libgcj maintenance for 4.0.x branch

Sat Apr 16 07:46:00 GMT 2005

On Fri, Apr 15, 2005 at 12:44:34PM -0700, Per Bothner wrote:
> Anthony Green wrote:
> >Michael Koch has volunteered to maintain the GCC 4.0.x branch for
> >libgcj. This basically involves moving approved patches from HEAD back
> >to the branch.
> >It would be good to set up some kind of policy for what patches should
> >be acceptable. Perhaps something like this already exists.
> >My suggestion is simply to make the golden rule:
> >"Programs built against libgcj 4.0.a should link correctly against
> >libgcj 4.0.b where a<b."

Anthony: Thanks for starting this thread.
We need be carefully with updating as already mentioned in this thread.
Breaking compatibility inside the 4.0.x release series is a non-option
to me. But getting fixes from HEAD gives us the possibility to point our
users to a stable release instead of pointing them to CVS HEAD all time,
as we did in the past after 3.4.0 release.
> I think 4.0.x is meant for small safe bug-fixes and improvements.
> So I think it is primarily the Release manager's job to set policy.
> What you're saying is a necessary condition for including a patch,
> but I'm not sure it's sufficient.

The release manager has shown that he thinks that java is not critical
for GCC and we are on our own when we break it.
With 3.4 release series we made the fault to nearly never fix bugs in
it. And we never ported back new stuff which people asked for. There
often people who say: we cant use 3.4 because its too old. They then use
one of the other free alternatives because they update more frequently.
I think not improving a release branch means stepping backward all the
time another free alternative does a new release.
Michael
-- 
Escape the Java Trap with GNU Classpath!
http://www.gnu.org/philosophy/java-trap.html
Join the community at http://planet.classpath.org/