From: rohrer@fncrd6.fnal.gov (Keith Rohrer) Subject: Re: X weirdness. Date: 28 Jul 1992 14:09:41 GMT
In article <ins295b.712319744@aurora.cc.monash.edu.au> ins295b@aurora.cc.monash.edu.au (Andrew Humphrey) writes:
>Hi,
>#include "pat.on.back.for.linus"
> I've been tinkering with getting X11 to work with my setup
> (50 MHz 486 with an Tseng Labs ET4000 card and a tatung multi scanning
>monitor) and for ages was unable to get a stable picture in any of the
>modes that the various resources had told me should work. Then
>completely by accident I tried to start X11 after I had selected one of
>the SVGA text modes at boot time (100x40) and It worked perfectly. Can
>anyone out there give me a vague idea of why this is so. If more
>details are required mail me and I'll supply what's needed.
I have a similar sort of thing going on with my Swan Palette Plus
(ET4000/HiColor) board. If you use the trick about commenting out
all "clocks" lines in your Xconfig and then startx'ing with stderr
redirected to a file, then sync and reboot and read the file to see
what your clocks are, you not only (1) get different frequencies on
my machine depending on what svga text mode you are/aren't in, you (2)
get different frequencies depending on whether you've already run X
(at least like that) or not!! What's worse is, X asks for an oscillator
speed by number, not by speed, so if your clocks line has 25 in the first
spot, and the real speed is 30 MHz, it will send a signal pattern calibrated
for a 25 MHz clock using the 30 MHz oscillator. THIS IS DRIVING ME
UP A WALL.
I have a feeling that, since I've got a variable oscillator (VCO?) on my
card, I'm in the same boat as the people with the new Speedstar 24s...
any progress on that situation yet?
C'MON PEOPLE, I CAN DO 1024x768 NO PROBLEM IN DOS MODE!!!
(that oughta get 'em...)
>Andrew Humphrey
>ins295b@aurora.cc.monash.edu.au
>ins295b@lindblat.cc.monash.edu.au
>humpy@yoyo.cc.monash.edu.au
>
>"Just because I'm paranoid, it doesn't mean the whole world isn't out to
>get me".
Keith
(rohrer@fncrd0.fnal.gov)