apr-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Nick Kew <n...@webthing.com>
Subject Re: Backport policy, WAS: Re: svn commit: r678140 - /apr/apr/branches/1.3.x/memory/unix/apr_pools.c
Date Sat, 19 Jul 2008 13:46:20 GMT
On Sat, 19 Jul 2008 14:18:11 +0200
Mladen Turk <mturk@apache.org> wrote:

> Sander Striker wrote:
> >>
> >> Modified:
> >>    apr/apr/branches/1.3.x/memory/unix/apr_pools.c
> > 
> > This is just an example, I'm seeing a change land on trunk and a
> > minute later on 1.3.x.  What's our backport policy?  Do we just
> > commit to trunk and 1.3.x, or are we going to go a route where we
> > get review prior to backporting?  Think httpd STATUS...
> > 
> 
> I usually wait for a day or two before committing to 1.3.x or 1.2.x
> from trunk. This particular one is copy/paste typo and was not
> observed until Bojan and myself did some actual hard core testing.

Which begs the question, how did the typo get in to 1.3?

I'm with Sander here: it is somewhat alarming to see things going
into should-be stable branches on little or no review.  That includes
things I've committed myself.

Thinking httpd STATUS, it can certainly get frustrating, but I find
it a useful check.  It helps document the review process, and means
I can take the time to review other people's work at my convenience
rather than (not) try to tackle every commit as it happens.

This seems to have been introduced in r647447, which was in /trunk/
at the time.  Thus it was never adequately reviewed.

-- 
Nick Kew

Application Development with Apache - the Apache Modules Book
http://www.apachetutor.org/

Mime
View raw message