[IPython-user] IPython 0.7.0 is out.

Charlie Moad cwmoad at gmail.com
Tue Jan 10 12:55:28 CST 2006


There was also a DeprecationWarning before we reverted as well.  In
the current release I don't get that warning, and the mpl/pylab
doesn't busy wait.  What does that change accomplish?

On 1/10/06, Fernando Perez <Fernando.Perez at colorado.edu> wrote:
> Hi Sean,
>
> Sean Moore wrote:
> > Fernando,
> >
> > the patch / correction I sent you for pygtk on windows threaded coded didn't
> > make it into the release:
> >
> > in Shell.py  line 626 (sorry - don't have patch)  You have:
> >
> >         if self.gtk.pygtk_version >= (2,4,0):
> >              import gobject
> >              gobject.timeout_add(self.TIMEOUT, self.on_timer)
> >          else:
> >              self.gtk.timeout_add(self.TIMEOUT, self.on_timer)
> >
> >
> > but I have this:
> >
> >          if self.gtk.pygtk_version >= (2,4,0):
> >              import gobject
> >              gobject.idle_add(self.on_timer)
> >          else:
> >              self.gtk.idle_add(self.on_timer)
> >
>
> could you please clarify for me again the rationale behind your patch?  The
> current version, using timeout_add, was reverted from your code very recently:
>
> 2006-01-04  Fernando Perez  <Fernando.Perez at colorado.edu>
>
>         * IPython/Shell.py (IPShellGTK.mainloop): fix bug in the GTK
>         backend for matplotlib (100% cpu utiliziation).  Thanks to Charlie
>         Moad for help with tracking it down.
>
> The problem was that GTK was using 100% cpu with the idle_add version, here's
> the thread on the matplotlib list where we tracked this down:
>
> http://sourceforge.net/mailarchive/forum.php?thread_id=9385296&forum_id=36187
>
> So we did indeed revert your patch out, but it was because the idle_add form
> was breaking matplotlib.   I guess we'll need to understand this better, so we
> can find a proper solution for all cases.  Perhaps Charlie also has some ideas
> on this front.
>
>
> Cheers,
>
> f
>
> _______________________________________________
> IPython-user mailing list
> IPython-user at scipy.net
> http://scipy.net/mailman/listinfo/ipython-user
>




More information about the IPython-user mailing list