I built a brand new clean java6 tree on Windows x86 and ran the full
test suite. Initially I had 9 failures and 1 error (in awt, swing and
net), but when I go and run those failing test cases standalone there is
only 1 test I see consistently failing:
javax.swing.text.html.HTMLDocument_Reader_ActionsTest - testHarmony_4582
fails with a NullPointerException and testHarmony_4615 fails with a test
assertion.
This failure looks exactly like the problem we had on the java5 branch
recorded in HARMONY-6392, so maybe the fix for this issue has not been
merged into the java6 branch. I've tagged the JIRA with a target of
java6 M1 so we keep track of this issue and I'll take a look and see if
I can figure out why it's still failing.
Apart from that we actually look in good shape on this branch from these
test results.
Regards,
Oliver
On 21/01/2010 13:58, Oliver Deakin wrote:
On 20/01/2010 11:18, Mark Hindess wrote:
In
message<4d9fb1a01001200154u780476d6k570d6bb3516ba...@mail.gmail.com>,
Ray
Chen writes:
Hi Mark,
I have a question, what conditions should java6 satisfy so that we can
release it?
We'll follow the same process as we do for java5/trunk.
All the tests passed? :)
That would be nice, but we don't hold the java5 releases to that
standard so I see no reason to apply it to java6.
Agreed - I think getting the M1 for java6 into a position where we can
agree that the failing tests are not blockers for release is really
what we're aiming for. I've just had a look in JIRA at the bugs
targeted for 6.0M1 and I only see 1 result, HARMONY-6330. One thing
that I think is important for us to get the M1 release out the door is
to have a realistic set of bugs in JIRA to represent the current state
of the build and I don't think that is true at the moment. I'm
rerunning the 6.0 tests on the latest federated build - I'll post the
results here once I have them and I aim to raise JIRAs for all the
failures (and fix some if I can) so we have a reasonable view of what
is required to reach M1.
Regards,
Oliver
Regards,
-Mark
On Wed, Jan 20, 2010 at 5:47 PM, Mark Hindess
<mark.hind...@googlemail.com> wrote:
Since we are not making much progress towards a java6 milestone
release
and the code has been frozen for much too long, I think we should
abandon the idea and postpone the release until we do the next java5
milestone.
If no one speaks up against this, then I will send a mail declaring
the
code freeze over tomorrow.
I'm not convinced having the milestones offset by a week really helps
much and suggest that next time we do a milestone release for java5 we
do a java6 release at the same time. =A0(I'm not convinced this
will work
either but it can't really be worse so I think we should try it.)
What do you think?
-Mark.
--=20
Regards,
Ray Chen
--
Oliver Deakin
Unless stated otherwise above:
IBM United Kingdom Limited - Registered in England and Wales with number 741598.
Registered office: PO Box 41, North Harbour, Portsmouth, Hampshire PO6 3AU
