stdin USING
DELIMITERS '\t';
The Cape 9.3999999999999999e+307 1.3000000000000001e+308
Ceylon 1.1999999999999999e+308 1e+308
Bantam 1.31e+308 1.095e+308
Batavia 1.3159999999999999e+308 1.0959999999999999e+308
Bengal 1.2400000000000001e+308 8.8400000000000003e+307
Coromandel 1.1999999999999999e+308 9.3500000000000003e+307
Texel 8.7999999999999998e+307 6.1100000000000002e+307
Rammekens 8.7000000000000004e+307 6.2999999999999997e+307
Although this code is produced by MapiClient's --dump
feature, the parser goes down on the very first row of
it when using MapiClient to restore it.
Maybe there should a test be added where a populated
table with all types Monet has is dumped after which it
is tried to import it again.
## Comment 629
Date: 2004-03-24 23:02:57 +0100
From: @njnes
Logged In: YES
user_id=43556
properly skip the white space between query and data.
ascii_io now also sends the 'more' prompt
## Comment 630
Date: 2005-10-06 16:31:17 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/inlined_copy_into.SF-922392.*
## Comment 631
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922392 at http://sourceforge.net/support/tracker.php?aid=922392
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: COPY x RECORDS FROM stdin USING DELIMITER '\t'; fails | https://api.github.com/repos/MonetDB/MonetDB/issues/119/comments | 0 | 2020-11-30T08:12:24Z | 2024-06-27T10:55:03Z | https://github.com/MonetDB/MonetDB/issues/119 | 753,257,802 | 119 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 01:09:27 +0100
From: @grobian
To: @sjoerdmullender
Version: -- development
Last updated: 2004-03-24 03:16:40 +0100
## Comment 624
Date: 2004-03-24 13:09:27 +0100
From: @grobian
If a file is piped into the MapiClient program (using
<) the last line is silently dropped if this line does
not end with a new line character.
This causes a commit bug if the last line of the file
was COMMIT;
:)
## Comment 625
Date: 2004-03-24 15:16:40 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
There were two bugs conspiring together to let this happen.
Both have been fixed.
The first was in MapiClient. It didn't handle incomplete
last lines gracefully. It just exited when it encountered
end of file without finishing the query it had started for
the last line.
The second was in sql_scan (SQL server) which kept asking
for more even after it was told there really was nothing more.
## Comment 626
Date: 2005-10-06 16:17:07 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/last_line_without_newline.SF-922390*
## Comment 627
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922390 at http://sourceforge.net/support/tracker.php?aid=922390
| MapiClient discards lasts line of file | https://api.github.com/repos/MonetDB/MonetDB/issues/118/comments | 0 | 2020-11-30T08:12:21Z | 2024-06-27T10:55:02Z | https://github.com/MonetDB/MonetDB/issues/118 | 753,257,766 | 118 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 11:03:43 +0100
From: @grobian
To: @sjoerdmullender
Version: -- development
Last updated: 2004-03-24 05:21:24 +0100
## Comment 618
Date: 2004-03-24 11:03:43 +0100
From: @grobian
Dunno if this is a real bug, but it is somehow weird.
create table test (
val double
);
insert into test values (1e+308);
the insert fails, parser error. If the value was
written as 1.0e+308 it succeeds. However, monet itself
returns 1e+308 on a select.
This means that the returned value cannot be used to
insert it again.
## Comment 619
Date: 2004-03-24 11:17:12 +0100
From: @grobian
Logged In: YES
user_id=963970
Since the MapiClient returns the 1+308 value as well on a
dump (--dump), this would result in an error on restore.
## Comment 620
Date: 2004-03-24 17:21:24 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
Fixed in revision 1.33 of sql_scan.mx.
## Comment 621
Date: 2004-03-24 20:03:39 +0100
From: @njnes
Logged In: YES
user_id=43556
Indeed 1e+308 is allowed in SQL-99
## Comment 622
Date: 2005-10-06 16:03:03 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/parsing_doubles.SF-922332.*
## Comment 623
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922332 at http://sourceforge.net/support/tracker.php?aid=922332
| double numbers | https://api.github.com/repos/MonetDB/MonetDB/issues/117/comments | 0 | 2020-11-30T08:12:18Z | 2024-06-27T10:55:01Z | https://github.com/MonetDB/MonetDB/issues/117 | 753,257,731 | 117 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 09:40:26 +0100
From: @grobian
To: SQL devs <<bugs-sql>>
Version: -- development
Last updated: 2004-03-24 11:04:49 +0100
## Comment 614
Date: 2004-03-24 09:40:26 +0100
From: @grobian
SELECT null AS test, "null" AS nullable FROM columns
results in:
test nullable
(char) (char)
NULL null
This is wrong, because:
NULL is not "null", the string Java would return if a
value was nil, as returned by Monet. (Has the syntax
been changed per http://sourceforge.net/support/tracker.php?aid=919528 perhaps?)
The column null from table columns is of boolean type.
What goes wrong here (used to go correct) is that the
parser sees it as a string, while the double quotes are
used to tell the parser this is a column. As far as I
know does SQL dictate that strings go in single quotes,
and double quotes are used for explicitly naming columns.
like select 'this is a string' and select
"table1"."column1" from "table1"
## Comment 615
Date: 2004-03-24 23:04:49 +0100
From: @njnes
Logged In: YES
user_id=43556
sql now only has single quoted strings
double quotes are now correctly (sql 99 spec) used
for 'quoted' idents
## Comment 616
Date: 2005-10-06 15:58:53 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/quoted_keywords.SF-922288.*
## Comment 617
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922288 at http://sourceforge.net/support/tracker.php?aid=922288
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 null and columns | https://api.github.com/repos/MonetDB/MonetDB/issues/116/comments | 0 | 2020-11-30T08:12:15Z | 2024-06-27T10:55:00Z | https://github.com/MonetDB/MonetDB/issues/116 | 753,257,694 | 116 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 08:01:15 +0100
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-03-28 10:44:28 +0200
## Comment 609
Date: 2004-03-24 08:01:15 +0100
From: @mlkersten
select 1<2;
select 'A'='a';
select not true;
select 1 from tables where 1<2;
select 1 from tables where true;
Atomary comparisons should work, just like
simple arithmetic addition.
## Comment 610
Date: 2004-03-24 09:05:34 +0100
From: @grobian
Logged In: YES
user_id=963970
from the same category:
select '1' where 1 = 0 seems to work, while
select * from tables where 1 = 0 doesn't.
## Comment 611
Date: 2004-03-28 10:44:28 +0200
From: @njnes
Logged In: YES
user_id=43556
was not handle, new code added to handle these cases
## Comment 612
Date: 2005-10-06 15:37:15 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/basic_arithmetics.SF-922252.s*
## Comment 613
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922252 at http://sourceforge.net/support/tracker.php?aid=922252
| SQL basic arithmetic | https://api.github.com/repos/MonetDB/MonetDB/issues/115/comments | 0 | 2020-11-30T08:05:37Z | 2024-06-27T10:54:59Z | https://github.com/MonetDB/MonetDB/issues/115 | 753,252,765 | 115 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-24 07:50:58 +0100
From: @drstmane
To: @njnes
Version: -- development
Last updated: 2004-03-24 08:35:06 +0100
## Comment 605
Date: 2004-03-24 07:50:58 +0100
From: @drstmane
It seems as if loading data in ascii_io seems to be
broken since yesterday's (Tuesday, March 23 2004)
changes; cf. tests
tests/suite01/Tests/test_40
tests/BugsViaSourgeforce/Tests/ID.642351
src/modules/plain/Tests/ascii_io
scripts/gold/Tests/*
Might be related to str2monet being replaced by
strToStr ...
## Comment 606
Date: 2004-03-24 08:35:06 +0100
From: @njnes
Logged In: YES
user_id=43556
white space skipping for non quote strings disappeared with
tuesdays removal
of str2monet. Restored that functionality.
## Comment 607
Date: 2005-10-06 15:14:46 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / REDUNDANT
refers to existing tests
## Comment 608
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922248 at http://sourceforge.net/support/tracker.php?aid=922248
| ascii_io data load broken | https://api.github.com/repos/MonetDB/MonetDB/issues/114/comments | 0 | 2020-11-30T08:05:35Z | 2024-06-28T07:36:25Z | https://github.com/MonetDB/MonetDB/issues/114 | 753,252,725 | 114 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-23 11:06:34 +0100
From: @njnes
To: @sjoerdmullender
Version: -- development
Last updated: 2004-03-24 12:58:33 +0100
## Comment 599
Date: 2004-03-23 23:06:34 +0100
From: @njnes
monet>var s4 := timestamp(nil);
monet>var s3 := isnil(s4);
monet>print (s3);
[ false ]
monet>print (s4);
[ nil 01:00:00.000 ]
monet>print ( isnil(timestamp(inil)) ); typo inil
instead of nil
!WARNING: interpret: convert identifier 'inil' to string.
[ true ]
monet>print ( isnil(timestamp(nil)) );
[ false ]
monet>
## Comment 600
Date: 2004-03-24 07:40:47 +0100
From: @drstmane
Logged In: YES
user_id=572415
timestamp is a {int,int} struct mapped onto a lng.
ts_nil is defined as {int_nil,int_nil}.
I suspect, that the default cast timestamp(nil) simply
produces a lng_nil instead of setting the proper ts_nil.
timestamp(str(nil)) & timestamp(int(nil)) do work OK (see
below).
similar problem occurs with tzone.
I haven't looked into details, how to fix this, yet.
I extended the monettime test script to keep us reminded.
print(timestamp(nil));
[ nil 01:00:00.000 ]
print(isnil(timestamp(nil)));
[ false ]
print(tzone(nil));
[ GMT-68:16 ]
print(isnil(tzone(nil)));
[ true ]
print(timestamp(int(nil)));
[ nil ]
print(isnil(timestamp(int(nil))));
[ true ]
print(tzone(int(nil)));
[ nil ]
print(isnil(tzone(int(nil))));
[ false ]
print(timestamp(lng(nil)));
[ nil 01:00:00.000 ]
print(isnil(timestamp(lng(nil))));
[ false ]
print(tzone(lng(nil)));
[ GMT-68:16 ]
print(isnil(tzone(lng(nil))));
[ true ]
print(timestamp(str(nil)));
[ nil ]
print(isnil(timestamp(str(nil))));
[ true ]
print(tzone(str(nil)));
[ nil ]
print(isnil(tzone(str(nil))));
[ false ]
print(timestamp(dbl(nil)));
[ nil 01:00:00.000 ]
print(isnil(timestamp(dbl(nil))));
[ false ]
print(tzone(dbl(nil)));
[ GMT-68:16 ]
print(isnil(tzone(dbl(nil))));
[ true ]
## Comment 601
Date: 2004-03-24 11:42:35 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
This bug can be fixed by this change in monettime.mx (and a
corresponding change to ts_isnil macro). I'm running make
check to see if this change breaks anything else.
This fix basically uses the lng_nil bit pattern for
timestamp(nil).
- ts_nil.days = date_nil;
- ts_nil.msecs = daytime_nil;
+ * (lng *) &ts_nil = lng_nil;
## Comment 602
Date: 2004-03-24 12:58:33 +0100
From: @sjoerdmullender
Logged In: YES
user_id=43607
Fixed with the previously mentioned solution, also applied
to tzone.
## Comment 603
Date: 2005-10-06 15:10:30 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / REDUNDANT
test script already available as
MonetDB/src/modules/plain/Tests/monettime.*
## Comment 604
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922062 at http://sourceforge.net/support/tracker.php?aid=922062
| isnil(timestamp(nil)) = false | https://api.github.com/repos/MonetDB/MonetDB/issues/113/comments | 0 | 2020-11-30T08:05:32Z | 2024-06-27T10:54:57Z | https://github.com/MonetDB/MonetDB/issues/113 | 753,252,687 | 113 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-23 09:27:54 +0100
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-03-23 09:28:51 +0100
## Comment 596
Date: 2004-03-23 21:27:54 +0100
From: @mlkersten
CREATE TABLE union_test (
id INT,
val1 VARCHAR(10),
val2 VARCHAR(10)
);
INSERT INTO union_test VALUES(1, 'abcdef','123456');
SELECT 'abc' AS c1, '123' AS c2
UNION ALL
SELECT val1 AS c1, val2 AS c2 FROM union_test;
Results in the message:
MAPI = monetdb@localhost:45123
ACTION= mapi_get_row
QUERY = SELECT 'abc' AS c1, '123' AS c2
ERROR = !ERROR: interpret: no matching MIL operator to
'access(str, int)'.
!MAYBE YOU MEAN:
! access(BAT[any::1,any::2]) : int
! access(BAT[any::1,any::2], int) :
BAT[any::1,any::2]
!ERROR: append(param 1): evaluation error.
!WARNING: BBPdecref: range error 89
Entering a non-supported area of the implementation.
## Comment 597
Date: 2005-10-06 15:02:02 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: NO TEST / DUPLICATE
same as http://sourceforge.net/support/tracker.php?aid=921985
## Comment 598
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 922001 at http://sourceforge.net/support/tracker.php?aid=922001
This bug was marked DUPLICATE in the database it was moved from.
Changing resolution to "MOVED"
| SQL union limitation | https://api.github.com/repos/MonetDB/MonetDB/issues/112/comments | 0 | 2020-11-30T08:05:29Z | 2024-06-28T07:36:25Z | https://github.com/MonetDB/MonetDB/issues/112 | 753,252,649 | 112 |
[
"MonetDB",
"MonetDB"
] | Date: 2004-03-23 09:19:02 +0100
From: @mlkersten
To: @njnes
Version: -- development
Last updated: 2004-03-23 10:40:08 +0100
## Comment 593
Date: 2004-03-23 21:19:02 +0100
From: @mlkersten
CREATE TABLE bounds_test (
id INT NOT NULL AUTO_INCREMENT PRIMARY KEY,
MAPI = monetdb@localhost:45123
ACTION= mapi_fetch_line
QUERY = CREATE TABLE bounds_test (
ERROR = Connection terminated
One of the Mysql Gotcha's problems
## Comment 594
Date: 2005-10-06 14:59:14 +0200
From: @drstmane
Logged In: YES
user_id=572415
BugDay_2005-10-06: CLAIMED BY stmane
BugDay_2005-10-06: TEST ADDED / SUCCESS ...
sql/src/test/BugDay_2005-10-06_2.9.3/Tests/union_limitation.SF-921996*
Actually, I do not understand, what the actually bug report
is about.
The given SQL snipped is obviously not complete and hence
triggers and error. That's exactly what the added test
script tests, now ...
## Comment 595
Date: 2010-05-04 09:32:09 +0200
From: Pseudo user for Sourceforge import <<sfimport>>
This bug was previously known as tracker item 921996 at http://sourceforge.net/support/tracker.php?aid=921996
| SQL parser crashes server | https://api.github.com/repos/MonetDB/MonetDB/issues/111/comments | 0 | 2020-11-30T08:05:27Z | 2024-06-28T07:36:24Z | https://github.com/MonetDB/MonetDB/issues/111 | 753,252,623 | 111 |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.