This is the mail archive of the
java-patches@gcc.gnu.org
mailing list for the Java project.
Re: Two more 4.1 branch patches (from 0.21-pre)
- From: Tom Tromey <tromey at redhat dot com>
- To: Mark Wielaard <mark at klomp dot org>
- Cc: java-patches at gcc dot gnu dot org
- Date: 06 Feb 2006 11:48:51 -0700
- Subject: Re: Two more 4.1 branch patches (from 0.21-pre)
- References: <1139091112.1962.33.camel@localhost.localdomain>
- Reply-to: tromey at redhat dot com
>>>>> "Mark" == Mark Wielaard <mark@klomp.org> writes:
Mark> Here are two more candidates for putting on the 4.1 branch. Again
Mark> small, unintrusive, have bug numbers associated with them and mauve
Mark> tests. OK?
Ok, thanks.
Mark> These are probably the last "easy" patches. There is of course a lot
Mark> more on the trunk and in classpath. In my eyes we should at least think
Mark> about the beans and regex updates. It looks like xml would be nice to
Mark> update also, but that is not a really clean patch since there has been a
Mark> lot of new development combined with bug fixing. There are also lots of
Mark> security, crypto, awt and swing updates, but those are so huge that it
Mark> only makes sense to take them all, or none at all imho.
Unfortunately given our current limitations (in particular, we aren't
yet ready to make the BC ABI the default), I think we simply have to
draw a line somewhere.
This is not the ideal situation, since Classpath updates all the time.
It sure would be great to be able to deliver updates on an ongoing
basis throughout the year. Maybe we can achieve this in the 4.2
cycle.
I looked at the regex fixes but put them off because I saw your
followup fix in URI, and assumed that there might be other internal
regex problems around.
Mark> So what are the "rules" for the 4.1 branch?
Right now, "don't mess things up", at least as far as I'm concerned.
We haven't had a more formal discussion.
After 4.1 is released I assume we will use the normal release branch
rules, i.e., don't break the (C++) ABI.
Personally I've focused largely on bugs needed to make things work for
FC5, with the occasional regression fix thrown in.
Tom