On 6/13/05, Matthew Weier O'Phinney <mweierophinney@...> wrote:
> On 6/6/05, Matthew Weier O'Phinney <mweierophinney@...> wrote:
> > I don't know why I haven't asked about this before, as I've noticed the
> > behaviour for a long time now...
> > This is with ROX-Filer >=3D 2.0.0 and OroboROX >=3D 0.9.6.
> > When I choose 'Window|New Window' from the context menu of a filer (whi=
> > have bound to the key 'W'), the new window opened opens in the upper le=
> > the screen, not directly under the cursor as I would expect. Is this
> > intentional, or isolated to my own system?
> > Any light Guido, Thomas, Jonatan, et.al could shed on the situation I'd
> > appreciate -- and, of course, I'll be a willing guinea pig for any code
> > changes resulting from this request!
> I found the offending piece of code. It's in filer.c -- I didn't produce =
> diff, as I was working off an old CVS version. Basically, in
> filer_window_set_size(), there was a section of code at the end that was =
> some pointer warping if the window was opened from a key event. It starts=
> if (event && event-type =3D=3D GDK_KEY_PRESS)
> I found that by commenting out that conditional area (the if statement an=
> code to execute on the condition), the behaviour is then correct.
> I'll see if I can create a patch later off of current CVS.
The patch is attached; it simply comments out the area that causes the unde=
behaviour. I'll leave it as an exercise for somebody else to figure out why=
section of code causes problems (somebody with more knowledge of C than
Matthew Weier O'Phinney