Re: September 2012 commitfest

The September 2012 commitfest has been running now for six weeks, and
we've been making some progress. That stalled a bit last week due to
the European conference; hopefully there are now some renewed energies
to let us finish it up soon.

One thing I do *not* want to do is send many patches to the next
commitfest. That will only make the final commitfests worse by piling
up stuff no one seems to care about (except their submitter). So if
you're able to help, please do.

There are now 65 patches listed; 20 of them have been committed so far,
and 22 have been returned with feedback. 1 was sent to the next
commitfest ("lock_timeout").

Waiting on Author: 1
Needs Review: 10
Ready for Committer: 7

One patch is "Waiting on Author" ("Incorrect behaviour when using a GiST
index on points"). This is a bug fix and so requires special
dedication, and backpatch. One of the GiST hackers should really get to
it soon, please.

Ten patches are in "Needs Review" state.

- From Amit Kapila:
* Performance Improvement in Buffer Management for Select operation
Jeff Janes has been looking at this. I think we should close it for
now. More performance figures are needed.

* Patch to compute Max LSN of Data Pages
Apparently it's difficult to get people excited about this. There was
no discussion.

- From Alexander Korotkov:

* Decrease GiST bloat when penalties are identical
* Range Types statistics
* Adjacent in SP-GiST for range-types

I think these three patches are pretty close to ready, according to the
review comments. If somebody can give them a final look to commit this
week, that'd be great. May need a final revision to be sent.

* 2d-mapping based GiST for ranges
This one doesn't seem to have gotten any review. I would like to send
this one to the next commitfest.

- Other guys:
* FOR KEY SHARE foreign keys (me)
This one is seeing some more discussion lately. I'm not closing it
just yet mainly to avoid killing the discussion.

* [PoC] Writable Foreign Tables (KaiGai Kohei)
This needs to be examined by someone with executor insight.

* Skip checkpoint on promoting from streaming replication (Kyotaro Horiguchi)
Simon is going to handle this at some point; I will punt to next CF.

Additionally, seven patches are "Ready for Committer":
* Updatable views
Tom said he was going to handle this one

* tuplesort memory usage: grow_memtuples
Greg Stark signed up for this

* Trim trailing NULL columns
Josh Berkus was going to do performance testing, but if he published
anything I can't find it. Robert said, in the previous commitfest,
that if the benchmarks were right then "this patch is ready to go in".
It's been long since any committer weighed in on this thread, though.

* Make pg_basebackup configure and start standby
Magnus signed up for this

* Fix console prompt encoding on Windows
Noah thinks this is pretty much ready to commit.
No one signed up, but a Windows committer needed. Magnus, Andrew?

* parallel pg_dump
Andrew Dunstan said he was going to handle this.

* plpgsql_check_function
Selena submitted an re-indented copy of the last one submitted by
Pavel but apparently something was wrong about what she did; got
advice about it but didn't get around to submitting another version.
There was another thread around message
4F7C7346(dot)2090407(at)enterprisedb(dot)com and it doesn't look like Selena
followed that one. I'm not sure where this patch stands, really; the
history needs to be checked carefully to ensure what Pavel last
submitted is in line with the review in the previous discussions.

There are 163 emails in three threads, each starting at
CAFj8pRDkkzSi611Eimp=AXj2HD46k-W46GDvW9MKAD2OgwoKag(at)mail(dot)gmail(dot)com
1330807185-sup-2696(at)alvh(dot)no-ip(dot)org
CAFj8pRAYVTQYCL8_NF_hDQjc0m+JBvbwR6E_ZJ0SJfkKQ9m2kA(at)mail(dot)gmail(dot)com

I vaguely recall Greg Stark signed up for another patch recently, but I
can't readily find which one it was.

--
Álvaro Herrera http://www.2ndQuadrant.com/
PostgreSQL Development, 24x7 Support, Training & Services