This is my mutt build cocktail status file.  Don't expect much, but it
might give you some clues :-)


Wed Feb 27 08:05:55 EST 2002
- Created status file


Wed Feb 27 08:09:18 EST 2002
- I had gone to rebuild 1.3.27 with new pgp patches but also included
  DGC's attach patch bugfixed for .23 -- which does not patch cleanly.
  Ah, well.


Wed Mar 13 09:44:49 EST 2002
- 1.3.28 released; begging feature patch maintainers for updates


Mon Mar 18 15:33:58 EST 2002
- I finally have 1.3.28 built.  Everything except ats.mark_old works, or
  at least compiles.  Now to test with it...


Wed Mar 20 11:43:30 EST 2002
- whoops; including Cedric's patch-1.3.28.cd.pattern_broken.1 causes my
  send-hooks to not fire; From: is not set and $editor is not unset in
  at least two cases.  It ain't just that patch, either, 'cuz they work
  fine when this patch is applied alone to bare source. *sigh*

- so it looks like it's down to one of dgc's patches. i think it might
  be isalias but i am not sure.  actually, i think it is MORE THAN one
  of the patches; i believe trying attach and pattern_broken was fine

- hokay, that's it. dgc.isalias, dgc.attach, and cd.pattern_broken all 
  combine to cause a problem with send-hook and my_hdr. rebuilding the
  clean tree without pattern_broken until this is resolved


Thu Mar 21 11:58:20 EST 2002
- yahoo; cedric has given me a tweaked version of pattern_broken that
  works just fine. new Clean builds built.

- there's still that annoying recipient-as-default-subject bug that i
  have track down; it appears to be in the patch2 set. maybe later...


Thu Mar 21 12:35:14 EST 2002
- yeah, it seems to be in patch-1.3.23.bj.my_hdr_subject.1 so let's see
  if I can track down byrial for an updated version...


Fri Mar 22 15:25:09 EST 2002
- nothing from byrial, so i will just take it out for now. too bad.


Thu Apr  4 09:16:14 EST 2002
- I have noticed problems with mutt's memory footprint on a large folder
  and have been tracking it.  The problem looks to be in patch2; could it
  be Byrial's hash_destroy patch, perhaps?  Checking (but it takes a
  couple of days for the problem to manifest itself...


Fri Apr  5 08:03:54 EST 2002
- Well, it appears to be bj.hash_destroy, all right. running a fully
  patched version without that is actually consistently SMALLER than the
  stock version and certainly smaller than the other increasing-levels
  cocktail versions...  Maybe Byrial can provide some input (haven't
  heard from him in a long time, though)


Fri Apr 12 13:09:45 EDT 2002
- hmmm... my =F.mutt process had a res size of only 12M but had a virtual
  size of 150M!  what gives?

- no update from Byrial after his mutt-users post providing a patch patch
  and promising to update his web site -- but he hasn't. applied the fix
  and recompiled (Clean/*-patched-with-hash_destroy-fix*) and now running
  with that one.


Tue May 14 07:40:05 EDT 2002
- 1.3.99 builds cleanly, and i have added DGC's fmtpipe patch to allow
    fcc-save-hook . "$HOME/.mutt/fcc-save-pipe.sh %O |"
  to handle tmda addresses (and anything else). stay tuned for a full
  build set.


Thu May 30 13:48:14 EDT 2002
- whipping up 1.4 Clean set; yahooie!


Mon Jul 22 04:59:05 EST 2002
- so now I'm on a freebsd server and have had to rebuild everything. it
  works; yay.  oh, yeah; this update is three weeks late :-)
- interesting that the freebsd regexp library does not like
    set alternates="(pattern|)..."
  but requires
    set alternates="(pattern)?..."
  for any expression (send-hook, folder-hook, fcc-save-hook, alternates,
  and so on). lots of muttrc rewriting; phooey.