Re: [Mingw-users] Bug #127047 is *still* eating my balls!

[<eg...@of...>]

Well, I can answer some of my own questions.  You're right that many
components in Mumit's patch have not been applied to the gcc tree.
(Why not?  It's been a year.  Maybe they've only been rolled into the
development version which is heading for 3.0?) 

However, the patch does not apply cleanly to 2.95.3-prerelease -- far from
it.  Dozens upon dozens of hunks are rejected.  Some small parts of the 
patch *have* been applied, apparently.  A lot of this can be fixed by a
manual application process, but a few things look a little trickier.

I don't think I'm qualified to sort out the mess (though I can try, but I
am not a gcc hacker), so I'm back where I started, asking the questions:

* Will mingw32 support gcc 2.95.3 any time soon?  (No, I don't mean binaries!)

* Can I accelerate this process by doing work myself, or by paying someone
  else (in cash or beer or ...) to do this work?

* Is it easier to try to selectively apply the sjlj-exceptions patch
  from 2.95.3 to the 2.95.2 source tree which *is* supported by mingw32?

* Is there some Master Documentation somewhere that I'm missing which
  explains the current state of mingw32, how it relates to the various
  versions of gcc (2.95.2 stable, 2.95.3 prerelease, various Cygwin
  releases, development snapshots, ...), what patches are necessary, etc.?

The "ate my balls" bug in question is a showstopper for us (without it,
exceptions are pretty much worthless), and we rely on mingw32 to generate 
code for Windows that isn't burdened by the cygwin nightmare.

Dan