<<< Date Index >>>     <<< Thread Index >>>

Re: Mutt Next Generation



Re: Thomas Roessler in <20050126110628.GC4589@xxxxxxxxxxxxxxxxxxxxxxxxxxxx>
> Thanks for the heads-up; I hadn't seen this project before.

(NB: My personal impression was that taking mutt-1.5.6 instead of CVS
HEAD to start from wasn't the brightest idea there.)

> My initial reaction would be that I'd like to work out a way to
> avoid forking things while moving ahead with mutt development.

In my view, there are some points that should be resolved to foster
mutt development:

* the mutt BTS
  The previous BTS collapsed due to spam. There are several options:
    - living without a BTS
      -> bad
    - setting up a debbugs instance elsewhere
      -> requires a well-trained spam filter (we have the offer to get
      the rules Debian's BTS uses) and constant hand-cleaning for the
      spam that got through (Blars Blarson told us that he spends
      1/2-1 hour per day for the Debian BTS).
    - using GForge (either at SF.net, or elsewhere)
      -> I have no idea on how well that scales to 100s of bugs
    - using Bugzilla
      -> dito
    - using trac (a Wiki with project management facilities)
      -> suggested by Brendan Cully

    As promised on IRC, I'd be willing to host the new system and help
    maintaining it; including spam-cleaning the old reports and
    migrating them. I haven't yet had the time to look at the systems
    mentioned above, so if anyone has a suggestion on which one to
    use, I'd be happy to hear.

* frequent development releases
  It's been quite a time since the release of 1.5.6.
  Thomas, do you have anything that you want to do on CVS before
  releasing 1.5.7?

* getting mutt 1.6 out
  CVS HEAD has been quite stable over the last months, and
  distributors have started packaging that instead of 1.5 (or even
  1.4). IMHO HEAD could be released really soon, preferably shortly
  after 1.5.7.
  Thomas, anything you want to include in 1.6 that's not in CVS yet?

* including patches
  There are tons of mutt patches out there. See for example the list
  posted on mutt-dev recently:
  http://www.df7cb.de/www.df7cb.de/mutt/message/20041221.091949.42f233d6.en.html

  Adding (some of) those should be postponed to 1.7, imho.
  Which to include should be discussed on mutt-dev.

* more developers
  most bugs in the current bug list are probably trivial to fix, but
  that's too much for one person to handle. By giving more people CVS
  access, development would gain a new pace. If there's a development
  team, the others will review the changes made and can comment. (It's
  a development branch, so occasional breakages are not a regression,
  but a sign of progress.)

  Again, this should be for 1.7 if we consider 1.5 to be frozen.

* manual restructuring
  The current manual lists all options alphabetically. By
  reordering/grouping, manual.txt would become more readable. (Rado
  has probably more to say about this.)

  Again, 1.7.


On a side note, since I could never get used to mhonarc archives, I've
set up a web archive using lurker for some mutt-related lists at
www.df7cb.de/mutt/ . It's somehow still in a test phase, so please
report any breakage there to me. Many thanks to those (notably Dave)
who provided their private archives for that back in 2003 (it took me
that much time to setup it since I didn't have a decent machine to run
it on until lately).

Christoph
-- 
cb@xxxxxxxx | http://www.df7cb.de/

Attachment: signature.asc
Description: Digital signature