matplotlib-devel Mailing List for matplotlib

Things are still very much in flux as we get our git legs -- so I didn't mean to imply any sort of "must do". It's not a problem to continue doing what you're doing, you just might see faster service by using github pull requests and the code review facility there is very cool. There's some information on this in the newly updated docs here:
http://matplotlib.github.com/devel/gitwash/patching.html
Cheers,
Mike
________________________________________
From: Christoph Gohlke [cg...@uc...]
Sent: Thursday, March 17, 2011 12:21 PM
Cc: matplotlib development list
Subject: Re: [matplotlib-devel] Patch for matplotlib-py3
I thought the workflow/process of contributing to matplotlib did not change:
<http://www.mail-archive.com/mat...@li.../msg07849.html>
<http://matplotlib.sourceforge.net/faq/howto_faq.html#contributing-howto>
Christoph
On 3/17/2011 6:40 AM, Michael Droettboom wrote:
> I have added these as pull requests to matplotlib/matplotlib-py3.
>
> To save us all time and make it easier to do code review, you may want to follow the instructions here to create your own pull requests:
>
> http://matplotlib.github.com/devel/gitwash/development_workflow.html
>
> Cheers,
> Mike
>
> ________________________________________
> From: Christoph Gohlke [cg...@uc...]
> Sent: Wednesday, March 16, 2011 3:01 PM
> To: matplotlib development list
> Subject: [matplotlib-devel] Patch for matplotlib-py3
>
> Hello,
>
> please consider the attached patch for the matplotlib-py3 fork on
> github. It corrects some build and test failures and removes unnecessary
> 'extern "C"' statements from PyMODINIT_FUNC functions.
>
> From<http://docs.python.org/release/2.6.6/extending/extending.html>:
> Note that PyMODINIT_FUNC declares the function as void return type,
> declares any special linkage declarations required by the platform, and
> for C++ declares the function as extern "C".
>
> Christoph
>
>
------------------------------------------------------------------------------
Colocation vs. Managed Hosting
A question and answer guide to determining the best fit
for your organization - today and in the future.
http://p.sf.net/sfu/internap-sfd2d
_______________________________________________
Matplotlib-devel mailing list
Mat...@li...
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

I thought the workflow/process of contributing to matplotlib did not change:
<http://www.mail-archive.com/mat...@li.../msg07849.html>
<http://matplotlib.sourceforge.net/faq/howto_faq.html#contributing-howto>
Christoph
On 3/17/2011 6:40 AM, Michael Droettboom wrote:
> I have added these as pull requests to matplotlib/matplotlib-py3.
>
> To save us all time and make it easier to do code review, you may want to follow the instructions here to create your own pull requests:
>
> http://matplotlib.github.com/devel/gitwash/development_workflow.html
>
> Cheers,
> Mike
>
> ________________________________________
> From: Christoph Gohlke [cg...@uc...]
> Sent: Wednesday, March 16, 2011 3:01 PM
> To: matplotlib development list
> Subject: [matplotlib-devel] Patch for matplotlib-py3
>
> Hello,
>
> please consider the attached patch for the matplotlib-py3 fork on
> github. It corrects some build and test failures and removes unnecessary
> 'extern "C"' statements from PyMODINIT_FUNC functions.
>
> From<http://docs.python.org/release/2.6.6/extending/extending.html>:
> Note that PyMODINIT_FUNC declares the function as void return type,
> declares any special linkage declarations required by the platform, and
> for C++ declares the function as extern "C".
>
> Christoph
>
>

I have added these as pull requests to matplotlib/matplotlib-py3.
To save us all time and make it easier to do code review, you may want to follow the instructions here to create your own pull requests:
http://matplotlib.github.com/devel/gitwash/development_workflow.html
Cheers,
Mike
________________________________________
From: Christoph Gohlke [cg...@uc...]
Sent: Wednesday, March 16, 2011 3:01 PM
To: matplotlib development list
Subject: [matplotlib-devel] Patch for matplotlib-py3
Hello,
please consider the attached patch for the matplotlib-py3 fork on
github. It corrects some build and test failures and removes unnecessary
'extern "C"' statements from PyMODINIT_FUNC functions.
 From <http://docs.python.org/release/2.6.6/extending/extending.html>:
Note that PyMODINIT_FUNC declares the function as void return type,
declares any special linkage declarations required by the platform, and
for C++ declares the function as extern "C".
Christoph

It turns out that it was also trivial to get the figure options editor 
working with PySide.
All that needed to be done was (in formlayout.py):
 * Replace the references to PyQt with PySide
 * Change: from PyQt.QtCore import (Qt, SIGNAL, SLOT, QSize, QString,
 pyqtSignature, pyqtProperty)
 to: from PySide.QtCore import (Qt, SIGNAL, SLOT, QSize, #QString,
 #pyqtSignature, pyqtProperty)
 Slot as pyqtSignature, Property as
 pyqtProperty)
 * Add the following after the import statements:
 # Hacks to emulate PyQt
 QString = str
 class QColorDialog(QColorDialog):
 @staticmethod
 def getRgba(color,parent):
 result =
 QColorDialog.getColor(QColor.fromRgba(color),parent,'')
 return result.rgba(),result.isValid()
Formlayout could probably be updated so it doesn't need the hacks but I 
wanted to keep changes to a minimum.
Gerald.
On 17/03/2011 11:21 AM, Gerald Storer wrote:
> On 01/18/2011 08:13 PM, Jed Ludlow wrote:
> > Please forgive me if I'm raising a heretical question with this since I
> > understand the topic of competing Qt bindings for Python gets a little
> > touchy in and of itself. Nonetheless, the elephant is in the room. I
> > searched the archives and found only a few comments on the subject:
> > mat...@li.../msg18652.html" 
> target="_new">http://www.mail-archive.com/mat...@li.../msg18652.html>
> > Has there been any additional discussion among the developers about
> > creating a formal backend for Pyside?
>
> Its actually a fairly easy to get the PyQt backend working with 
> PySide. It would have been laughably easy if not for a couple of bugs 
> in PySide that took awhile to track down.
>
> To get it working PySide working you need to:
>
> * Obviously replace the reference to PyQt with PySide
> * Remove the reference to PyQt/Pyside.Qt and Qt.qApp by replacing
> it with QtGui.qApp (I think this is already done in the most
> recent Git version)
> * Replace the toolbar message signal with a new style signal by:
> o adding 'message = QtCore.Signal(str)' to the
> NavigationToolbar2QT class definition
> o replacing: QtCore.QObject.connect(self.toolbar,
> QtCore.SIGNAL("message"),
> 
> self.window.statusBar().showMessage)
> with:self.toolbar.message.connect(self.window.statusBar().showMessage)
> o replacing: self.emit(QtCore.SIGNAL("message"), s)
> with: self.message.emit(s)
> * Work around the PySide bug with QImage and convert the string
> passed from the Agg backend into a python buffer with
> buffer(stringBuffer) or wait for PySide to fix bug 489.
> * Work around a PySide bug (738) by creating functions to perform
> the slider.setMaxiumum/setMinimum tasks or ignore the runtime
> errors for now and wait for a bug fix.
> * I haven't bothered with the figure options editor at this point
> - I just commented out the references to it.
>
> It might also be a good idea to convert all the signals/slots into the 
> new style but it seems to work just fine with only the above changes.
>
> Regards,
> Gerald.

On 01/18/2011 08:13 PM, Jed Ludlow wrote:
 > Please forgive me if I'm raising a heretical question with this since I
 > understand the topic of competing Qt bindings for Python gets a little
 > touchy in and of itself. Nonetheless, the elephant is in the room. I
 > searched the archives and found only a few comments on the subject:
 > mat...@li.../msg18652.html" 
target="_new">http://www.mail-archive.com/mat...@li.../msg18652.html>
 > Has there been any additional discussion among the developers about
 > creating a formal backend for Pyside?
Its actually a fairly easy to get the PyQt backend working with PySide. 
It would have been laughably easy if not for a couple of bugs in PySide 
that took awhile to track down.
To get it working PySide working you need to:
 * Obviously replace the reference to PyQt with PySide
 * Remove the reference to PyQt/Pyside.Qt and Qt.qApp by replacing it
 with QtGui.qApp (I think this is already done in the most recent
 Git version)
 * Replace the toolbar message signal with a new style signal by:
 o adding 'message = QtCore.Signal(str)' to the
 NavigationToolbar2QT class definition
 o replacing: QtCore.QObject.connect(self.toolbar,
 QtCore.SIGNAL("message"),
 
 self.window.statusBar().showMessage)
 with:self.toolbar.message.connect(self.window.statusBar().showMessage)
 o replacing: self.emit(QtCore.SIGNAL("message"), s)
 with: self.message.emit(s)
 * Work around the PySide bug with QImage and convert the string
 passed from the Agg backend into a python buffer with
 buffer(stringBuffer) or wait for PySide to fix bug 489.
 * Work around a PySide bug (738) by creating functions to perform
 the slider.setMaxiumum/setMinimum tasks or ignore the runtime
 errors for now and wait for a bug fix.
 * I haven't bothered with the figure options editor at this point -
 I just commented out the references to it.
It might also be a good idea to convert all the signals/slots into the 
new style but it seems to work just fine with only the above changes.
Regards,
Gerald.