[Sugar-devel] [Dextrose] [PATCH sugar-toolkit] Cursor moves to beginning while trying to edit activity instance name #2608

Gonzalo Odiard gonzalo at laptop.org
Tue Feb 15 15:25:50 EST 2011


Is this useful for you?

http://bugs.sugarlabs.org/attachment/ticket/1948/0001-Save-title-when-closing-1948.patch

Gonzalo

On Tue, Feb 15, 2011 at 5:19 PM, Bernie Innocenti <bernie at codewiz.org>wrote:

> On Tue, 2011-02-15 at 21:00 +0100, Sascha Silbe wrote:
> > Excerpts from Aleksey Lim's message of Mon Feb 14 23:23:25 +0100 2011:
> >
> > >      def __jobject_updated_cb(self, jobject):
> > > +        prev_pos = self.entry.props.cursor_position
> > >          self.entry.set_text(jobject['title'])
> > > +        self.entry.set_position(prev_pos)
> >
> > The GTK documentation doesn't say anything about what the cursor position
> > will be after a set_text() call, so we need to take care of saving and
> > restoring the cursor position (i.e. your patch does the right thing).
> >
> > But the documentation for set_position() says the position must not
> > exceed the number of characters currently in the widget, so we need some
> > bounds checking here (the title might have been shortened).
>
> The above patch isn't sufficient in this race-condition between the user
> typing and the metadata update: even if we reset the cursor position to
> the right place, the text that we set back into the widget is likely to
> miss one or two keystrokes from the user.
>
> We must absolutely avoid triggering a save while the user could still be
> typing in the widget, which is easily achieved by *NOT* having an
> arbitrary timeout of 2 seconds and instead hooking to the
> focus-out-event.
>
> The problem with the focus-out-event is that we don't get one when the
> user clicks directly on the Stop button after editing the name. That
> case is tricky because the Stop button does not have an easy way to find
> the name widget even though it's in the same toolbar.
>
> Someone (tomeu?) pointed out that, normally, gtk sends a focus-out-event
> in this case. There was even a small test case demonstrating it. For
> some reason, it doesn't work in Sugar. Fixing this bug would let us
> remove all these half-baked workarounds.
>
> --
>   // Bernie Innocenti - http://codewiz.org/
>  \X/  Sugar Labs       - http://sugarlabs.org/
>
>
> _______________________________________________
> Dextrose mailing list
> Dextrose at lists.sugarlabs.org
> http://lists.sugarlabs.org/listinfo/dextrose
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.sugarlabs.org/archive/sugar-devel/attachments/20110215/c4b796ad/attachment.html>


More information about the Sugar-devel mailing list