Path: utzoo!attcan!uunet!know!samsung!zaphod.mps.ohio-state.edu!rpi!bu.edu!bu-pub.bu.edu
From: tjh@bu-pub.bu.edu (Tim Hall)
Newsgroups: comp.sys.sgi
Subject: scrsave weirdness
Message-ID: <65560@bu.edu.bu.edu>
Date: 4 Oct 90 17:20:10 GMT
Sender: news@bu.edu.bu.edu
Organization: Boston University
Lines: 32

In article <1990Oct4.134253.16475@eagle.lerc.nasa.gov>, xxjgh@dali.lerc.nasa.gov (Jay G. Horowitz) writes:
|> WARNING to those looking towards scrsave as an answer to their screen-saving
|> problem:
|> 
|>                         Snapshot and others that require a graphics
|> port to allow you do establish your capture window, eventually do a:
|> 
|> 	system ("scrsave 0 1280 0"); /* or whatever xmin, xmax, ymin, ymax */
|> 
|> which spawns a new process.
|> 
|> I assume that doing a succesfull screen capture (i.e., capture pixel rgb
|> regardless whether that pixel was colormap, rgb, window border, background,
|> or whatever) can only be done by a process that is not tied to a particular
|> window. 

I recently modified snapshot to save images in our own local image format.
I still use most of the code in scrsave, in particular gl_readscreen,
which I think you are talking about.  Anyhow, I did not use a system call
or create a new process.  The scrsave code is compiled right into snapshot
and I haven't had any problems with it.  This is on a 240 running 3.2 and
P.I. running 3.3.1.

-Tim



--
-Tim Hall
tjh@bu-pub.bu.edu

The night is filled with the cries of dispossessed children in search