|
From: | P. Alejandro Lopez-Valencia |
Subject: | Re: [groff] PDF problems - see attached file |
Date: | Wed, 22 May 2002 12:56:52 -0500 |
At 11:03 a.m. 22/05/2002 -0400, Stewart Russell wrote:
P. Alejandro Lopez-Valencia wrote: > > I'd suggest you file a report at SourceForge's GS bug tracker. ^^mind you, it could be an aspect of grops output. Feeding sample text through Solaris 8 troff/dpost, the ps2pdf (7.04) output looks fine in acroread. groff 1.17.2's output does not, and seems to trigger some class of subsetting weirdness in gs7.
As far as I know grops postscript dictionary hasn't changed between 1.7.2 and the latest development version. And the fact that gs 6.5.2 can produce good output is what I use to draw my conclusions. As well, and I forgot to mention it this morning, I did not use ps2pdf but rather Gsview 4.2 internal hooks to the GS 7 API to produce the PDF.
It could be anything related to the interpreter, thus the real test would be trying to print the file to a true Adobe Postscript printer. grops' dictionary is level 1 but to make the test valid against GS 7.x, the printer should have a true Adobe Level 3 Postscript interpreter.
Another test to make sure that groff's ps output is correct, would be to run the postscript file through PSAlter, if anyone has it.
On the subsetting weirdness, it might be rather that GS 7.x has trouble with font custom vector re-encoding. There's definitely at least one serious bug in GS related to font embedding, i.e., it insists on including Times Roman even if not used in the original PS file. Or are you per chance using font outlines generated with any version of Fontographer?
[Prev in Thread] | Current Thread | [Next in Thread] |