Re: GCC 3.0 Release Criteria

[Jeffrey A Law <law at cygnus dot com>]

  In message <200005040953.LAA33819@max.zk-i.med.uni-muenchen.de>you write:
  > > Err, no.  egcs-1.1.2 was a bugfix release to egcs-1.1.1, in no way was it
  > > specifically slated to support Linux.  If you don't know something for a
  > > fact, do not state it as a fact, especially on a public list.
  > 
  > Maybe Charles misremembered and meant egcs-1.0.2, not 1.1.2?  _That_
  > really was an example of a release whose purpose you declared as `to
  > fix problems for RedHat and that's it'.  (So, not just `Linux only'
  > but even RedHat only..)  This provoked at least some discussion, see
  > http://gcc.gnu.org/ml/gcc/1998-04/msg00923.html and followups.
It is possible that he confused it with egcs-1.0.3 which was a release
thatprimarly  fixed a set of bugs that were preventing major Linux vendors
from switching from gcc-2.7.2 to egcs as their system compiler.  Note that
even though the bugs were things that showed up on Linux systems initially,
all but one of the changes also benefited non-Linux systems.


  > Fortunately, since then this has remained the only such `slated'
  > release AFAICS.
If something like this happens again, I wouldn't consider it bad.  We made a
strategic decision to spin a release that would have the effect of drastically
increasing the user base for egcs.

I see the ability and willingness to do that as a positive, not a negative and
I'm certainly willing to consider doing it again.

jeff