[comp.windows.x] X11 and Magic

tjs@dadent (tom sullivan) (07/07/89)

We've recently upgraded our uVAX-II/GPX to Ultrix 3.0, which runs X11.
I need to get Magic running on this beast and have gotten the mods from
the University of Washington (no relation to my University); however,
there is a problem with these mods. As you move the mouse from one 
window to another, the inactive window goes blank. This makes life a
bit difficult, as you are unable to read the text window when working
in the Graphics window. Other than this, everything appears to be working.

Any ideas as how to fix this, or a different X11 port would be appreciated.

thanks,

tom sullivan

tjs@dadent.wustl.edu

arons@IRIS.UCDAVIS.EDU (07/07/89)

>>We've recently upgraded our uVAX-II/GPX to Ultrix 3.0, which runs X11.
>>I need to get Magic running on this beast and have gotten the mods from
>>the University of Washington (no relation to my University); however,
>>there is a problem with these mods. As you move the mouse from one 
>>window to another, the inactive window goes blank. This makes life a
>>bit difficult, as you are unable to read the text window when working
>>in the Graphics window. Other than this, everything appears to be working.

I'm not sure what the mods from the University of Washington are, but we
found that the code as supplied in the V11R3 contrib directory had
real problems with colormap handling and are in the process of correcting
things.  Here's what's probably happening:

The orignal code made assumptions about where things were in the default
colormap that were not true on all server implementations.  One "correct"
way to do things is to allocate and install a new colormap. This has
the unwanted side affect of making things go "technicolor" (your
dissapearing text window) when it is installed.  Since magic can probably
get by with as many colors as are available in the default map this is
the path we are taking.  The disadvantage of this is that you are not
guaranteed that the pixel values are contiguous which messes up the code
a little bit.

I expect we will have the modifications ready in the next couple of weeks,
tested on a variety of server/client combinations:
Servers: HP9k/300 v6.5; Sun3 4.0; Vax MIT, Decwindows;  IBM PC/RT AIX 2.?
Clients: above plus HP9k/800 v3.10

Tom Arons                       Internet: arons@iris.ucdavis.edu
EE/CS Department                Bitnet:   arons@ucdavis
University of California        UUCP:     {lll-crg, ucbvax}!ucdavis!iris!arons
Davis, CA 95616                 Phone:    (916) 752-1750