Dear Werner,
On Sun, 12 Feb 2006, Werner LEMBERG wrote:
But the similar routine in dvips has
{ 1 index /FID ne 2 index /UniqueID ne and
{def} {pop pop} ifelse } forall
Did you consider that UniqueID issue?
No -- the RE macro in groff is Clark's original code. Until now there
were never any complaints, but it is probably an enhancement to use
the code from dvips. Opinions?
There could not be any complaints since any font was reencoded only once,
thus there were no two fonts with the same UniqueID in PS output.
But as we are going Unicode there may be several reencodings of
the same font used in grops output.
And all reencodings get the same UniqueID.
It seems UniqueID has something to do with caching.
If there is no UniqueID no caching is done.
Since performance is not a high priority and the effect of UniqueID is
questionnable it may be simply omitted from PS fonts we obtain by
reencoding.
See the first result of the following search for insight
("Conflicting encoding question" thread in comp.lang.postscript started
Jun 6 2001):
http://groups.google.com/group/comp.lang.postscript/search?group=comp.lang.postscript&q=font+reencoding+UniqueID&qt_g=1&searchnow=Search+this+group
Or the second result, "Bug in Adobe PostScript version 2012.017!?
(Distiller)" of Dec 30 1996.
Sincerely, Michail
_______________________________________________
Groff mailing list
Groff@gnu.org
http://lists.gnu.org/mailman/listinfo/groff
