Re: 8r encoding + xdvi :-(

[Thomas Esser <te@informatik.uni-hannover.de>]

Now, where I have my render.ps right with the fix from Paul Vojta, there 
was some time for the "real" test. In my last mail, I claimed that all 
things are workling well with gsftopk and this is really true for 
ptmr.gsf and the commercial type1 fonts, I have tested.

But with some (most?) other .gsf fonts, some problems remain. Here an 
example:

Using the psfonts.map entry of:
	pbkl8r  Bookman-Light "TeXBase1Encoding ReEncodeFont" <8r.enc
and pbkl8r.tfm from CTAN, I produced 4 printouts of the fonttable 
(produced by 'tex testfont'):

a) using the commercial Bookman-Light.pfa + gsftopk + dvilj4
b) using the commercial Bookman-Light.pfa + dvips + gs
c) using pbkl.gsf + gsftopk + dvilj4
d) using pbkl.gsf + dvips + gs

The results in a) and b) are both fine and give the same "correct" 
output.

Using c) some positions in the fonttable are missing or broken:
	- the ligatures for fi and fl are just a small dot 
	- empty positions: 204, 205, 206, 214, 226, 227 and more ...
	- the accents are not correctly placed over the characters, e.g. 
	  position 300
	
Using d) some positions are missing and almost the accents are not 
correctly placed over the characters.

So, it seems like there are major problems with ghostscript which need to 
be fixed (mentioned in d) ) and maybe some remaining problems with gsftopk.

Thomas

--
Thomas Esser                                 te@informatik.uni-hannover.de
Univ. of Hanover, Germany
Institute for Informatics (Database and Information Systems Group)