Bruce's Patch Queue

From PostgreSQL wiki

Revision as of 14:59, 9 September 2008 by Alvherre (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search
Patch URL Author Reviewer Status
updated hash functions for postgresql v1 1 Kenneth Marshall <ktm@rice.edu> Alvaro Herrera  ?

Tom Lane:

I think it's a large patch with a small patch struggling to get out :-(. Surely we don't need so many variants of the same code.

Alvaro Herrera:

This seems to be a reviewable patch.
why postgresql over other RDBMS 1 Gregory Stark <stark@enterprisedb.com> Bruce Momjian  ?

Bruce Momjian:

I figured it should stay with the patch if we are going to review it.

Tom Lane:

Why is this even in the queue?

Bruce Momjian:

This relates to concurrent psql not being applied and why.
Reviewing new index types (was Re: Updated bitmap indexpatch) 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
SET TRANSACTION conformance to SQL:2003 1 Simon Riggs <simon@2ndquadrant.com> ?  ?
Incomplete docs for restore_command for hot standby 1 2 Simon Riggs <simon@2ndquadrant.com> ?  ?
Add function for 1 Bruce Momjian <bruce@momjian.us> ?  ?
pg_standby observation 1 Erik Jones <erik@myemma.com> ?  ?
Add function for quote_qualified_identifier? 1 2 Bruce Momjian <bruce@momjian.us> ?  ?
Ordered Append Node 1 Gregory Stark <stark@enterprisedb.com> ?  ?
Truncate Triggers 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 Decibel! <decibel@decibel.org> ?  ?
stored procedure stats in collector 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
[PERFORM] partitioned table and ORDER BY indexed_field DESC LIMIT 1 1 "Heikki Linnakangas" <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

The other copy got removed, why not this one?

Tom Lane:

This is the same non-patch that's a few messages above.
Updated bitmap index patch 1 2 3 4 5 6 Mark Kirkwood <markir@paradise.net.nz> ?  ?
pg_next_dst_boundary optimization 1 Heikki Linnakangas <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

All of this is done already --- both the search optimization and the tzcode re-sync were committed last month.
ipcclean is excepted by windows. 1 "Hiroshi Saito" <z-saito@guitar.ocn.ne.jp> Bruce Momjian  ?

Bruce Momjian:

OK, I now agree --- let's remove it. People can always run ipcs and ipcrm themselves, and probably safer.

Tom Lane:

ipcclean is so broken that I'd never recommend anyone use it. The Linux-specific branch has an incorrect test for whether the postmaster is still running (these days it could be postgres instead) while the other-Unixen branch fails to have any test at all. Neither branch has any intelligence for the case of multiple postmasters on one machine. And in any case it's been years since the last report of PG failing to cope with pre-existing SysV resources.

Magnus Hagander:

no, ipcclean never does anything usable on Windows. a) it's a shellscript. Simply doesn't work on Windows. And it uses tools that don't exist on Windows. b) given how our shared memory work, there would be no use for it. Once the last process goes away, so does the memory. Automatically. So, for windows it should certainly be removed. but Toms comment still applies - if we should remove it completely, there is no need to remove it just for windows.

Bruce Momjian:

Uh, during 8.3 feature freeze someone did what IPC clean does for Win32 to figure out why their shared memory wasn't mapped in the child, so I think perhaps for diagnostics it is still used.

Tom Lane:

There is discussion in another thread of just removing ipcclean. Does anyone really use it?

Bruce Momjian:

This seems ready to apply
script binaries renaming 1 Zdenek Kotala <Zdenek.Kotala@Sun.COM> Bruce Momjian  ?

Tom Lane:

As best I can tell, the consensus in the pgsql-patches discussion was that we don't want this.

Bruce Momjian:

Ah, the utility command patch again.
V1.1 patch for TODO Item: SQL-language reference parameters by name. 1 "Gevik Babakhani" <pgdev@xs4all.nl> ?  ?
PQParam version 0.5 1 Andrew Chernow <ac@esilo.com> Andrew Chernow  ?

Bruce Momjian:

Yep, kept just as background material.

Andrew Chernow:

This is a very out-dated version and should only be viewed for historical background. The latest version of the libpq type system is 0.9a and can be found here: http://archives.postgresql.org/pgsql-patches/2008-03/msg00057.php It includes full documentation, implementation and regression test.
Proposal - libpq Type System beta-0.8a (was PGparam) 1 2 3 4 5 "Merlin Moncure" <mmoncure@gmail.com> ?  ?
Reviewing new index types (was Re: Updated bitmap indexpatch) 1 "Simon Riggs" <simon@2ndquadrant.com> ?  ?
[FEATURE REQUEST] Streaming Onlinebackup (Maybe OFFTOPIC) 1 "Florian G. Pflug" <fgp@phlo.org> ?  ?
V0.2 patch for TODO Item: SQL-language referenceparameters by name. 1 2 3 "Gevik Babakhani" <pgdev@xs4all.nl> ?  ?
[PATCH] Add size/acl information when listing databases 1 Andrew Gilligan <andy@tcpd.net> Greg Smith  ?

Tom Lane:

As long as the database size collection is relegated to \l+, I have no objection on performance grounds. We need to be sure it's not broken for non-superusers though.

Greg Smith:

Patch looks finished and it adds a psql feature people might like. Not clear to me if Tom was satisfied with the performance impact of collecting the info though.
Proposal: real procedures again (8.4) 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
{**Spam**} Re: Proposed patch: synchronized_scanning GUC variable 1 Dimitri Fontaine <dfontaine@hi-media.com> ?  ?
psql slash# command 1 "Sibte Abbas" <sibtay@gmail.com> ?  ?
[8.4] Updated WITH clause patch (non-recursive) 1 Neil Conway <neilc@samurai.com> ?  ?
proposal for 8.4: PL/pgSQL - statement CASE 1 "Pavel Stehule" <pavel.stehule@gmail.com> ?  ?
psql slash# command 1 "Sibte Abbas" <sibtay@gmail.com> ?  ?
V0.1 patch for TODO Item: SQL-language reference 1 pgsql-patches-owner@postgresql.org ?  ?
Truncate Triggers 1 Simon Riggs <simon@2ndquadrant.com> Greg Smith  ?

Greg Smith:

Just discussion here, patch exists later in this list
Maintaining cluster order on insert 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Heikki Linnakangas <heikki@enterprisedb.com> ?  ?
Reducing Transaction Start/End Contention 1 2 Alvaro Herrera <alvherre@commandprompt.com> ?  ?
Add function for quote_qualified_identifier? 1 2 "Brendan Jurd" <direvus@gmail.com> ?  ?
Auto-explain patch 1 2 3 4 5 David Fetter <david@fetter.org> ?  ?
SET TRANSACTION conformance to SQL:2003 1 Simon Riggs <simon@2ndquadrant.com> Tom Lane  ?

Tom Lane:

Oh, here's the patch. But see my comment above --- I'm not sure I buy the premise.
Fragments in tsearch2 headline 1 "Sushant Sinha" <sushant354@gmail.com> ?  ?
V0.1 patch for TODO Item: SQL-language reference parameters by name. 1 "Gevik Babakhani" <pgdev@xs4all.nl> ?  ?
Visibility map thoughts 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
proposal, plpgsql, 8.4, for record in cursor 1 "Pavel Stehule" <pavel.stehule@gmail.com> ?  ?
Text <-> C string 1 2 3 4 5 6 7 8 Gregory Stark <stark@enterprisedb.com> ?  ?
[FEATURE REQUEST] Streaming Onlinebackup (Maybe 1 2 3 4 5 6 7 8 9 "Kevin Grittner" <Kevin.Grittner@wicourts.gov> Tom Lane  ?

Bruce Momjian:

This is about truncating archived WAL files by removing the empty space at the end. I think it needs to be a separate /contrib directory because it can be used for archiving and standby modes. In fact I imagine it is more useful for archiving than standby because standby is just going to discard the WAL logs once they are processed.

Tom Lane:

What are we thinking of doing with this? Contrib module? Maybe dump it in with pg_standby?
Minor changes to Recovery related code 1 "Simon Riggs" <simon@2ndquadrant.com> Tom Lane  ?

Bruce Momjian:

Simon, which of these were not done for 8.3 and need TODO entries?

Tom Lane:

No patch here.
WIP: rewrite numeric division 1 Tom Lane <tgl@sss.pgh.pa.us> Tom Lane  ?

Tom Lane:

Not sure why we didn't just apply this in 8.3. The last couple of messages contemplate a feature addition (integer division operator) that's really a separate issue, but beyond that I'm not seeing any alternative choices.
Proposed patch: synchronized_scanning GUC 1 2 Alvaro Herrera <alvherre@commandprompt.com> Greg Smith  ?

Greg Smith:

No patch here, is TODO: Remove pgdump -i (Ignore version mismatch) and only allow cases where the dumping version is compatible with the database.
WIP: rewrite numeric division 1 2 3 4 5 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
pg_next_dst_boundary optimization 1 2 Heikki Linnakangas <heikki@enterprisedb.com> ?  ?
Updated bitmap index patch 1 Gavin Sherry <swm@alcove.com.au> ?  ?
Array behavior oddities 1 Bruce Momjian <bruce@momjian.us> Tom Lane  ?

Tom Lane:

I see no TODO here, unless possibly clarify the documentation.
POSIX shared memory support 1 2 3 Chris Marcellino <cmarcellino@apple.com> Bruce Momjian  ?

Michael Renner:

I think the main benefit is freeing people from having to manually increase the amount of memory available for sysv shared memory. This can be even more annoying when running in container-based virtualization since you've to increase the sysv shm pages in the guest/instance _and_ on the host OS. Largely cosmetic, but probably one less hurdle to stumble over for new users and helpful in constrained environments.

Tom Lane:

It's certainly far from clear what the point is...

Bruce Momjian:

Did we decide we don't want this?
actualized SQL/PSM patch 1 "Pavel Stehule" <pavel.stehule@gmail.com> Bruce Momjian  ?

Bruce Momjian:

So I guess this is the final patch version.
postgresql-icu patch status 1 Ireneusz Pluta <ipluta@wp.pl> Bruce Momjian  ?

Martijn vO:

We've been here before: the comparison routine can take iterators. [quote]ICU4C's ucol_strcollIter API allows for comparing two strings that are supplied as character iterators (UCharIterator). This is useful when you need to compare differently encoded strings using strcoll. In that case, converting the strings first would be probably be wasteful, since strcoll usually gives the result before whole strings are processed.[/quote] http://www.icu-project.org/userguide/Collate_API.html

Tom Lane:

I agree that it's unlikely we'd write our own locale support; but it also seems unlikely that we'll buy into ICU's limitations (in particular the there-is-no-god-but-UTF16 angle). Whatever your thoughts on that argument, you must realize that this isn't going to be resolved during this commit fest. Please push to the TODO list.

Bruce Momjian:

I think it is extremely unlikely that we will write collation support on our own so if we want to solve this, I think we have to use ICU.

Martijn vO:

Right now ICU is mentioned in the TODO under COLLATE support. Perhaps it changed into: - Allow collation per collumn (SQL COLLATE) - Write custom collation library providing same features as ICU There's no point adding ICU support itself to the TODO since even optional support has been rejected in the past. I don't think this patch has any real future, unfortunatly.

Bruce Momjian:

The patch that never dies.  ;-) TODO again?
Auto Partitioning Patch - WIP version 1 1 NikhilS <nikkhils@gmail.com> ?  ?
[PATCH] automatic integer conversion 1 "Merlin Moncure" <mmoncure@gmail.com> Tom Lane  ?

Tom Lane:

Associated with the PQParam threads...
CopyReadLineText optimization 1 "Heikki Linnakangas" <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

Looks like a TODO entry to me.

Bruce Momjian:

Yep. Do we keep it here or just delete it?

Tom Lane:

Consensus is that this still needs work.

Andrew Dunstan:

this is different

Tom Lane:

I think Andrew applied this --- or was that a different patch?
Maintaining cluster order on insert 1 Heikki Linnakangas <heikki@enterprisedb.com> ?  ?
Add function for quote_qualified_identifier? 1 "Brendan Jurd" <direvus@gmail.com> Tom Lane  ?

Tom Lane:

Ah, I'd forgotten that a patch was indeed submitted. But do we want it? Reviewing the original thread, nobody except Brendan seemed to want it.

Bruce Momjian:

I was missing the actual patch; I have added it to the queue.

Tom Lane:

Not a patch, just a TODO proposal that I find pretty uninteresting.
[FEATURE REQUEST] Streaming Onlinebackup (Maybe 1 "Kevin Grittner" <Kevin.Grittner@wicourts.gov> ?  ?
Remove pg_dump -i option (was Re: Proposed patch: 1 Alvaro Herrera <alvherre@commandprompt.com> ?  ?
Some ideas about Vacuum 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 "Gokulakannan Somasundaram" <gokul007@gmail.com> ?  ?
proposed patch for function parameters name refs 1 "Gevik Babakhani" <pgdev@xs4all.nl> ?  ?
Show INHERIT in \du 1 "Brendan Jurd" <direvus@gmail.com> ?  ?
Integer datetimes 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Ordered Append Node 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Configurable Penalty Costs for Levenshtein 1 Volkan YAZICI <yazicivo@ttnet.net.tr> Bruce Momjian  ?

Tom Lane:

Seems to have a few bugs ... I don't think the offered correction is self-consistent.

Bruce Momjian:

This patch looks good.
New style of hash join proposal 1 Gregory Stark <stark@enterprisedb.com> ?  ?
Free Space Map thoughts 1 Heikki Linnakangas <heikki@enterprisedb.com> Alvaro Herrera  ?

Bruce Momjian:

I assume we will review this as part of the dead space man visibility TODOs.

Alvaro Herrera:

There is no patch here -- just a discussion on possible future development.
Visibility map thoughts 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 "Gokulakannan Somasundaram" <gokul007@gmail.com> ?  ?
Auto Partitioning Patch - WIP version 1 1 NikhilS <nikkhils@gmail.com> ?  ?
EXECUTE USING for plpgsql (for 8.4) 1 2 3 4 "Heikki Linnakangas" <heikki@enterprisedb.com> ?  ?
Reference by in \d <table_name> out 1 "kenneth d'souza" <kd_souza@hotmail.com> Tom Lane  ?

Tom Lane:

This was resubmitted (in a saner format) below.
[8.4] Updated WITH clause patch (non-recursive) 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Fix pg_dump dependency on postgres.h 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Text <-> C string 1 "Brendan Jurd" <direvus@gmail.com> Tom Lane  ?

Tom Lane:

Looks like text_to_cstring et al have a plurality. I'll get started on this now.

Bruce Momjian:

Yea, this is being discussed on hackers now.

Tom Lane:

Need to pick different names for the functions...
V0.2 patch for TODO Item: SQL-language reference parameters by name. 1 "Pavel Stehule" <pavel.stehule@gmail.com> ?  ?
PGparam proposal v2 1 Andrew Chernow <ac@esilo.com> ?  ?
Configurable Penalty Costs for Levenshtein 1 Volkan YAZICI <yazicivo@ttnet.net.tr> ?  ?
PGparam proposal 1 2 3 4 Andrew Chernow <ac@esilo.com> ?  ?
Sorting Improvements for 8.4 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
WIP: plpgsql source code obfuscation 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 "Dave Page" <dpage@postgresql.org> ?  ?
Some ideas about Vacuum 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Auto-explain patch 1 Dean Rasheed <dean_rasheed@hotmail.com> Greg Smith  ?

Greg Smith:

Interesting patch, but without any examples of what it does I can't figure out if it's outputting in a useful format. Could probably use some review from somebody like Guillaume Smet to make sure the resulting logs will parse well, and that they'll come out right in syslog/CVSlog formats as well.
proposal for 8.4: PL/pgSQL - statement CASE 1 2 3 4 5 6 Josh Berkus <josh@agliodbs.com> ?  ?
PGparam timestamp question 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Reference by output in : \d <table_name> 1 "kenneth d'souza" <kd_souza@hotmail.com> Bruce Momjian  ?

Bruce Momjian:

Ah, this is a patch for a new TODO entry.
Add function for quote_qualified_identifier? 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Proposed patch for ANALYZE overcounting dead rows 1 Tom Lane <tgl@sss.pgh.pa.us> Tom Lane  ?

Tom Lane:

Note that the next thread discusses changes to this patch --- I'm not at all sure we should apply it as-is.
[DOCS] Full Text Search examples patch 1 "Erik" <er@xs4all.nl> Tom Lane  ?

Tom Lane:

The proposed patch seems fundamentally misguided to me --- a one-liner example query is not expected to have complete sample data underlying it. Else we have a whole lot of work to do in a whole lot of places. Perhaps there's a TODO here to improve the tsearch docs, but I'm not sure how to phrase it.
Visibility map thoughts 1 Heikki Linnakangas <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

Although there's no ready-to-review patch here, this whole area is something that I'd like to see discussed more, so we should probably keep it in the to-look-at-during-commit-fest list.
Reference by output in : \d <table_name> 1 2 "kenneth d'souza" <kd_souza@hotmail.com> ?  ?
[FEATURE REQUEST] Streaming Onlinebackup (MaybeOFFTOPIC) 1 2 3 4 Heikki Linnakangas <heikki@enterprisedb.com> ?  ?
WIP: variadic function, named params 1 "Pavel Stehule" <pavel.stehule@gmail.com> ?  ?
PGparam extension version 0.4 1 Andrew Chernow <ac@esilo.com> ?  ?
[FEATURE REQUEST] Streaming Onlinebackup (MaybeOFFTOPIC) 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
{**Spam**} Re: Proposed patch: synchronized_scanning GUC variable 1 Tom Lane <tgl@sss.pgh.pa.us> Bruce Momjian  ?

Bruce Momjian:

I have submitted a patch for this.
Dead Space Map version 3 (simplified) 1 2 3 ITAGAKI Takahiro <itagaki.takahiro@oss.ntt.co.jp> ?  ?
Sorting Improvements for 8.4 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 Jeff Davis <pgsql@j-davis.com> ?  ?
Bitmapscan changes - Requesting further feedback 1 Heikki Linnakangas <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

No patch here either.
PGparam timestamp question 1 2 3 4 5 6 Robert Treat <xzilla@users.sourceforge.net> ?  ?
Patch for pg_dump 1 "Dany DeBontridder" <dany118@gmail.com> Bruce Momjian  ?

Bruce Momjian:

Yea, I guess we need to give the author feedback.

Tom Lane:

It may be a TODO, but we don't want *this* patch. Have you read it?

Bruce Momjian:

This is a TODO so we probably want this patch.
proposal, plpgsql, 8.4, for record in cursor 1 2 3 Peter Eisentraut <peter_e@gmx.net> ?  ?
GIT patch 1 2 3 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
[DOCS] Full Text Search examples patch 1 2 3 4 "Erik" <er@xs4all.nl> ?  ?
Including Snapshot Info with Indexes 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 "Gokulakannan Somasundaram" <gokul007@gmail.com> ?  ?
Array behavior oddities 1 2 3 Bruce Momjian <bruce@momjian.us> ?  ?
SHA1 on postgres 8.3 1 2 3 4 Bruce Momjian <bruce@momjian.us> Tom Lane  ?

Bruce Momjian:

I thought we are considering moving the one-way (hash) encryption functions into the core, but leave the others in pgcrypto.

Tom Lane:

I thought the conclusion of this thread was that we weren't changing anything.
WIP: rewrite numeric division 1 2 3 4 5 6 7 Gregory Stark <stark@enterprisedb.com> ?  ?
PGparam proposal 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
script binaries renaming 1 2 3 4 5 6 7 8 9 10 11 12 13 Magnus Hagander <magnus@hagander.net> ?  ?
proposal, plpgsql, 8.4, for record in cursor 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Proposal: PL/pgSQL EXECUTE INTO USING (for 8.4) 1 "Pavel Stehule" <pavel.stehule@gmail.com> Tom Lane  ?

Tom Lane:

Still another update posted 2008/03/24

pavel stehule:

This older patch was deprecated. I sent new variant http://momjian.us/mhonarc/patches/msg00359.html

Tom Lane:

There's no patch here.
Friendly help for psql 1 2 3 4 5 6 7 8 Peter Eisentraut <peter_e@gmx.net> ?  ?
Proposal: real procedures again (8.4) 1 "Pavel Stehule" <pavel.stehule@gmail.com> Alvaro Herrera  ?

Tom Lane:

We probably need to look it over again to be sure there's consensus on the development direction.

Alvaro Herrera:

No patch here -- just discussion and a couple of TODO items.
V0.2 patch for TODO Item: SQL-language referenceparameters by name. 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Better psql tab-completion support for schemas and tables 1 Greg Sabino Mullane <greg@turnstep.com> ?  ?
updated hash functions for postgresql v1 1 2 3 4 5 6 7 8 Simon Riggs <simon@2ndquadrant.com> ?  ?
Implemented current_query 1 2 3 4 Neil Conway <neilc@samurai.com> ?  ?
Indexam interface proposal 1 Heikki Linnakangas <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

No patch here.
VACUUM/ANALYZE counting of in-doubt tuples 1 2 3 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Incomplete docs for restore_command for hot standby 1 "Markus Bertheau" <mbertheau.pg@googlemail.com> Tom Lane  ?

Bruce Momjian:

Yep, Simon?

Tom Lane:

Patch is clearly not ready to apply as-is. Seems to me this is Simon's turf...
Key for grant attributes (was: Re: Show INHERIT in \du) 1 Decibel! <decibel@decibel.org> ?  ?
Incomplete docs for restore_command for hot standby 1 "Markus Bertheau" <mbertheau.pg@googlemail.com> ?  ?
Stream bitmaps 1 Heikki Linnakangas <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

No patch here.
Dead Space Map version 3 (simplified) 1 2 3 4 5 6 "Pavan Deolasee" <pavan.deolasee@gmail.com> ?  ?
Integer datetimes 1 2 3 4 5 6 Peter Eisentraut <peter_e@gmx.net> ?  ?
V0.1 patch for TODO Item: SQL-language reference parameters by name. 1 2 3 "Pavel Stehule" <pavel.stehule@gmail.com> ?  ?
BUG #3387: mod on non-integer returns bad result 1 2 Gregory Stark <stark@enterprisedb.com> ?  ?
bitmapscan changes patch review 1 Alexey Klyukin <alexk@commandprompt.com> ?  ?
CopyReadLineText optimization 1 "Heikki Linnakangas" <heikki@enterprisedb.com> ?  ?
WIP: plpgsql source code obfuscation 1 2 3 4 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
WIP: plpgsql source code obfuscation 1 "Pavel Stehule" <pavel.stehule@gmail.com> Bruce Momjian  ?

Bruce Momjian:

Do we want a TODO for this?
V0.1 patch for TODO Item: SQL-language reference parameters 1 Andrew Dunstan <andrew@dunslane.net> ?  ?
Index Tuple Compression Approach? 1 2 3 4 Heikki Linnakangas <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

This is discussion related to the GIT patch, but there's no independent patch here.
Sorting Improvements for 8.4 1 Simon Riggs <simon@2ndquadrant.com> Bruce Momjian  ?

Bruce Momjian:

Is there a TODO in here somewhere?
V0.2 patch for TODO Item: SQL-language reference 1 David Fetter <david@fetter.org> ?  ?
Implemented current_query 1 Tomas Doran <bobtfish@bobtfish.net> Tom Lane  ?

Tom Lane:

The reason we haven't had complaints is that it's not exposed ...

Bruce Momjian:

You mean the issue with multiple commands separated by semicolons? Well, I wonder if we should change debug_query_string to actually print the executing statement now that they are treated as separate transactions. Or we could just add this SQL function and document it --- it is not like we have had tons of complaints about the current debug_query_string behavior.

Tom Lane:

The fundamental issue here is whether we really want to expose the current warty behavior of debug_query_string at the SQL level.
[PATCH] Add size/acl information when listing databases 1 2 3 4 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
VACUUM/ANALYZE counting of in-doubt tuples 1 2 3 4 5 6 7 Russell Smith <mr-russ@pws.com.au> ?  ?
PQParam version 0.5 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
V0.2 patch for TODO Item: SQL-language reference parameters by name. 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Show INHERIT in \du 1 "Brendan Jurd" <direvus@gmail.com> Greg Smith  ?

Tom Lane:

Brendan will submit a new version of this patch.

Tom Lane:

Ooops ... this idea won't work, unless we can solve how to localize the output. See my message just now.

Greg Smith:

Looks useful and ready for potential commit review, it's a pretty simple patch. My vote would be for !login instead of no login BTW.
V0.1 patch for TODO Item: SQL-language reference parameters by name. 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
V0.2 patch for TODO Item: SQL-language reference parameters by name. 1 "Gevik Babakhani" <pgdev@xs4all.nl> ?  ?
VACUUM/ANALYZE counting of in-doubt tuples 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Jagged Rows (was Re: Proposal: real procedures again 1 David Fetter <david@fetter.org> ?  ?
Dead Space Map version 3 (simplified) 1 ITAGAKI Takahiro <itagaki.takahiro@oss.ntt.co.jp> Bruce Momjian  ?

Bruce Momjian:

Being able to move these emails around in groups is certainly a capability I would like to have.

Tom Lane:

Well, I guess if we process everything else the space-map-related threads will end up together ...

Bruce Momjian:

No, I don't know how to group threads together because the web pages are rebuilt everytime the patch archive changes. I would have to save them in a separate mbox file and have a separate URL.

Tom Lane:

Well, it's certainly not getting committed in this form, but given all the open issues it seems like it's review material. Can we somehow group together the other related threads? DSM, FSM, known-frozen pages, etc all need to be considered together IMHO.

Bruce Momjian:

What are we doing with all these dead space emails? Is this just a TODO?
Ordered Append Node 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 Markus Schiltknecht <markus@bluegap.ch> ?  ?
Reducing Transaction Start/End Contention 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Proposal: real procedures again (8.4) 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 Gregory Stark <stark@enterprisedb.com> ?  ?
[DOCS] Full Text Search examples patch 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Concurrent psql patch 1 2 3 4 5 6 7 Gregory Stark <stark@enterprisedb.com> Bruce Momjian  ?

Tom Lane:

Greg asked for review on the lists today, so I guess we do need to look at it ...

Tom Lane:

Is there any useful review to be done during this commit fest? If not, it doesn't belong in this list.

Bruce Momjian:

This is not ready for application. Do we just add it to TODO?
Integer datetimes 1 Neil Conway <neilc@samurai.com> Bruce Momjian  ?

Tom Lane:

We agreed to switch configure's default for 8.4; I'm not convinced about anything longer-term than that. Neil, are you going to do the honors?

Bruce Momjian:

I think we agreed to make this change for 8.4.
script binaries renaming 1 2 3 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Text <-> C string 1 "Brendan Jurd" <direvus@gmail.com> ?  ?
[FEATURE REQUEST] Streaming Onlinebackup 1 2 3 4 5 6 Simon Riggs <simon@2ndquadrant.com> ?  ?
PQParam version 0.5 1 2 3 4 5 Alvaro Herrera <alvherre@alvh.no-ip.org> ?  ?
Text <-> C string 1 2 3 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Array behavior oddities 1 2 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Maintaining cluster order on insert 1 Heikki Linnakangas <heikki@enterprisedb.com> Tom Lane  ?

Tom Lane:

This is superseded by a later submission.
Friendly help for psql 1 Greg Sabino Mullane <greg@turnstep.com> Greg Smith  ?

Greg Smith:

Patch is simple and looks handy, but could use a quick revision based on the feedback received
Kludge in pg_standby.c 1 Gregory Stark <stark@enterprisedb.com> Tom Lane  ?

Magnus Hagander:

I'll add it to my list.

Tom Lane:

Somebody in the Windows crowd needs to look at this ...
Friendly help for psql 1 2 3 4 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Show login privilege in psql \du command 1 Bernd Helmle <mailings@oopsware.de> Greg Smith  ?

Bruce Momjian:

Right, I just left it here for reference.

Greg Smith:

There is a patch with a superset of this feature later in the queue http://momjian.us/mhonarc/patches/msg00024.html
TODO item: add \# which lists line numbers, and allows command execution 1 "Sibte Abbas" <sibtay@gmail.com> Tom Lane  ?

Tom Lane:

Background for patch in next thread.
pg_next_dst_boundary optimization 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Ordered Append WIP patch v1 1 Gregory Stark <stark@enterprisedb.com> ?  ?
pg_standby observation 1 Jeff Davis <pgsql@j-davis.com> Tom Lane  ?

Tom Lane:

No patch, just a possible TODO about improving pg_standby's logging.
Truncate Triggers 1 Simon Riggs <simon@2ndquadrant.com> Bruce Momjian  ?

Bruce Momjian:

Smells like a patch to me.  ;-)
Reviewing new index types (was Re: Updatedbitmap indexpatch) 1 "Simon Riggs" <simon@2ndquadrant.com> ?  ?
Proposal - libpq Type System beta-0.8a (was PGparam) 1 Andrew Chernow <ac@esilo.com> ?  ?
Show INHERIT in \du 1 2 3 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Consistent \d commands in psql 1 Greg Sabino Mullane <greg@turnstep.com> Alvaro Herrera  ?

Alvaro Herrera:

There is a useful patch here.
BUG #3387: mod on non-integer returns bad result 1 2 3 Tom Lane <tgl@sss.pgh.pa.us> Tom Lane  ?

Tom Lane:

There's no usable patch here --- it's just background for the rewrite numeric division thread below.
stored procedure stats in collector 1 2 3 4 5 6 7 Neil Conway <neilc@samurai.com> Alvaro Herrera  ?

Tom Lane:

I'm still desperately unhappy with putting this into ExecMakeFunctionResult; I do not believe for a moment that the overhead will be unnoticeable.

Alvaro Herrera:

Author will be submitting new patch. This should probably be punted to the next commit fest. http://archives.postgresql.org/pgsql-hackers/2008-03/msg00727.php

Alvaro Herrera:

There is a patch here, and it seems consensus on it was reached by the end of the thread, but the patch needs to be updated for 8.4 (enum GUC for one thing).
Key for grant attributes (was: Re: Show INHERIT in \du) 1 "Brendan Jurd" <direvus@gmail.com> ?  ?
Transaction Snapshot Cloning 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Chris Browne <cbbrowne@acm.org> Bruce Momjian  ?

Bruce Momjian:

Is this a TODO?
PGparam timestamp question 1 Andrew Chernow <ac@esilo.com> ?  ?
Show INHERIT in \du 1 2 3 4 5 Bernd Helmle <mailings@oopsware.de> ?  ?
Fix pg_dump dependency on postgres.h 1 Zdenek Kotala <Zdenek.Kotala@Sun.COM> Alvaro Herrera  ?

Alvaro Herrera:

There is a useful patch here.
Fix pg_dump dependency on postgres.h 1 2 3 4 5 6 7 8 9 10 11 Alvaro Herrera <alvherre@commandprompt.com> ?  ?
plpgsql and qualified variable names 1 Bruce Momjian <bruce@momjian.us> Tom Lane  ?

Tom Lane:

This should not be in the queue; it was already done in 8.3.
Proposal: PL/pgSQL EXECUTE INTO USING (for 8.4) 1 2 3 4 "Merlin Moncure" <mmoncure@gmail.com> ?  ?
Transaction Snapshot Cloning 1 2 3 4 5 6 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Reviewing new index types (was Re: Updated 1 Heikki Linnakangas <heikki@enterprisedb.com> ?  ?
GIT patch 1 2 3 4 5 6 Heikki Linnakangas <heikki@enterprisedb.com> ?  ?
EXECUTE USING for plpgsql (for 8.4) 1 "Pavel Stehule" <pavel.stehule@gmail.com> Zdenek Kotala  ?

Alvaro Herrera:

New patch is here: http://archives.postgresql.org/pgsql-patches/2008-03/msg00310.php

Tom Lane:

Latest update posted 2008/03/24

Zdenek Kotala:

Pavel, I cannot apply this patch. Can you please send updated version with HEAD?
PGparam proposal 1 Andrew Chernow <ac@esilo.com> Andrew Chernow  ?

Bruce Momjian:

Yes, I am keeping these older emails for background material.

Andrew Chernow:

This is a very out-dated version and should only be viewed for historical background. The latest version of the libpq type system is 0.9a and can be found here: http://archives.postgresql.org/pgsql-patches/2008-03/msg00057.php It includes full documentation, implementation and regression test.
Reviewing new index types (was Re: Updatedbitmap indexpatch) 1 "Simon Riggs" <simon@2ndquadrant.com> ?  ?
[8.4] Updated WITH clause patch (non-recursive) 1 2 3 4 5 6 7 Gregory Stark <stark@enterprisedb.com> ?  ?
Reducing Transaction Start/End Contention 1 "Simon Riggs" <simon@2ndquadrant.com> Tom Lane  ?

Bruce Momjian:

I have asked Simon if this is a TODO.

Tom Lane:

There's no patch here, and even the TODO proposals need to be rethought in the light of Florian Pflug's later work on reducing ProcArray contention.
Proposal - libpq Type System beta-0.8a (was PGparam) 1 Andrew Chernow <ac@esilo.com> ?  ?
Proposal - libpq Type System beta-0.8a (was PGparam) 1 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Maintaining cluster order on insert 1 2 3 4 5 6 7 8 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
TODO item: add \# which lists line numbers, and allows command execution 1 "Sibte Abbas" <sibtay@gmail.com> ?  ?
suggested wording improvement in psql 1 "Harald Armin Massa" <haraldarminmassa@gmail.com> ?  ?
plpgsql, forc statement 1 "Pavel Stehule" <pavel.stehule@gmail.com> ?  ?
Truncate Triggers 1 2 3 4 5 6 7 Tom Lane <tgl@sss.pgh.pa.us> ?  ?
Free Space Map thoughts 1 2 3 4 5 6 7 8 9 10 11 12 13 14 Simon Riggs <simon@2ndquadrant.com> ?  ?

Personal tools