Thu 10 May 2012 08:33:41 PM UTC, comment #10:
I checked in the following change:
http://hg.savannah.gnu.org/hgweb/octave/rev/13cc11418393
If you notice any new problems with incorrect filling on resizing, please open a new bug report.
|
Fri 27 Apr 2012 05:28:21 AM UTC, comment #9:
I don't know why the push did not work anyway,
if it is not the correct fix I'm glad it didn't ;)
c.
|
Fri 27 Apr 2012 01:39:56 AM UTC, comment #8:
It doesn't look like you actually pushed the changeset.
If you did not, please don't push it, because it is not the rigth fix. If you did push it, please undo it.
I have a fix for this problem, but it is not complete yet. I'll try to finish and push it soon. Sorry about the failing test.
|
Thu 26 Apr 2012 03:11:14 PM UTC, comment #7:
OK, pushed.
c.
|
Thu 26 Apr 2012 01:55:10 PM UTC, comment #6:
Looks like a bad merge, Carlo.
Push the fix. I was talking with jwe in #octave about this diagonal matrix problem. I guess something went wrong during the merge.
|
Thu 26 Apr 2012 12:48:09 PM UTC, comment #5:
Ben,
I don't understand the code myself either, I just noticed
that when transplanting the changeset from stable to default
only half of the patch went through.
As the change was done by JWE I'd rather wait for him to comment before pushing.
c.
|
Thu 26 Apr 2012 11:59:42 AM UTC, comment #4:
The patch works for me.
I'm not familiar with this part of the code, so I'm hesitant to push.
Carlo, do you think it is safe to push?
|
Thu 26 Apr 2012 11:18:51 AM UTC, comment #3:
after more careful inspection I noticed that with
http://hg.savannah.gnu.org/hgweb/octave/rev/3a9a56999ce5
only part of the changes introduced in
http://hg.savannah.gnu.org/hgweb/octave/rev/5bb5fcffa29d
were transferred from stable to default:
the new tests were added to data.cc but the
rest of the modifications to handle cell array
args in diag were not.
the attached patch fixes the test failures for me.
(file #25742)
|
Thu 26 Apr 2012 10:30:11 AM UTC, comment #2:
I see now that the tests were introduced with changeset 5bb5fcffa29d so I don't think this should be marked as a regression anymore.
|
Thu 26 Apr 2012 10:26:19 AM UTC, comment #1:
oops, I referred to the wrong changeset, I have to roll-back to before
to make the failures go away.
Still, this changeset is just a merge of stable to default
so the responsible changeset is one of the previous ones committed to stable.
|
Thu 26 Apr 2012 10:09:09 AM UTC, original submission:
With current HEAD I see the following test failures in data.cc
with "make check":
If I roll-back to before this changeset:
the same tests do not fail, therefore I am marking this as a regression.
|