Thread: Missing bug-report numbers?
There's somebody over in pgsql-bugs complaining about lack of response
to his bug report #4593. Well, no such bug number came by here, nor is
it in the archives. The adjacent bug numbers are also missing, and I
notice quite a few other gaps. What might be the cause of missing
bug numbers?
For reference sake, the last couple dozen bugs that did arrive here are
Subject: [BUGS] BUG #4566: pg_stop_backup() reports incorrect STOP WAL LOCATIOSubject: [BUGS] BUG #4567: Clustering on
GISTINDEX clobbers records in table Subject: [BUGS] BUG #4568: sporadic error doing COUNT(*) on empty tableSubject:
[BUGS]BUG #4572: Incorrect error message when using wrong password wSubject: [BUGS] BUG #4573: download problemSubject:
[BUGS]BUG #4574: LIKE fails on non-varying character with no wildcarSubject: [BUGS] BUG #4575: All page cache in
shared_bufferspinned (duplicatedSubject: [BUGS] BUG #4576: Postgres Windows service crashesSubject: [BUGS] BUG #4578:
Notreleaseing memory from PQmakeEmptyPGresult whenSubject: [BUGS] BUG #4579: to_char & utf8Subject: [BUGS] BUG #4580:
selectsubstring ('a' from '(b)?(a)') returns nullSubject: [BUGS] BUG #4581: Difficult uninstallSubject: [BUGS] BUG
#4582:Renaming sequences and default valueSubject: [BUGS] BUG #4583: Db stopSubject: [BUGS] BUG #4584: PostgreSQL
servicedoesn't startSubject: [BUGS] BUG #4585: out parameter name cuases disruption in custom aggrSubject: [BUGS] BUG
#4586:Supporting of Binary instead Bytea for Primary KeysSubject: [BUGS] BUG #4588: post gres installation
problemSubject:[BUGS] BUG #4589: Postgressql database server 8.3 service fails to stSubject: [BUGS] BUG #4591:
postgres.exemore than 150 processes runnedSubject: [BUGS] BUG #4596: information_schema.table_privileges is way too
slowSubject:[BUGS] BUG #4597: Can't join the mailing lists!Subject: [BUGS] BUG #4599: bugfix for contrib/dblink module
regards, tom lane
Most likely, it got caught in the spamfilter. Anything that comes
through the bug report form is just sent to pgsql-bugs@, so it'll pass
through both the maia filters and potentially through moderation as well.
Maybe Marc has some logs that'll show if it was dropped? If there are,
it should be easy enough to search on the bugnumber...
//Magnus
Tom Lane wrote:
> There's somebody over in pgsql-bugs complaining about lack of response
> to his bug report #4593. Well, no such bug number came by here, nor is
> it in the archives. The adjacent bug numbers are also missing, and I
> notice quite a few other gaps. What might be the cause of missing
> bug numbers?
>
> For reference sake, the last couple dozen bugs that did arrive here are
>
> Subject: [BUGS] BUG #4566: pg_stop_backup() reports incorrect STOP WAL LOCATIO
> Subject: [BUGS] BUG #4567: Clustering on GIST INDEX clobbers records in table
> Subject: [BUGS] BUG #4568: sporadic error doing COUNT(*) on empty table
> Subject: [BUGS] BUG #4572: Incorrect error message when using wrong password w
> Subject: [BUGS] BUG #4573: download problem
> Subject: [BUGS] BUG #4574: LIKE fails on non-varying character with no wildcar
> Subject: [BUGS] BUG #4575: All page cache in shared_buffers pinned (duplicated
> Subject: [BUGS] BUG #4576: Postgres Windows service crashes
> Subject: [BUGS] BUG #4578: Not releaseing memory from PQmakeEmptyPGresult when
> Subject: [BUGS] BUG #4579: to_char & utf8
> Subject: [BUGS] BUG #4580: select substring ('a' from '(b)?(a)') returns null
> Subject: [BUGS] BUG #4581: Difficult uninstall
> Subject: [BUGS] BUG #4582: Renaming sequences and default value
> Subject: [BUGS] BUG #4583: Db stop
> Subject: [BUGS] BUG #4584: PostgreSQL service doesn't start
> Subject: [BUGS] BUG #4585: out parameter name cuases disruption in custom aggr
> Subject: [BUGS] BUG #4586: Supporting of Binary instead Bytea for Primary Keys
> Subject: [BUGS] BUG #4588: post gres installation problem
> Subject: [BUGS] BUG #4589: Postgressql database server 8.3 service fails to st
> Subject: [BUGS] BUG #4591: postgres.exe more than 150 processes runned
> Subject: [BUGS] BUG #4596: information_schema.table_privileges is way too slow
> Subject: [BUGS] BUG #4597: Can't join the mailing lists!
> Subject: [BUGS] BUG #4599: bugfix for contrib/dblink module
>
> regards, tom lane
>
Magnus Hagander wrote: > Most likely, it got caught in the spamfilter. Anything that comes > through the bug report form is just sent to pgsql-bugs@, so it'll pass > through both the maia filters and potentially through moderation as well. I think the problem is that there aren't enough moderators for pgsql-bugs (and a bunch of other lists as well). -- Alvaro Herrera http://www.CommandPrompt.com/ PostgreSQL Replication, Consulting, Custom Development, 24x7 support
Alvaro Herrera wrote: > Magnus Hagander wrote: > >> Most likely, it got caught in the spamfilter. Anything that comes >> through the bug report form is just sent to pgsql-bugs@, so it'll pass >> through both the maia filters and potentially through moderation as well. >> > > I think the problem is that there aren't enough moderators for > pgsql-bugs (and a bunch of other lists as well). > I'm willing to do some list moderation if that will help. -- Chander Ganesan Open Technology Group, Inc. One Copley Parkway, Suite 210 Morrisville, NC 27560 919-463-0999/877-258-8987 http://www.otg-nc.com
On Mon, Jan 5, 2009 at 4:11 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote: > There's somebody over in pgsql-bugs complaining about lack of response > to his bug report #4593. Well, no such bug number came by here, nor is > it in the archives. The adjacent bug numbers are also missing, and I > notice quite a few other gaps. What might be the cause of missing > bug numbers? Many are spam. Because we don't require a community registration to log bugs, we often got junk submissions from bots/idiots. Because the form sends the bugs as though they were from the submitter, most are then caught in the moderation queue where they get zapped. It's possible that a real one got lost that way of course. I'd prefer to require a community account, but others have wanted to keep the bug-reporting bar as low as possible in the past. -- Dave Page EnterpriseDB UK: http://www.enterprisedb.com
On Mon, 2009-01-05 at 08:30 +0000, Dave Page wrote: > On Mon, Jan 5, 2009 at 4:11 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote: > > There's somebody over in pgsql-bugs complaining about lack of response > > to his bug report #4593. Well, no such bug number came by here, nor is > > it in the archives. The adjacent bug numbers are also missing, and I > > notice quite a few other gaps. What might be the cause of missing > > bug numbers? > > Many are spam. Because we don't require a community registration to > log bugs, we often got junk submissions from bots/idiots. Because the > form sends the bugs as though they were from the submitter, most are > then caught in the moderation queue where they get zapped. It's > possible that a real one got lost that way of course. > > I'd prefer to require a community account, but others have wanted to > keep the bug-reporting bar as low as possible in the past. Perhaps it is time to re-open that discussion. Joshua D. Drake > > -- > Dave Page > EnterpriseDB UK: http://www.enterprisedb.com > -- PostgreSQL Consulting, Development, Support, Training 503-667-4564 - http://www.commandprompt.com/ The PostgreSQL Company,serving since 1997
Joshua D. Drake wrote: > On Mon, 2009-01-05 at 08:30 +0000, Dave Page wrote: >> On Mon, Jan 5, 2009 at 4:11 AM, Tom Lane <tgl@sss.pgh.pa.us> wrote: >>> There's somebody over in pgsql-bugs complaining about lack of response >>> to his bug report #4593. Well, no such bug number came by here, nor is >>> it in the archives. The adjacent bug numbers are also missing, and I >>> notice quite a few other gaps. What might be the cause of missing >>> bug numbers? >> Many are spam. Because we don't require a community registration to >> log bugs, we often got junk submissions from bots/idiots. Because the >> form sends the bugs as though they were from the submitter, most are >> then caught in the moderation queue where they get zapped. It's >> possible that a real one got lost that way of course. >> >> I'd prefer to require a community account, but others have wanted to >> keep the bug-reporting bar as low as possible in the past. > > Perhaps it is time to re-open that discussion. while keeping the bar as low as possible is important I fully support the idea as restricting the bug report form to community accounts only. The success of the wiki (which requires community login) has proven that this kind of "barrier" is not really a problem in practice imho. Stefan
Stefan Kaltenbrunner <stefan@kaltenbrunner.cc> writes:
> while keeping the bar as low as possible is important I fully support
> the idea as restricting the bug report form to community accounts only.
How exactly would that help the problem at hand? If the mail generated
by the bug form continues to go through the spam filters then we'll
still have the same issue.
regards, tom lane
On Mon, 2009-01-05 at 19:07 -0500, Tom Lane wrote: > Stefan Kaltenbrunner <stefan@kaltenbrunner.cc> writes: > > while keeping the bar as low as possible is important I fully support > > the idea as restricting the bug report form to community accounts only. > > How exactly would that help the problem at hand? If the mail generated > by the bug form continues to go through the spam filters then we'll > still have the same issue. If we use community login, we could white list. Joshua D. Drake > > regards, tom lane > -- PostgreSQL Consulting, Development, Support, Training 503-667-4564 - http://www.commandprompt.com/ The PostgreSQL Company,serving since 1997
Joshua D. Drake wrote: > On Mon, 2009-01-05 at 19:07 -0500, Tom Lane wrote: >> Stefan Kaltenbrunner <stefan@kaltenbrunner.cc> writes: >>> while keeping the bar as low as possible is important I fully support >>> the idea as restricting the bug report form to community accounts only. >> How exactly would that help the problem at hand? If the mail generated >> by the bug form continues to go through the spam filters then we'll >> still have the same issue. > > If we use community login, we could white list. exactly ;-) Stefan
-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 > while keeping the bar as low as possible is important I fully support > the idea as restricting the bug report form to community accounts only. > The success of the wiki (which requires community login) has proven that > this kind of "barrier" is not really a problem in practice imho. Sorry, but the success of the wiki is not a good indication of other things, such as bug reports from random people. Forcing an account creation just to report a bug is one of the most annoying and counter-productive practices in the history of open-source projects, and I really hope Postgres has no intention of going there. - -- Greg Sabino Mullane greg@turnstep.com End Point Corporation PGP Key: 0x14964AC8 200901072127 http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iEYEAREDAAYFAkllZI8ACgkQvJuQZxSWSsjtWgCfRX5Ijj3JPCEMLl781T4gySIp hTEAoPm0pVj5r3yrhjSw9zGEF0MSt1dH =WP2m -----END PGP SIGNATURE-----
-----BEGIN PGP SIGNED MESSAGE----- Hash: RIPEMD160 >> Most likely, it got caught in the spamfilter. Anything that comes >> through the bug report form is just sent to pgsql-bugs@, so it'll pass >> through both the maia filters and potentially through moderation as well. > I think the problem is that there aren't enough moderators for > pgsql-bugs (and a bunch of other lists as well). There is no "potential": all bug reports from the form are filtered for moderation, as they should be, due to the amount of form spam. However, the real problem with that lists (and the others) is not so much the lack of moderators, but the hassle the existing ones have to go through, due to the very limited spam filtering done on the lists. I've raised this problem a few times, but this time I'll make links instead of posting the sample spam itself, to prevent further filtering. Here is the most recent message sent to pgsql-bugs. Note the total lack of any spam hits (see lines 28 and 29), despite the many freely available spamassassin recipes that should have caught this: http://privatepaste.com/35TMBvAiHz Here's an interesting one that just came in to -announce: http://privatepaste.com/96hmnRRrlK Note the multiple greylisting, and the fact that this spam at least hit one of the spam tests, but at 1.37 for the DCC check, with a final criteria of 5, it's unlikely to ever be effective. My recomendations? Add a lot more SA tests, lower the overall threshhold to under 5, and tweak the numbers on some of the existing tests (I recall that some of the tests are at .01 or some useless number) - -- Greg Sabino Mullane greg@turnstep.com End Point Corporation PGP Key: 0x14964AC8 200901072134 http://biglumber.com/x/web?pk=2529DF6AB8F79407E94445B4BC9B906714964AC8 -----BEGIN PGP SIGNATURE----- iEYEAREDAAYFAkllZnIACgkQvJuQZxSWSshTVQCg7Km8fzMUftTGIqyEBabA+BqM c1EAoJAR5LgJ/4aNmprc1dS1rkSkrFlM =AMK8 -----END PGP SIGNATURE-----
Greg Sabino Mullane a écrit : > Sorry, but the success of the wiki is not a good indication of other things, > such as bug reports from random people. Forcing an account creation just > to report a bug is one of the most annoying and counter-productive practices > in the history of open-source projects, and I really hope Postgres has no > intention of going there. +1. Why not just add a simple question to the form to prevent the bots from submitting it? Something like "Write PostgreSQL in this field:" should be sufficient IMHO. It's just an example, I'm pretty sure we can come up with something simple enough for everyone. I don't think this form has sufficient value for spammers to make them write something specific to it. -- Guillaume