[GRASS-dev] 6.4rc1

Glynn Clements glynn at gclements.plus.com
Tue Dec 2 05:09:47 EST 2008


Glynn Clements wrote:

> > > so what remains todo befor 6.4rc1? IMO lib API and module list should be
> > > frozen at that point, which means creating releasebranch_6_4. No need to
> > > wait on that anymore IMO. bugfixes can continue on until the end.
> > > if 6.4 is the last, that means for all of GRASS 6.x so last chances...
> > 
> > I also vote for creating releasebranch_6_4 within the next days...
> > 
> > what about
> > 
> > http://trac.osgeo.org/grass/ticket/72
> > 
> > and
> > 
> > http://trac.osgeo.org/grass/ticket/58
> > 
> > temporal solution would to include pseudodc.cpp and pseudodc.h to
> > gui/wxpython/vdigit and relay on the given version of wxWidgets.

If you want to use a local copy, you will probably have to rename the
classes to prevent conflicts.

> The ideal solution is to figure out how to call the wxPseudoDC
> method(s) via PyEval_CallObject(). I can't see that it could be so
> complex as to either hold up the release or to require a hackaround.

Scratch that. I had assumed that wxPseudoDC was a subclass of wxDC;
unfortunately it isn't. It's a distinct class which just happens to
provide the same methods as wxDC.

That's good enough for Python, but for C++ you can't just treat it as
a wxDC for most operations and call out to Python for the
wxPseudoDC-specific methods (SetId and SetIdBounds). You would have to
invoke *all* methods via Python to eliminate the direct _gdi_.so
dependency.

This probably explains why it isn't part of wxWidgets proper.

-- 
Glynn Clements <glynn at gclements.plus.com>


More information about the grass-dev mailing list