Hello all,
It’s been a long time since I’ve written, but I run into a bit of a
concern that I hoped y’all could help with. I am currently in the
design phase of putting together a linux system that will require 3
separate video outputs and 2 separate touch screen inputs. Is it
possible to parse the input using SDL and determine which of the
touchscreens is being pressed? Also, has anyone used a quad head video
card with SDL? Finally, is it possible in linux, to use an array of
video cards instead of a single quad head video card can I use two dual
head cards? I assume the answer to this one is you much use a single
card, but correct me if I am wrong.
Thank you for your time,
Robert Diel
Robert Diel wrote:
Hello all,
It’s been a long time since I’ve written, but I run into a bit of a
concern that I hoped y’all could help with. I am currently in the
design phase of putting together a linux system that will require 3
separate video outputs and 2 separate touch screen inputs. Is it
possible to parse the input using SDL and determine which of the
touchscreens is being pressed?
Right now SDL dosn’t support multiple windows, nevermind multiple
displays. What you may want to do is have multiple programs (or copies
of the program) running that communicate with each other via shared
memory or signals - then each process would be handing the input and
display of their specific window (or passing it along to another
process, which would imply the screen/input by context). How feasable
this is depends on the application.
Also, has anyone used a quad head video
card with SDL? Finally, is it possible in linux, to use an array of
video cards instead of a single quad head video card can I use two dual
head cards? I assume the answer to this one is you much use a single
card, but correct me if I am wrong.
I’d actually assume the opposite. There is the possibility that you may
have to run multiple X servers to accomplish this… but I’m nearly
positive multiple graphics cards would work on a linux box. And by
"would work" I mean there’s easy(ish) to use ready to go support
programmed for it.> Thank you for your time,
Robert Diel
SDL mailing list
SDL at libsdl.org
http://www.libsdl.org/mailman/listinfo/sdl
Michael B. Edwin Rickert wrote:
Robert Diel wrote:
Also, has anyone used a quad head video
card with SDL? Finally, is it possible in linux, to use an array of
video cards instead of a single quad head video card can I use two dual
head cards? I assume the answer to this one is you much use a single
card, but correct me if I am wrong.
I’d actually assume the opposite. There is the possibility that you may
have to run multiple X servers to accomplish this… but I’m nearly
positive multiple graphics cards would work on a linux box. And by
"would work" I mean there’s easy(ish) to use ready to go support
programmed for it.
Sorry for the doublepost but check out:
http://www.faqs.org/docs/Linux-HOWTO/Xinerama-HOWTO.html#PCIBUSSECTION
http://www.faqs.org/docs/Linux-HOWTO/Xinerama-HOWTO.html#EDITXF86CONFIG
-Mike
Right now there is no multi-monitor support in SDL. This means that SDL
apps do horrible things when they go into fullscreen on my dual-monitor
setup. This includes, to my dismay, UT2004, and my own program.
Since I have development access to Mac OS X, Windows, and Linux, I’ve
been debating creating a patch for SDL that essentially allows you to
choose which monitor your context is created on, probably by creating a
new function that you would call instead of SDL_SetVideoMode. It would
be essentially the same, except take one additional argument to choose
which screen to use. This would also require new analog functions for
SDL_ListModes and the like, as well as a function to query available
screens.
If this patch were written, would there be interest in including it with
the main SDL distribution? Has someone else already written this, and
if so, could I have a diff?
Richard Schreyer
Michael B. Edwin Rickert wrote:> Michael B. Edwin Rickert wrote:
Robert Diel wrote:
Also, has anyone used a quad head video
card with SDL? Finally, is it possible in linux, to use an array of
video cards instead of a single quad head video card can I use two dual
head cards? I assume the answer to this one is you much use a single
card, but correct me if I am wrong.
I’d actually assume the opposite. There is the possibility that you
may have to run multiple X servers to accomplish this… but I’m
nearly positive multiple graphics cards would work on a linux box. And
by “would work” I mean there’s easy(ish) to use ready to go support
programmed for it.
Sorry for the doublepost but check out:
http://www.faqs.org/docs/Linux-HOWTO/Xinerama-HOWTO.html#PCIBUSSECTION
http://www.faqs.org/docs/Linux-HOWTO/Xinerama-HOWTO.html#EDITXF86CONFIG
-Mike
SDL mailing list
SDL at libsdl.org
http://www.libsdl.org/mailman/listinfo/sdl
Are you by any chance running on nVidia hardware? Personnaly I have no
problems running fullscreen SDL apps (like UT2004) in a dual-monitor setup,
as the nVidia drivers allow setting up “meta-modes” where only one screen is
being used when switching to a specific resolution.On January 13, 2005 0136, Richard Schreyer wrote:
Right now there is no multi-monitor support in SDL. This means that SDL
apps do horrible things when they go into fullscreen on my dual-monitor
setup. This includes, to my dismay, UT2004, and my own program.
I use nvidia hardware with twinview, all my metamodes specify all active
monitors. UT2004 starts between the two screens, which I can then 'pan’
to the wrong one. UT2004 also does not list any resolution other than
640x480. Regardless of ut2004, being unaware of Xinerama is still a bug
in SDL.
Richard Schreyer
Simon Roby wrote:> On January 13, 2005 0136, Richard Schreyer wrote:
Right now there is no multi-monitor support in SDL. This means that SDL
apps do horrible things when they go into fullscreen on my dual-monitor
setup. This includes, to my dismay, UT2004, and my own program.
Are you by any chance running on nVidia hardware? Personnaly I have no
problems running fullscreen SDL apps (like UT2004) in a dual-monitor setup,
as the nVidia drivers allow setting up “meta-modes” where only one screen is
being used when switching to a specific resolution.
SDL mailing list
SDL at libsdl.org
http://www.libsdl.org/mailman/listinfo/sdl
Since I have development access to Mac OS X, Windows, and Linux, I’ve
been debating creating a patch for SDL that essentially allows you to
choose which monitor your context is created on, probably by creating a
new function that you would call instead of SDL_SetVideoMode. It would
be essentially the same, except take one additional argument to choose
which screen to use. This would also require new analog functions for
SDL_ListModes and the like, as well as a function to query available
screens.
I’d be interested in seeing such a patch, for SDL 1.3 development.
See ya!
-Sam Lantinga, Software Engineer, Blizzard Entertainment