CommitFest 2008-05

This page contains historical information or deprecated articles.

⇐ Previous CommitFest	This CommitFest	Next CommitFest ⇒
2008-03	2008-05	2008-07

CommitFest: May 2008

This commitfest is now closed All items in this commitfest have been reviewed and actioned. If you have a new patch to submit, please add it at CommitFestOpen.

Pending patches

Patch	Status	Author	Reviewers
Map Forks	WIP	Heikki Linnakangas	Nobody
tgl says: newer version here
tgl says: my short review is here; isn't anyone else gonna comment?
libpq
libpq object hooks	WIP	Merlin Moncure, Andrew Chernow	Nobody
merlinm says: Initial review here. Status?
tgl says: close but not ready to commit now

Committed patches

Patch	Status	Author	Reviewers
Moving snapshot code around	Committed	Álvaro Herrera	Nobody
Suppress compiler warnings on mingw	Committed 2008-04-16	Takahiro Itagaki	Nobody
More datum-vs-pointer cleanups	Committed 2008-04-17	Gavin Sherry	Nobody
Make int8, float8, timestamp, timestamptz pass-by-value on 64-bit architectures	Committed 2008-04-20	Zoltán Böszörményi	Nobody
Improve shutdown during online backup	Committed 2008-04-23	Laurenz Albe	Simon Riggs
ADD COLUMN PRIMARY KEY bug	Committed 2008-04-25	Brendan Jurd	Nobody
generate_subscripts	Committed 2008-04-28	Pavel Stehule	Nobody
Remove typename field from A_Const	Committed 2008-04-29	Brendan Jurd	Nobody
Remove NONSEG code, add configurable BLCKSZ and segment size	Committed 2008-05-01	Zdenek Kotala	Nobody
separate psql commands from arguments	Committed 2008-05-02	Bernd Helmle	petere
Add table size to \d+	Committed 2008-05-02	Dickson dos Santos Guedes	Nobody
GUC parameter cursor_tuple_fraction	Committed 2008-05-02	Robert Hell	Nobody
RETURN QUERY EXECUTE	Committed 2008-05-02	Pavel Stehule	Nobody
pl/python OUT param support	Committed 2008-05-02	Hannu Krosing	Nobody
Add statement_timeout to pg_dump	Committed 2008-05-03	Joshua Drake	Nobody
Replace more manual text <-> cstring conversions	Committed 2008-05-04	Brendan Jurd	Nobody
pg_conf_load_time() patch	Committed 2008-05-04	George Gensure	Nobody
generate_series for Timestamp	Committed 2008-05-04	Hitoshi Harada	Nobody
Fix \dT enum in psql	Committed 2008-05-04	David Fetter	Nobody
Multiline privileges in \z	Committed 2008-05-04	Brendan Jurd	Nobody
libpq win32.mak fix	Committed 2008-05-05	Hiroshi Saito	mha Nobody
psql wraps at window width	Committed 2008-05-08	Bryce Nesbitt	Bruce Momjian
Proper handling of inherited CHECK constraints	Committed 2008-05-09	Nikhil Sontakke, Alex Hunsaker	Nobody
Snapshot registering and tracking	Committed 2008-05-12	Álvaro Herrera	Nobody
New printTable API	Committed 2008-05-12	Brendan Jurd	Nobody
Show role attributes in \du	Committed 2008-05-12	Brendan Jurd	Nobody
additional options for RAISE statement	Committed 2008-05-13	Pavel Stehule	Nobody
stored procedure stats in collector	Committed 2008-05-14	Martin Pihlak	Nobody
plpgsql CASE statement	Committed 2008-05-15	Pavel Stehule	Nobody
localization infrastructure for ecpg	Committed 2008-05-16	Euler Taveira de Oliveira	petere Nobody
shorten startup banner	Committed 2008-05-16	Joshua Drake	bmomjian Nobody
GIN partial match, tsearch prefix search	Committed 2008-05-16	Teodor Sigaev, Oleg Bartunov	Nobody
ALTER SEQUENCE RESTART, TRUNCATE TABLE with IDENTITY	Committed 2008-05-16	Zoltán Böszörményi	Nobody
lc_time and localized dates	Committed 2008-05-19	Euler Taveira de Oliveira	Nobody
libpq Mutex on Win32	Committed 2008-05-21	Andrew Chernow	mha Nobody

Returned with Feedback

Patch	Status	Author	Reviewers
pg_standby auxiliary command	Rejected	Christopher Johnson	Simon Riggs
simon says: capability already available - patch not needed in core
Fix Windows kludge in pg_standby	Analyzing...	Greg Stark	Simon Riggs Nobody
somebody says: Needs to be handled by someone who's Windows-knowledgeable
alvherre says: Magnus mentioned via IM that the kludge might have been working around the Windows stat() bug that Magnus just fixed; that needs to be investigated
simon says: The current code was put there by patch from Dave Page and was tested by him. AFAICS there is no bug and no patch, so I'm moving this off the queue until a patch arrives, recognising that it remains an open issue
dave says: The code works as-is, but I think Greg does have a valid point. I'm not sure this is the stat() bug however - that involved stat() returning old data, whereas what I saw when testing was the space for the file being pre-allocated and then the data being copied. The stat() bug would show a file size lower than the actual size as I understand it, this showed the full size before the copy completed.
Sorting writes during checkpoint	On hold pending testing	Takahiro Itagaki	Nobody
tgl says: updated patch here, but I need some performance proof to apply it
gsmith says: will take a look at quantifying potential improvement on another system in late May
SE-PostgreSQL	Needs major work	Kohei KaiGai	Nobody
alvherre says: this message updates the patches
tgl says: reviews here, here, here, and here
LDAPS support w/o TLS	Needs revision	Stephen Layland	Nobody
tgl says: I think this shouldn't use a GUC variable
mha says: I'm looking at a proposal that will change how pg_hba deals with the parameters, by name-values instead. Suggest we put this patch on hold until this proposal is ready, since it would require a different approach for the patch.
tgl says: Agreed --- in any case, the patch isn't getting committed as-is.
Add support for Renesas SuperH	Waiting on author	Nobuhiro Iwamatsu	Nobody
tgl says: Author has not responded to this complaint
Column Level Privileges	Returned for rewrite	Guodong Liu, Andrew Dunstan	Nobody
tgl says: unfortunately it still needs a lot of work
Auto Partitioning Patch	WIP version 1	Nikhil Sontakke	Nobody
tgl says: What this is really waiting on is a roadmap for future partitioning features. Maybe it's a compatible first step, or maybe not
somebody says: March 2008 pgsql-patches thread
tgl says: Since there's no more roadmap now than there was in March, we still don't know what to do with this patch.
wildspeed contrib module	Returned for rework	Teodor Sigaev, Oleg Bartunov	Nobody
tgl says: performance is just too awful at the moment, mainly because of index size
psql command aliases support	Returned for rework	Bernd Helmle	Nobody
tgl says: still has some bugs