Re: [E-devel] E CVS: libs/evas raster

[Carsten H. (T. R. <ra...@ra...>]

On Tue, 13 Nov 2007 22:19:44 +0100 Simon TRENY <sim...@fr...> babbled:

> On Tue, 13 Nov 2007 12:15:39 +1100,
> Carsten Haitzler (The Rasterman) <ra...@ra...> wrote :
> 
> > On Wed, 10 Oct 2007 18:23:56 +0200 Christian Hunke
> > <Chr...@we...> babbled:
> > 
> > > Its only the out-commenting of the constants in evas_object_main.c.
> > > When you just undo this it already works:
> > > 
> > > http://enlightenment.org/viewvc/e17/libs/evas/src/lib/canvas/evas_object_main.c?r1=1.60&r2=1.61&sortby=date
> > 
> > yes - those changes broke etk. there is no good reason i can think of
> > why they should beak it. unless its some reliance on behavior that
> > wasn't necessarily good/intended, but the changes are an improvement
> > to evas. i found things that broke in e17 and fixed them. they were
> > silly things that were bugs waiting to happen, so the changes were
> > good.
> 
> Actually, Etk uses the smart resize-callback to update a widget when
> the widget has to be updated (by calling
> evas_object_resize(widget->smart, widget->w, widget->h). Since the
> callback is not called anymore if the widget has not really been
> resized, it broke down Etk when widgets needed a content-update while
> the widget was not resized. To fix this, now we just call manually the
> resize-callback when we need to update a widget that has not been
> resized.

oooh so THATS why! you should have a separate widget update call really.
piggybacking the resize callback is really wrong :(

-- 
------------- Codito, ergo sum - "I code, therefore I am" --------------
The Rasterman (Carsten Haitzler)    ra...@ra...