On Sat, Mar 22, 2008 at 5:53 PM, Brian Dessent <brian@dessent.net> wrote:
The problem is that MinGW gcc releases contain a huge number of local
patches by Danny.  They are not just vanilla FSF releases.  The patches
fix a number of issues that aren't fixed upstream.  It's a ton of work
to maintain a local tree like that, and I wouldn't be surprised to find
that Danny finds it quite painful to make releases.

Of course, a great deal of the patchset in Danny's 4.2 release were
backports from 4.3: ...

So the real issue is not that FSF gcc doesn't build or doesn't work, but
rather that vanilla FSF gcc has never been stable enough to call it the
system compiler, for either MinGW or Cygwin.

My big question after reading this is this: if the patches are necessary to make GCC stable, why are they not in the upstream source repositories?  Is it that the FSF won't accept them for some reason, or that the MinGW and/or Cygwin communities haven't pushed hard enough for their inclusion?  In other words, are we just trying to get through a little red tape, or have we been flat-out denied?

-- Early Ehlinger, President, ResPower, Inc.