You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
{{ message }}
Repository navigation
macosx backend broken by #901: QuadMesh fails so colorbar fails #996
backend_macosx.py and _macosx.m need to be updated to handle the QuadMesh change in which edgecolors are passed in place of showedges. As it is, it fails with an exception. This is illustrated by any attempt to make a colorbar, for example.
We fixed the issue causing the exception by updating _macosx.m to expect a PyObject * instead of an int. See commits 02f8a65 and 3c7b5ee.
However, the colorbar is still drawn blank. Does anyone have any tips as to where the colobar drawing code should be for the OSX renderer? I'd like to fix the whole issue before submitting a pull request.
We fixed the issue causing the exception by updating _macosx.m to expect a PyObject * instead of an int. (See commits 02f8a65 and 3c7b5ee.)
It requires much more than that; Michiel de Hoon, the macosx maintainer,
is going to take care of it shortly.
However, the colorbar is still drawn blank. Does anyone have any tips as to where the colobar drawing code should be for the OSX renderer? I'd like to fix the whole issue before submitting a pull request.
The problem is not with the colorbar code, it is a matter of fixing
macosx correctly, so please just wait for Michiel's work.
Reply to this email directly or view it on GitHub: #996 (comment)
Reply to this email directly or view it on GitHub: #996 (comment)
Thank you! It works for me, so I committed it; I did not see any reason
to wait for more review, since we don't seem to have many people looking
at this code.
I started a new issue, however, because both before and after this
change, compilation yields warnings. Maybe not a big deal, but it is
nicer to see a perfectly clean compilation.
Hi Eric,
These compilation warnings only occur with Mac OS X 10.6 or later. I can look at this later this week when I get access to a mac with a recent OS.
-Michiel
On Sun, Jul 15, 2012 3:39 PM EDT Eric Firing wrote:
Reply to this email directly or view it on GitHub: #996 (comment)
Thank you! It works for me, so I committed it; I did not see any reason
to wait for more review, since we don't seem to have many people looking
at this code.
I started a new issue, however, because both before and after this
change, compilation yields warnings. Maybe not a big deal, but it is
nicer to see a perfectly clean compilation.
Activity
oxling commentedon Jul 10, 2012
We fixed the issue causing the exception by updating _macosx.m to expect a PyObject * instead of an int. See commits 02f8a65 and 3c7b5ee.
However, the colorbar is still drawn blank. Does anyone have any tips as to where the colobar drawing code should be for the OSX renderer? I'd like to fix the whole issue before submitting a pull request.
efiring commentedon Jul 10, 2012
On 2012/07/10 9:25 AM, Amy wrote:
It requires much more than that; Michiel de Hoon, the macosx maintainer,
is going to take care of it shortly.
The problem is not with the colorbar code, it is a matter of fixing
macosx correctly, so please just wait for Michiel's work.
mdehoon commentedon Jul 15, 2012
See Pull Request #1011
Merge pull request #1011 from mdehoon/fix_for_issue_996
efiring commentedon Jul 15, 2012
On 2012/07/15 3:22 AM, mdehoon wrote:
Thank you! It works for me, so I committed it; I did not see any reason
to wait for more review, since we don't seem to have many people looking
at this code.
I started a new issue, however, because both before and after this
change, compilation yields warnings. Maybe not a big deal, but it is
nicer to see a perfectly clean compilation.
#1013
Eric
mdehoon commentedon Jul 16, 2012
Hi Eric,
These compilation warnings only occur with Mac OS X 10.6 or later. I can look at this later this week when I get access to a mac with a recent OS.
-Michiel
On Sun, Jul 15, 2012 3:39 PM EDT Eric Firing wrote:
WeatherGod commentedon Jul 17, 2012
Should this issue be closed?