JDK 7 Updates: Policy Changes

[Andrew Hughes]

----- Original Message -----
> Hi Dalibor,
> 
> Thanks for the responses! I have a few follow on questions that mostly
> stem from my unfamiliarity with JIRA.
> 
> * dalibor topic <dalibor.topic at oracle.com> [2015-07-22 13:05]:
> > On 20.07.2015 17:36, Omair Majid wrote:
> > >Could be there be conflicts would Open and
> > >Proprietary bugs? If so, any tips to minimize those conflicts?
> > 
> > An interesting potential source of conflict is that hgupdater creates
> > backports entries in JBS. So you want to ensure that for jdk7u forests it's
> > configured in a way that lets you easily distinguish issues.
> > 
> > That's why we had this thread back in March:
> > http://mail.openjdk.java.net/pipermail/jdk7u-dev/2015-March/010279.html
> 
> What about manual OpenJDK7-specific issues? Can we create those
> manually?

I don't follow this. Are you saying something will happen when e.g.
JDK-8043202 from the current security update [0] is pushed to the OpenJDK 7
tree?

> 
> > For example, there is no special representation of the milestones like
> > Rampdown Phase 2 in the JDK JBS Project. Instead, they were tracked
> > separately - see https://wiki.openjdk.java.net/display/jdk7u/JDK+7u80 for
> > an
> > example.
> 
> Thank you, this clarifies things. Is there a way to say "We expect this
> bug to be resolved in OpenJDK 7u99" in JIRA? Bugzilla calls this "Target
> Milestone" [1].

There is in JIRA, generally; we already do this in the OpenJDK 6 JIRA. [1]

> 
> > That way, when an issue X is marked as Fixed in Version OpenJDK 7, Resolved
> > in Build b85, you could distinguish it from one that is marked as fixed in
> > version 7u85, Resolved in Build b01 (as would for example be the case for
> > issues addressed in Oracle's JDK).
> 
> Can we mark an issue as Fixed in Version OpenJDK 7, Resolved in Build
> 7u85? I expect future "releases" of OpenJDK7u to continue using the 7uXY
> versioning scheme that has been used in the past (I don't know if we
> will use many build numbers, however).

We're using OpenJDK 7u85 for the new update. I've currently bumped the
build numbers only when a new merge to IcedTea is required, so I don't
expect them to reach double digits unless we use them in a different
way.

IcedTea 2.6.1 [0] is based on 7u85 b01. There was a b00, but a
couple of additional issues (OPENJDK7-04 and OPENJDK7-05)
were found, resulting in b01 being tagged and merged.

For the bug database, I'd need to be able to target bugs to a particular
version so we can see easily what needs to be fixed in the update and not
miss anything. This is how we currently operate with both IcedTea Bugzilla
and (to a lesser extent) the OpenJDK 6 JIRA. We also use tracker bugs in
IcedTea [2] but this is not essential.

The OpenJDK 6 project was largely winding down before the change in
leadership. If we expect the OpenJDK 7 project to be more active than
OpenJDK 6 has been, then the bug database will be more active in turn.
If it is going to be used for more than just quarterly security updates
(I've seen suggestions on this list of moving IcedTea patches into it
and having it replace that repository [3] ), then it's essential that
we can point people to the bug database to file bugs. As far as I'm aware,
the OpenJDK bug database is only accessible to OpenJDK committers, so this
would be a blocker to using it for the OpenJDK 7 project.


> 
> Thanks,
> Omair
> 
> [1] https://bugzilla.readthedocs.org/en/latest/using/understanding.html
> 
> --
> PGP Key: 66484681 (http://pgp.mit.edu/)
> Fingerprint = F072 555B 0A17 3957 4E95  0056 F286 F14F 6648 4681
> 

[0] http://bitly.com/it20601
[1] https://java.net/jira/browse/OPENJDK6/fixforversion/17281/
[2] http://icedtea.classpath.org/bugzilla/show_bug.cgi?id=1698
[3] http://mail.openjdk.java.net/pipermail/jdk7u-dev/2015-July/010325.html
-- 
Andrew :)

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

PGP Key: ed25519/35964222 (hkp://keys.gnupg.net)
Fingerprint = 5132 579D D154 0ED2 3E04  C5A0 CFDA 0F9B 3596 4222

PGP Key: rsa4096/248BDC07 (hkp://keys.gnupg.net)
Fingerprint = EC5A 1F5E C0AD 1D15 8F1F  8F91 3B96 A578 248B DC07