Bernhard Fischer - Re: Fortran patch committed during freeze (original) (raw)

This is the mail archive of the gcc-patches@gcc.gnu.orgmailing list for the GCC project.

• From: Bernhard Fischer
• To: Brooks Moses
• Cc: fortran at gcc dot gnu dot org, gcc-patches at gcc dot gnu dot org
• Date: Mon, 12 Feb 2007 13:30:42 +0100
• Subject: Re: Fortran patch committed during freeze
• References: <200702112108.l1BL8SfJ012805@sparrowhawk.codesourcery.com> <Pine.LNX.4.44.0702111911540.9990-100000@www.eyesopen.com> <45D0080D.8030707@codesourcery.com>

On Sun, Feb 11, 2007 at 10:24:13PM -0800, Brooks Moses wrote:

Roger Sayle wrote:

On Sun, 11 Feb 2007, Mark Mitchell wrote:

However, please be mindful of the release branch freeze state in future.

One comment I would make though, is how difficult it is to discover the 4.1 branch's current status. Neither the main page's "Current release status" at http://gcc.gnu.org/ nor the most recently linked to status e-mail from that page explicitly mention the/a freeze. When reviewing/approving a recent regression patch I had to infer that since there was mention of a release candidate but there hadn't been a full release, we were most likely frozen. The 4.0 branch, however, is clearly marked as frozen.

I'll further note that it would really be useful to cc the fortran email list on announcements of branch freezes and the like, given that many (probably most) of the Fortran maintainers and contributors don't monitor the patches and devel lists. Usually someone happens to notice and forwards a copy of the announcement over, but I think that got overlooked this time, with predictable results.

Just to put emphasis on whom to "blaim" for this oversight:

It was not me who checked that fix in but jerryd who backported it to 4.1 (with a bad timing, it seems).

IIRC there was a convention to list the person who initially prepared the patch as the first entry and then the actual committer:

date original <orig@> committer <blame_me@>

or something to that effect. The rules for expected changelog formats are a bit hidden, IIRC. Not sure if the convention on how to format authors for backports like in the above case is even listed there. Neither contribute.html nor http://www.gnu.org/prep/standards/standards.html#Change-Logs give examples for this particular case, AFAICS.

• Follow-Ups:
  • Re: Fortran patch committed during freeze
    * From: François-Xavier Coudert
• References:
  • Fortran patch committed during freeze
    * From: Mark Mitchell
  • Re: Fortran patch committed during freeze
    * From: Roger Sayle
  • Re: Fortran patch committed during freeze
    * From: Brooks Moses