"Serve Lau" <nihao@[EMAIL PROTECTED]
> wrote in message
news:342ff$48478d0d$541fc2ec$5135@[EMAIL PROTECTED]
>
> "jbwest" <jbwest@[EMAIL PROTECTED]
> schreef in bericht
> news:zsWdnZV8pau_q9rVnZ2dnUVZ_h3inZ2d@[EMAIL PROTECTED]
>>> so the render function would look like this in pseudo:
>>>
>>> select_context(screen1)
>>> draw_scene1()
>>>
>>> select_context(screen2)
>>> draw_scene2()
>>>
>>>
>>> the screen are unrelated, different things will be rendered not one
>>> scene over 2 screens
>>>
>>
>> Yes, but, almost all Nvidia and ATI accelerated cards have native
>> dual-screen sup****t where the pair of screens is treated as one jumbo
>> framebuffer. Don't usually have to mess with multiple anything, just
>> treat 2 screens as one extra-wide (or tall) one.
>
> I see, but this will be for an embedded system. I dont know what
videocard
> we will have but I'm pretty sure there will be 2 in the machine.
>
Well, I suggest you dig deeper. The builders may not realize that Xinerama
and OpenGL often don't get along well, or at all. What this means then,
is,
either No acclerated OpenGL (ugh, useless) with Xinerama on or, any plain
old X window on screen 0 CANNOT be moved to screen 1 with Xinerama off
(very, well, old-fa****oned & primitive). Many apps don't understand
multiple
screens & visuals. An app running on screen 0 can't just simply create
another window on screen 1 -- you have to be careful to match XVisual's &
use a different root & etc. PITA, trust me.
-jbw
|
