From: Sylvain Joyeux <sylvain.joyeux@...4x.org>
Date: 2007-06-20T16:11:35+09:00
Subject: Re: Question about the patchlevel release cycle

> could you refer to bug #s?

> > 1.8.6 thread support was broken in bad ways. It stayed for three
> > months
  unfortunately, I don't find the bugreport. It has been discussed first on 
  ruby-core and then opened (as far as I remember)
  See ruby-core:10594

> > 1.8.6-p36 is broken w.r.t. compilation of extensions and has a
> > thread bug.
  bug #11507
  there has been no bugreport (that I know of) for the compilation problem: 
  it has been fixed in SVN before anybody sent a report.
  See ruby-core:11499

> > Wouldn't it be possible to have "hotfixes" (i.e.: quick patchlevel
> > releases) when this kind of bugs happen ?
The problem is: how to make sure that a patched version is quickly released 
when Ruby is in a bad shape ? Some obvious bugs are in general fixed in 
svn before having a bugreport, ... and we can't assume that the release 
manager can follow all the commits in *both* ruby_1_8 and ruby_1_8_6

Moreover, maybe you need a process (tagging the commit logs, ...) to make 
sure that the bugfixes in ruby_1_8 that should go into 1_8_6 *do* go into 
the 1.8.6 branch (I'm thinking about bug #9376 for instance, which missed 
the p36 release. This one is quite minor, so that's not that much of a 
problem)

btw, neither the english version ruby-lang.org nor the project page on 
rubyforge mention p36.
-- 
Sylvain Joyeux