Re: TODO list updates

[Havoc Pennington <hp redhat com>]

Ettore Perazzoli <ettore@helixcode.com> writes:
>   I also have some other things to say on the look & feel side.  This
> is probably not high priority, but these problems should be fairly
> easy to fix...
> 
>   - I think the extra border size for Enter-activated buttons should
>     be user-specifiable.  Actually, I would like it not to exist at
>     all by default: it looks ugly, and contributes to the general
>     "chubbiness" (as Miguel put it) of the GNOME UI.
>

IIRC Owen's stock response to this is that we need the Windows way of
indicating the default if we do this (dotted black line), and the
dotted black line thing has to be used for all widgets that can be the
default. I'm not sure what those are besides buttons.
 
>   - There should be a way to disable prelight in the gtkrc.
>   Currently, themes that try to be prelight-less just specify the
>   same color for the prelight and normal states and this causes
>   flickering.  (Probably this is not a major issue anymore with
>   Owen's no-flicker work, but still...)
> 
>   - Changing theme on the fly sometimes does not update the size of
>     all the existing widgets in running applications correctly.  For
>     example, if you switch from a theme with "small" scrollbars to one
>     with "large" scrollbars and then switch to the one with the
>     "small" scrollbars again, usually the scrollbars in running
>     applications remain large.  (At least on 1.2.x, I don't know about
>     1.3 enough -- maybe it has been fixed already?)
>

This is because changing scrollbar width isn't really supported, and
those themes are doing some really insane hacks in order to do it.
 
>   It would also be nice if theme engines could specify thickness
> per-widget instead of just specifying x/ythickness in the theme struct
> as it happens now.  If I recall correctly, this is one of the reasons
> why, for example, you cannot really have Windows 95 toolbars in the
> Windows 95 theme.
> 

The theme engines interface needs a major rework; I'm not sure it's worth
breaking the existing theme engines in small ways, since the next GTK
after this one will probably break them in large ways.

Havoc