This is the mail archive of the
cygwin-xfree@sources.redhat.com
mailing list for the Cygwin project.
Re: Color depth question.
- To: Clemmitt Sigler <siglercm at alphamb2 dot phys dot vt dot edu>
- Subject: Re: Color depth question.
- From: Alexander Gottwald <Alexander dot Gottwald at informatik dot tu-chemnitz dot de>
- Date: Fri, 16 Feb 2001 18:17:52 +0100 (MET)
- Cc: cygwin-xfree at sourceware dot cygnus dot com
On Fri, 16 Feb 2001, Clemmitt Sigler wrote:
> Hi all,
>
> I've been able to get XFree working on NT4 SP6a, but only in 8-bit color
> depth. 16-bit and 24-bit "run," but the colors are wrong. When you
> export netscape from Linux to the XFree on NT display, the resulting
> colors are strange, I'm not sure how to explain it. It looks
> exactly the same as trying to display a 16-bit or 32-bit color app
> on an 8-bit only display. There's a weird "rainbow" effect to icons
> displayed, for example.
>
Yes, the colors are broken. But since Harold works on a complete new
XServer, I don't expect this will get fixed soon. Maybe you can create
a workaround by modifying the rgb.txt or you have to live with it.
> I've tried changing color depth to 8-bit, 16-bit, and 24-bit on the
> Windows desktop, and also changing resolution to 800x600 and 1024x768.
> This doesn't change the color behavior I'm seeing for XFree. [BTW,
> I can confirm that XFree showing the Windows arrow cursor and crashing
> on the first mouse click is caused by trying to run XFree in the
> same resolution and color depth as Windows is running.]
I think this is a result of the long startup time. When changing video
mode, windows is busy for a while and the XServer has time to lock
the screen. But that's only a guess.
>
> I'm wondering if this color depth problem means something is wrong with
> the drivers for my video card (ancient Matrox Mystique, last drivers
> posted in 1998). Thanks for any feedback :^)
>
I don't think so.
I think the color problem is caused by difference how the color is
transmited in the XProtocol and how it's stored in the videocard ram.
I expect that the needed change in the code weren't to big. Since the
only XProtocol functions using absolute color values (red,green,blue) are
AllocColor, AllocColorPlanes, StoreColors, QueryColors and LookupColor.
I think that only these functions need to get altered to map the 16bit
values for red, green and blue to the 32bit pixel value.
ago
--
Alexander.Gottwald@informatik.tu-chemnitz.de
http://www.gotti.org
+49 3725 3498080