[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: blot_diameter

From: Han-Wen Nienhuys
Subject: Re: blot_diameter
Date: Tue, 5 Feb 2002 00:25:58 +0100

address@hidden writes:
> On Fri, 1 Feb 2002, Han-Wen Nienhuys wrote:
> > ...
> > Good !  I've put in the patch, and, at the same time split out all
> > ancient entries into a separate font: tex fonts (and PS fonts as well)
> > can only have 256 glyphs. Almost half of the feta font was taken up by
> > the various ancient glyphs.
> >
> Interesting.  I had no problems to compile the feta fonts with more than
> 256 characters.  I even sucessfully produced proof sheets via gftodvi,
> which, IIRC, contained all (i.e. >256) characters.  It was really not
> before lily crashed with the cited assertion failure that I recognized any
> problem.

TeX book, app. F, page 427. "there are exactly 128 different
characters in each of CM fonts, although TeX can work up to 256
characters per font".

I don't know how to do 16 bit fonts: I don't know how such stuff works
(which is also the reason that lily doesn't have support for different

Anyway, can you see what happens if you do

--- Fri Feb  1 19:16:15 2002
+++  Tue Feb  5 00:23:49 2002
@@ -15,7 +15,7 @@
   checksum_ = 0;
   font_inf_ = fi;

-  for (int i= 256; i--;)
+  for (int i= 256 >? fi->numOfChars ; i--;)
     ascii_to_metric_idx_.push (-1);

   for (int i=0; i < fi->numOfChars; i++)

> Anyway, how do you handle 16 bit unicode characters in PostScript?

> And, FYI, if you look at D.E.Knuth's Pre-fascicle 2a of The Art of
> Computer Programming, volume 4 (see
> http://Sunburn.Stanford.EDU/~knuth/, you will see lots of
> Chinese, Arabian and Hebrew characters (e.g. on page 65).  There are some
> thousand Chinese characters, but I can not imagine that Knuth switches
> between dozens of fonts just to print a single name...

Again, I don't know. I'd appreciate it if you'd try to find  out.


Han-Wen Nienhuys   |   address@hidden    |

reply via email to

[Prev in Thread] Current Thread [Next in Thread]