Re: Commitfest 2023-03 starting tomorrow! - Mailing list pgsql-hackers

Status summary:
  Needs review: 116.
  Waiting on Author: 30.
  Ready for Committer: 32.
  Committed: 94.
  Moved to next CF: 17.
  Returned with Feedback: 6.
  Rejected: 6.
  Withdrawn: 18.
Total: 319.



Ok, here are the patches that have been stuck in "Waiting
on Author" for a while. I divided them into three groups.

* The first group have been stuck for over a week and mostly look like
  they should be RwF. Some I guess just moved to next CF. But some of
  them I'm uncertain if I should leave or if they really should be RfC
  or NR.

* The other two groups have had some updates in the last week
  (actually I used 10 days). But some of them still look like they're
  pretty much dead for this CF and should either be moved forward or
  Rwf or Rejected.

So here's the triage list. I'm going to send emails and start clearing
out the patches pretty much right away. Some of these are pretty
clearcut.


Nothing in over a week:
----------------------

* Better infrastructure for automated testing of concurrency issues

- Consensus that this is desirable. But it's not clear what it's
  actually waiting on Author for. RwF?

* Provide the facility to set binary format output for specific OID's
per session

- I think Dave was looking for feedback and got it from Tom and
  Peter. I don't actually see a specific patch here but there are two
  patches linked in the original message. There seems to be enough
  feedback to proceed but nobody's working on it. RwF?

* pg_visibility's pg_check_visible() yields false positive when
working in parallel with autovacuum

- Bug, but tentatively a false positive...

* CAST( ... ON DEFAULT)

- it'll have to wait till there's something solid from the committee"
-- Rejected?

* Fix tab completion MERGE

- Partly committed but
  v9-0002-psql-Add-PartialMatches-macro-for-better-tab-complet.patch
  remains. There was a review from Dean Rasheed. Move to next CF?

* Fix recovery conflict SIGUSR1 handling

- This looks like a suite of bug fixes and looks like it should be
  Needs Review or Ready for Commit

* Prefetch the next tuple's memory during seqscans

- David Rowley said it was dependeny on "heapgettup() refactoring"
  which has been refactored. So is it now Needs Review or Ready for
  Commit? Is it likely to happen this CF?

* Pluggable toaster

- This seems to have digressed from the original patch. There were
  patches early on and a lot of feedback. Is the result that the
  original patches are Rejected or are they still live?

* psql - refactor echo code

- "I think this patch requires an up-to-date summary and explanation"
  from Peter. But it seems like Tom was ok with it and just had some
  additional improvements he wanted that were added. It sounds like
  this might be "Ready for Commit" if someone familiar with the patch
  looked at it.

* Push aggregation down to base relations and joins

- Needs a significant rebase (since March 1).

* Remove self join on a unique column

- An offer of a Bounty! There was one failing test which was
  apparently fixed? But it looks like this should be in Needs Review
  or Ready for Commit.

* Split index and table statistics into different types of stats

- Was stuck on "Generate pg_stat_get_xact*() functions with Macros"
  which was committed. So "Ready for Commit" now?

* Default to ICU during initdb

- Partly committed, 0001 waiting until after CF

* suppressing useless wakeups in logical/worker.c

- Got feedback March 17. Doesn't look like it's going to be ready this CF.

* explain analyze rows=%.0f

- Patch updated January, but I think Tom still had some simple if
  tedious changes he asked for

* Fix order of checking ICU options in initdb and create database

- Feedback Last November but no further questions or progress

* Introduce array_shuffle() and array_sample() functions

- Feedback from Tom last September. No further questions or progress



Status Updates in last week:
----------------------------

* Some revises in adding sorting path

- Got feedback Feb 21 and author responded but doesn't look like it's
  going to be ready this CF

* Add TAP tests for psql \g piped into program

- Peter Eisentraut asked for a one-line change, otherwise it looks
  like it's Ready for Commit?

* Improvements to Meson docs

- Some feedback March 15 but no response. I assume this is still in
  play



Emails in last week:
-------------------

* RADIUS tests and improvements

- last feedback March 20, last patch March 4. Should probably be moved
  to the next CF unless there's progress soon.

* Direct SSL Connections

- (This is mine) Code for SSL is pretty finished. The last patch for
  ALPN support needs a bit of polish. I'll be doing that momentarily.

* Fix alter subscription concurrency errors

- "patch as-submitted is pretty uninteresting" and "patch that I don't
  care much about" ... I guess this is Rejected or Withdrawn

* Fix improper qual pushdown after applying outer join identity 3

- Tom Lane's patch. Active discussion as of March 21.

* Error "initial slot snapshot too large" in create replication slot

- Active discussion as of March 24. Is this now Needs Review or Ready
  for Committer?

* Transparent column encryption

- Active discussion as of March 24

* Make ON_ERROR_STOP stop on shell script failure

- I rebased this but I think it needs a better review. I may have a
  chance to do that or someone else could. The original author
  bt22nakamorit@oss.nttdata.com seems to have disappeared but the
  patch seems to be perhaps committable?

* pg_stats and range statistics

- Updated patch as of March 24, should be Needs Review I guess?

* TDE key management patches

- Actively under discussion

* Reconcile stats in find_tabstat_entry() and get rid of
PgStat_BackendFunctionEntry

- Actively under discussion

--
Gregory Stark
As Commitfest Manager