Final call for testing of notmuch 0.3.1

[Carl Worth]

On Tue, 27 Apr 2010 10:36:44 -0700, Carl Worth <cworth at cworth.org> wrote:
> I feel comfortable with what's in place in master now for the 0.3.1
> release. But I'd like to wait for a couple of hours (at least) this time
> to ensure nobody notices any last-minute bug that should be fixed.
> 
> So please test out the code in master, especially the new features
> and bug fixes mentioned below, and let me know if you find anything that
> should be fixed.

When writing the above, I had meant to ask specifically if people
thought it was a good idea to include a couple of new features in this
0.3.1 release or if it should be restricted to bug fixes only.

Even without me asking that, I got feedback from three primary notmuch
contributors, (David E., Dirk, and Jesse), who all recommended that it
should be a bug-fix only release.

So I'm convinced and have now setup a "0.3.x" branch from which the
release will be made. It was formed by cherry-picking from master, but
omitting the commits that were described as "features" in the release
notes, (new hook for washing inline patches and auto-tagging of replied
messages), as well as the 2 or 3 commits to fix up those features.

If people could give the 0.3.x branch a careful look and give me a
"thumbs up" on it being ready-to-go that would be appreciated. In
particular, now would be a great time for people who care about building
notmuch on platforms other than Linux to speakup.

In the future, I think I'd like to do this sort of "fallout release" on
master rather than a side branch. (And do side branches only for when
fixing up some release long after master has moved on.)

Thanks to everyone for helping us figure out how to make this project
work.

-Carl

PS. Here are the new (still preliminary) release notes for 0.3.1

General bug fix
---------------
Fix an infinite loop in "notmuch reply"

  This bug could be triggered by replying to a message where the
  user's primary email address did not appear in the To: header and
  the user had not configured any secondary email addresses. The bug
  was a simple re-use of the same iterator variable in nested loops.

Emacs bug fixes
---------------
Fix calculations for line wrapping in the primary "notmuch" view.

Fix Fcc support to prompt to create a directory if the specified Fcc
directory does not exist.

Build fix
---------
Fix build on OpenSolaris (at least) due to missing 'extern "C"' block.

  Without this, the C++ sources could not find strcasestr and the
  final linking of notmuch would fail.


-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: not available
URL: <http://notmuchmail.org/pipermail/notmuch/attachments/20100427/cb3a7d11/attachment.pgp>