Xref: utzoo comp.windows.x:7455 comp.sys.hp:1470
Path: utzoo!attcan!uunet!lll-winken!ncis.llnl.gov!helios.ee.lbl.gov!pasteur!ucbvax!decwrl!labrea!rutgers!mailrus!ncar!tank!mimsy!dftsrv!palantir!schuler
From: schuler@palantir.gsfc.nasa.gov ( RMS)
Newsgroups: comp.windows.x,comp.sys.hp
Subject: X windows R3 on catseye monitor
Message-ID: <578@palantir.gsfc.nasa.gov>
Date: 21 Jan 89 03:47:35 GMT
Reply-To: schuler@palantir.gsfc.nasa.gov (Larry Schuler - RMS)
Distribution: na
Organization: NASA/Goddard Space Flight Center   Greenbelt, MD
Lines: 42


Shhhhh ! 

Don't tell the people I work with about this, they will think I spent days 
hacking the Xserver to get it to work BUT to help others out....

 During my hacking on the X11R3 server, trying to get it to work on the HP
catseye monitor, I realized (recalled) that the X11R2 server was using the 
topcat declaration in the /usr/lib/X11/X0screens file, and that it would 
not even run if it was declared as a catseye. (An important clue)

 So after days of scrutinizing, mapping, hacking, and an evening of debugging, 
(I was originaly trying to integrate the X11R2 catseye code into the R3 server)
I decided to try an oooohh so simple approach. I replaced the definition
of which routines to call for HP screen dependant functions for the catseye
with topcat calls. (The HP server uses a table of function names to determine
which routines to execute based on a HP device number [Where this comes from
I still do not know]). Crossing my fingers, I did a make and and ran the server
and lo and behold it worked like a champ. SURPRISE SURPRISE!! 

 Kids, you should try this at home. 
 
 The file to modify is screentab.c in $(TOP)/server/ddx/hp/hp. The table 
can be located by searching for "catseye".  Just replace the three routines
"catseyeScreenInfo", "catseyeScreenInit", and "catseyeScreenClose", with
"topcatScreenInfo", "topcatScreenInit", and "topcatScreenClose" resepectively.

 I am still very curious as to why this works, any ideas? The machine is
a HP 9000/360CH. Is it possible that this machine has a catseye monitor, but
a topcat controller chip ?  As HP does not distribute information about
their hardware with the system (how many vendors do this though ?), I was
not even aware they had different screen types. 

Good luck

-Larry
schuler@palantir.gsfc.nasa.gov

One last thing, The server does seem to have problems with the colormap and
window managers. twm completely loses it (Not quite as bad as on the Apollo
- which may be a twm bug, but even uwm will leave tracks from it's popup
windows occasionaly.)