[comp.sys.sgi] gl_readscreen, scrsave, icut, snapshot bug

blbates@AERO4.LARC.NASA.GOV ("Brent L. Bates AAD/TAB MS361 x42854") (11/14/90)

     We have a 4D/210VGX and a 4D/320VGX both OS 3.3.1.  I have found
a bug in gl_readscreen, which causes scrsave, icut, and snapshot not
to work properly.  If you try to save a full screen width, the full
1280 pixels wide, the image is garbaged with vertical lines. If I make
the image one pixel narrower everything works fine.
     Perhaps, before SGI makes gl_readscreen a "documented" and "supported"
routine, they could fix this bug in the next release, 3.3.2?

	Brent L. Bates
	NASA-Langley Research Center
	M.S. 361
	Hampton, Virginia  23665-5225
	(804) 864-2854
	E-mail: blbates@aero4.larc.nasa.gov or blbates@aero8.larc.nasa.gov
--

					Brent

jim@baroque.Stanford.EDU (James Helman) (11/14/90)

Hmm...  I've had lots of trouble with scrsave giving spurious
horizontal lines on our 4D/220GTX, but never vertical.  Also, the
problem is not new.  I've seen it happen in previous releases.  If
anyone has a workaround, I'd like to know.  On a couple occasions, it
has cost me a several hours of work because I *thought* I was making
good screen dumps.  Now, I'm more careful.

Jim Helman
Department of Applied Physics			Durand 012
Stanford University				FAX: (415) 725-3377
(jim@KAOS.stanford.edu) 			Work: (415) 723-9127

blbates@AERO4.LARC.NASA.GOV ("Brent L. Bates AAD/TAB MS361 x42854") (11/15/90)

     The only thing I can think of at the moment to patch
scrsave, is to change the source (which is in 4Dgifts) so that the
maximum screen width saved is one less than actual screen maximum.
--

	Brent L. Bates
	NASA-Langley Research Center
	M.S. 361
	Hampton, Virginia  23665-5225
	(804) 864-2854
	E-mail: blbates@aero4.larc.nasa.gov or blbates@aero8.larc.nasa.gov

fsfacca@AVELON.LERC.NASA.GOV (Tony Facca) (11/17/90)

> Hmm...  I've had lots of trouble with scrsave giving spurious
> horizontal lines on our 4D/220GTX, but never vertical.  Also, the
> problem is not new.  I've seen it happen in previous releases.  If
> anyone has a workaround, I'd like to know.  On a couple occasions, it
> has cost me a several hours of work because I *thought* I was making
> good screen dumps.  Now, I'm more careful.

On the 320 and 340 VGX systems I have notices horizontal lines in all of
screen dumps saved with scrsave.  As a workaround, use pixel coordinates
in the command as follows:

	scrsave ~/screen.dump 1 1279 1 1023

You loose one row and one colomn of pixels, if you can live with that.
Also, someone had posted a nifty program to activate the "Print Screen"
key.  This comes in handy for a lot of applications which do not do there
own screen dumps when they are in full screen mode.

The program has been hacked up a bit, but it works pretty well:

-------------------------- cut here for best results ----------------------

#!/usr/NeWS/bin/psh

%
% activate the 'Print Screen' key and bind it to the scrsave function.
%
% You can remove the '##' sign if you want to automatically spool the print
% to the Tektronix network printer.  Or change 'tek77' to your local network
% printer, you get the idea.   -- tony facca
%

systemdict begin

/ScreenPrinter {
    newprocessgroup
    createevent begin
	/Name 16#6F9E def	% PrintScreen key scan code
	/Action [/UpTransition /DownTransition] def
	/Exclusivity true def
	currentdict
    end expressinterest
    {
	awaitevent begin
	    Action /UpTransition eq {
		console rand (Saving screen%\n) fprintf
 	        (scrsave ~/screen.dump 1 1279 1 1023			     \
	 	 ; echo Screen saved as ~/screen.dump \ > /dev/console     \
##		 ; lp -dtek4693 ~/screen.dump				     \
		) forkunix
	    } if
            (Screen saved as screen.dump) print
	end
    } loop
} fork def

(Ready -- press 'Print Screen' for screen dump\n) print

end

-------------------------- cut here for best results ----------------------
--
-----------------------------------------------------------------------------
Tony Facca   |   fsfacca@avelon.lerc.nasa.gov      |     phone: 216-433-8318
-----------------------------------------------------------------------------
      You are at Witt's end.  Passages lead off in *all* directions.