[510] in Release_7.7_team

home help back first fref pref prev next nref lref last post

Re: sun4 8.0A: I can't preview or print from ez

daemon@ATHENA.MIT.EDU (Albert Dvornik)
Thu Apr 25 15:56:42 1996

From: "Albert Dvornik" <bert@MIT.EDU>
To: Greg Hudson <ghudson@MIT.EDU>
Cc: Bill Cattey <wdc@MIT.EDU>, sol24@MIT.EDU, Carla Fermann <carla@MIT.EDU>,
        release-team@MIT.EDU, irix53@MIT.EDU
In-Reply-To: ghudson@mit.edu's message of 25 Apr 1996 (12:20:37 EDT).
             <9604251620.AA05304@small-gods.MIT.EDU> 
Date: Thu, 25 Apr 1996 15:56:20 EDT


Greg said:

> All of your tests for pslpr, psdraft, and psnup use as input the
> result of "psroff -t -man man/psdraft.1p".  If I take the output of
> psroff -t -man man/psdraft.1p on an SGI and then run, say, psdraft on
> it on a Sun, I get the same (bad) output as if I run psdraft on an
> SGI.
> 
> That is to say, the problem appears to be in the psroff output on the
> SGIs.  psroff uses troff, so it appears to be a problem with the SGI
> troff output.  Both the Sun and SGI troff output use currentmatrix and
> setmatrix; I'm guessing that the SGI troff output uses them in an
> unethical manner (read: not conforming with the Adobe Document
> Structuring Conventions).

<PICKY importance=low>

Strictly speaking, troff doesn't output PostScript, it outputs Magical
Mystery Ditroff Stuff, which then needs to be shoved through a back-end in
order to produce device-specific code such as PostScript.  In the case of
PostScript, the back end (psdit) is a part of Transcript; in fact, psroff
is a script which is a glorified wrapper to run

   troff [font info and other options] | psdit [psdit options]

(possibly further piped to lpr).

It is possible that the SGI problem is, in some way, related to psdit
(Transcript) and not troff (vendor).  Note that I did not say "likely". =)
You may want to send SGI troff output through Sun psdit to verify this.

</PICKY>

> When I played with psdraft and psnup on other postscript files, they
> worked fine on an SGI.
> 
> Although it's bothersome that the SGI troff doesn't seem to produce
> DSC-compliant output, I don't think it needs to affect our release
> plans.

This I certainly agree with; this is a fairly obscure, low-key problem and
we could certainly go on without even looking at whether the problem is
with troff or psdit, let alone fixing it.  But if you have a free minute
and a set of binaries+fonts handy (I don't), it would be nice to know...

--bert

home help back first fref pref prev next nref lref last post