Re: Changing WAL Header to reduce contention during ReserveXLogInsertLocation()

On Fri, Mar 23, 2018 at 11:06:48AM +0000, Simon Riggs wrote:
> Your assumption that I would commit a new patch that was 29 mins old
> is frankly pretty ridiculous, so yes, lets keep calm.

When a committer says that a patch is "ready for commit" and that he
calls for "last objections", I am understanding that you would be ready
to commit the patch from the moment such an email has been sent. Am I
the only one to think so? Now let's look at the numbers:
- The last patch sent is a v2, which implements a completely new
approach compared to v1. This is a non-trivial patch which touches
sensitive parts of the code.
- v2 has been sent exactly two weeks after the last email exchanged on
this thread.
- Per the data publicly available, it took less than 30 minutes to
review the patch, and there are zero comments about its contents.
I do patch review on a more-or-less daily basis, and look at threads on
hackers on a daily basis, but I really rarely see such an "efficient"
review pattern. You and Pavan have likely discussed the patch offline,
but nobody can guess what has been discussed and what have been the
arguments exchanged.

> Enjoy your weekend and I'll be happy to read your review on Monday.

Er. So this basically means that I need to do a commitment to look at
this patch in such a short time frame? If you are asking for reviews,
doing such requests by asking a proper question rather than by implying
it in an affirmation would seem more adapted to me, so this email bit is
making me uncomfortable. My apologies if I am not able to catch the
nuance in those words.
--
Michael