why bugs only fixed in trunk?

[Stephen J. Turnbull]

Leo writes:

 > One of the incentives to report bugs in the pretest is to have them
 > squashed before the release. Unfortunately it seems most bugs are just
 > fixed in trunk (which could mean wait another year for the fix) even if
 > reported against a pretest build.
 > 
 > Is there an implicitly agreed practice here?

Emacs is following best practice, AFAICS.  The only thing I personally
would do differently is to maintain a bugfix-only branch for the
previous release, but that wouldn't make a difference to your
situation.

If you want the most recent bug-fixes, the answer is obvious: use
trunk.

 > Anyway, I think I am discouraged to report bugs and inclined to just
 > code a workaround in my .emacs.

That seems self-defeating to me.  Not going to special effort to test
pre-releases would actually save you time, effort, and aggravation.

But apparently you're talking about not reporting bugs in a version
you are going to be using anyway.  It would take only a minute to
write a function that would copy the region containing the workaround,
open a bug report buffer, yank the copy into the bug report, and send.
It would take a minute per workaround to write a comment explaining
what infelicity is being addressed, making the .emacs code itself a
self-explanatory bug report.

 > Long ago I used to report bugs in the org mode that was bundled with
 > emacs but then the project's practice was to only fix in the development
 > trunk.

Completely different case, though.  It makes sense for a downstream
distributor to simply hand such bug reports back to upstream, and
resync when upstream releases.  Too much divergence of downstream from
any upstream branch just makes life a lot harder for everybody, to fix
a few bugs that don't matter much to most users, who would prefer an
earlier release and sync.