Re: [gtk-list] Re: Multiple Display support in GTK+?
- From: Kenneth Albanowski <kjahds kjahds com>
- To: gtk-list redhat com
- cc: raster redhat com, jharmon telecnnct com, ats acm org
- Subject: Re: [gtk-list] Re: Multiple Display support in GTK+?
- Date: Tue, 14 Apr 1998 23:32:51 -0400 (EDT)
On Sat, 11 Apr 1998 raster@redhat.com wrote:
> In X you cannot have windows span 2 screens on a multi-head display.
> they are efectivel like 2 different displays.. they just happen to go
> through the same Xserver. This is completely removed form a window
> manager and its functions.
I'm reminded of the Mac, which has handled multiple heterogenous displays
with a shared display space for ages.
Conceptually, it boils down to having a "physical" relationship between
displays (so it knows how the edges should be glued together), and when a
window overlaps a screen border, it is drawn independantly on each screen,
using a separate pass through the drawing code.
Unfortunately, under X where things like windows, colors, and cursors are
intimately tied to the X server, this just isn't very simple to implement.
I should think that the main trick needed in Gtk is the ability to
abstract widget drawing completely away from X, so that there is one
function to call "draw widget using this particular visual, this depth,
and this set of colors", which can be repeated for any screens that the
widget is displayed on.
Since that doesn't sound very efficient (especially with delta display
functions like button clicks and so forth), some sort of push-pop
graphical context stack might make more sense. All in all, it's not
trivial, especially if everything is based on X.
Of course, if the X server already does all the work itself, for
same-depth multi-heads, there's nothing that need be done. X is nice that
way.:-)
--
Kenneth Albanowski (kjahds@kjahds.com, CIS: 70705,126)
[
Date Prev][
Date Next] [
Thread Prev][
Thread Next]
[
Thread Index]
[
Date Index]
[
Author Index]