🚀 go-pugleaf

Re <9208200428.AA14167@cookie-crisp>
<9208192024.AA14584@life.ai.mit.edu>:

> I compiled GhostScript 2.5 on a IBM/PC clone with a 80486 running MSDOS 5.0
> using the Borland C++ 6.1 compiler using the tbcplus.mak makefile.
> It compiled without error and seems to run fine aside from problems with
> the HP Color DeskJet 500C drivers.  When I use either of the two
> contributed drivers (cdeskjet/cdjcolor or djet500c), I get the error:
> 
> Unrecoverable error: VMerror in array
> 
> A similar problem occured with the GS2.4.1 using these drivers.
> When I reported it, Peter Deutsch send me mail saying that this was a well
> known bug, and he obviously assumed that I have Usenet access because he
> seemed pissed off that I would bother him with such a well known phenomenon.
> He said that the error was a memory allocation problem.
> He told me that the next version of GS would fix the bug.
> Is there something I missed or is this not fixed?

As I recall, your original message didn't make it clear as to whether
you were using the Borland compiler or the Watcom compiler.  The
"well known bug" was a "VMerror in setdevice" occurring with the
Watcom compiler; I believe this bug has been fixed.  All other
VMerror bugs, such as the one above, are the result of overflowing
MS-DOS' pitiful 640K address space.  Try compiling Ghostscript for MS
Windows, and running it under Windows in enhanced mode; if this
doesn't help, you need a 32-bit compiler, such as the Watcom
compiler, to get past this.  djgpp would do the trick, except
Ghostscript (at least the version I distribute) doesn't support it
yet.

L. Peter Deutsch :: Aladdin Enterprises :: P.O. box 60264, Palo Alto, CA 94306
ghost@aladdin.com, ...decwrl!aladdin!ghost ; 415-322-0103 voice ; no fax yet
	    "Implementation is the sincerest form of flattery."