Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!cs.utexas.edu!sun-barr!decwrl!chico.pa.dec.com!klee
From: klee@chico.pa.dec.com (Ken Lee)
Newsgroups: comp.windows.x
Subject: Re: optimized X11 for server/clients on same host
Message-ID: <1933@bacchus.dec.com>
Date: 17 Oct 89 21:55:00 GMT
References: <1989Oct17.191614.10659@jdyx.UUCP>
Sender: news@decwrl.dec.com
Reply-To: klee@decwrl.dec.com
Distribution: usa
Organization: DEC Western Software Laboratory
Lines: 20

In article <1989Oct17.191614.10659@jdyx.UUCP>, tpf@jdyx.UUCP (Tom
Friedel) writes:
> I have to believe that shared memory is the most efficient solution to
> this problem.
> 
> Has anyone placed such code in the public domain?

A few shared memory X transports have been floating around, but all have
problems.  
For several reasons, shared memory is not as great as it seems:
1.  memory still isn't that cheap
2.  the semaphors (or whatever) used as controls are expensive
3.  many of the Xlib algorithms (e.g., graphics batching) get more complicated

I'm not saying that shared memory can't work well.  It's just not easy.

Ken Lee
DEC Western Software Laboratory, Palo Alto, Calif.
Internet: klee@decwrl.dec.com
uucp: uunet!decwrl!klee