Re: Wine patch

[Stefan Dösinger]

> Hi Stefan, did you ever get any comment on this patch,
> as why it hasn't been committed??
>
> http://www.winehq.org/hypermail/wine-patches/2005/03/att-0504/01-x11drv.dif
> f
>
> It seems as if it fixes quite a few bugs (e.g. Steam)
> . Was it ever discussed further? Regards Louis

No, it was not discussed in detail. I think this patch is incorrect, it fixes 
some apps, but it breaks others, the QuickTime Player for example behaves 
incorrect with this patch.

One suggestion was to fix non-managed windows instead, but according to some 
Documentation this is not possible because every window manager behaves 
differently with non-managed windows.

Maybe someone has a better solution

Cheers,
Stefan

Wine patch

[Saulius Krasuckas]

* On Wed, 12 Oct 2005, Stefan [iso-8859-1] D�singer wrote:
> > http://www.winehq.org/hypermail/wine-patches/2005/03/att-0504/01-x11drv.diff
>
> I think this patch is incorrect, it fixes some apps, but it breaks 
> others
>
> One suggestion was to fix non-managed windows instead, but according to 
> some Documentation this is not possible because every window manager 
> behaves differently with non-managed windows.

I haven't looked at the patch, but hey, what about run-time detecting of 
what exactly WM (and maybe its version) is running?

Sounds like refactoring WM-dependent code into some WM driver, in a vein 
of some Sound-Output driver, idea of which was given by Eric Pouech, I 
believe.

Wine patch

[Stefan Dösinger]

> I haven't looked at the patch, but hey, what about run-time detecting of
> what exactly WM (and maybe its version) is running?
>
> Sounds like refactoring WM-dependent code into some WM driver, in a vein
> of some Sound-Output driver, idea of which was given by Eric Pouech, I
> believe.

How many WMs are there? I think writing a WM 'driver' would be too much work. 
And as far as I know it's not the WMs that are broken, such unmanaged windows 
shouldn't be used due to the X11 design. But I am not an expert in this. 
Unmanaged windows have another problem: They will never show up in the WMs 
window list(Task bar, for example).

So in my personal opinion, we should try to find common properties of 
boarderless windows which make problems and make those windows managed. This 
is what my patch tries to do, but the rule isn't correct
(OR it is correct, but it reveals some other bugs. I haven't checked this).

Stefan