Moving Forward with IcedTea

[Andrew Hughes]

----- Original Message -----
> * Andrew Hughes <ahughes at redhat.com> [2012-05-02 10:39]:
> > ----- Original Message -----
> > 
> > snip...
> > 
> > > 2.  IcedTea7 for u4.
> > > 
> > > Oracle have already released u4:
> > > http://www.oracle.com/technetwork/java/javase/7u4-relnotes-1575007.html
> > > 
> > > I propose to sync IcedTea7 to jdk7u4 b13 (the point u4 branched)
> > > and
> > > then create a 2.2 branch which can
> > > then be prepared and tested for a 2.2 release.
> > > 
> > > PROPOSED BRANCH DATE: 2nd of May, 2012.  13h00 UTC (tomorrow).
> > 
> > The forest branch is now available:
> > 
> > http://icedtea.classpath.org/hg/release/icedtea7-forest-2.2/
> > http://icedtea.classpath.org/hg/release/icedtea7-forest-2.2/corba
> > http://icedtea.classpath.org/hg/release/icedtea7-forest-2.2/jaxp
> > http://icedtea.classpath.org/hg/release/icedtea7-forest-2.2/jaxws
> > http://icedtea.classpath.org/hg/release/icedtea7-forest-2.2/jdk
> > http://icedtea.classpath.org/hg/release/icedtea7-forest-2.2/langtools
> > http://icedtea.classpath.org/hg/release/icedtea7-forest-2.2/hotspot
> > 
> > The corresponding IcedTea tree will be created in due course to
> > facilitate testing.
> 
> Shouldn't we have first merged u4 into HEAD and then branched? Now we
> have to do the work twice.. am I missing something?
> 

It's a three-stage process mirroring upstream, of which this is stage 1.

1.  As mentioned above, HEAD is merged up to jdk7u4 b13 prior to branch.
This is done.  Both HEAD and 2.2 are at jdk7u4 b13.  This is the majority
of u4 changes (most of which were in the earlier merge, not today's).

2.  The changes from the upstream jdk7u4 trees need to be pulled into the
2.2 forest.

3.  The changes from the upstream jdk7u trees need to be pulled into the
HEAD forest.

In theory, 2 is a subset of 3, but I'd prefer to mirror what Oracle did
upstream to avoid mistakes (and their branchpoints aren't exactly clear;
I had to ask to obtain the b13 branchpoint).

The important one is stage 2.  That will give us a 2.2 that should be
fairly close to u4.

Stage 3 will move HEAD onto u6 which isn't immediately urgent, but would
be good to get in sync so we and Oracle are both working on the same thing
and we can find bugs and get them fixed upstream.

You said you wanted to do some work on this merge, so I've left 2 & 3 for
you to do.  We can discuss this further on IRC but basically it's just a
cycle of:

1. hg pull http://hg.openjdk.java.net/jdk7u/jdk7u{,4}/{.,corba,jaxp,jaxws,jdk,langtools,hotspot}
2. hg merge
3. Fix any conflicts with hg resolve -l / hg resolve -m
4. Commit the merge: hg commit -m Merge
5. Build the forest when all trees are merged.
6. Push all trees if it builds.

4 and 5 are interchangeable.  I feel a little safer having the merge committed and just using
hg rollback if there's an issue, but the build shouldn't be making any changes to the source tree.

> Deepak
> 
> > --
> > Andrew :)
> > 
> > Free Java Software Engineer
> > Red Hat, Inc. (http://www.redhat.com)
> > 
> > PGP Key: 248BDC07 (https://keys.indymedia.org/)
> > Fingerprint = EC5A 1F5E C0AD 1D15 8F1F  8F91 3B96 A578 248B DC07
> > 
> 

-- 
Andrew :)

Free Java Software Engineer
Red Hat, Inc. (http://www.redhat.com)

PGP Key: 248BDC07 (https://keys.indymedia.org/)
Fingerprint = EC5A 1F5E C0AD 1D15 8F1F  8F91 3B96 A578 248B DC07