Re: dvips(k) state and maintainer

2000-01-24 Thread Tomas G. Rokicki

Just so you know:

the dvips in the pretest works *great* on the files that Karsten sent out.
That new writet1 looks good!  Who did the work?

-tom



dvips(k) state and maintainer

2000-01-20 Thread Karsten Tinnefeld

Who is the one who currently maintains dvips? I have some trouble with 
partially encoded fonts with non-standard encoding and late postscript 
interpreters in recent hp printers.

unfortunately, I have no idea who is the current maintainer. ctan has 
dvips versions up to 5.67, dvipsk to 5.78a, and 
http://www.radicaleye.com/dvipsalpha.html talks of 5.83 as being 
current. However, tetex 1.0.6 comes with 5.86.

Below the description of the problem:
I have written a short test text which uses a few characters of an 
expert font. When running dvips on the resulting dvi file with partial 
font encoding on, it shortens the definition of the Encoding vector to

/Encoding 256 array
dup 58 /ff put
dup 63 /ffi put
dup 64 /ffl put
readonly def

. Our postscript printers hp 2000n and clj 4050 complain and do not 
print: "Error: typecheck; OffendingCommand: definefont". They are 
right, since the postscript specs say "All unused posistions must
be filled with the name .notdef"

When inserting the following line after "... array", the error vanishs:
0 1 255 {1 index exch /.notdef put} for

Ideas or references? Thank you, greetings,
-- 
Karsten Tinnefeld   [EMAIL PROTECTED]
Fachbereich Informatik, Lehrstuhl 2   T +49 231 755-4737
Universität Dortmund, D-44221 Dortmund, Deutschland   F +49 231 755-2047



Re: dvips(k) state and maintainer

2000-01-20 Thread Tom Kacvinsky

Hi Karsten,

I am CC'ing this to the tex-k mailing list.

To answer your question, Tom Rokicki is the official maintainer of
dvips(k).

To answer your question about the encoding problem:

Does the original Type 1 font you are subsetting (you wrote partially
encoded but I think you really meant subsetted) have the following
lines in it?

/Encoding 255 array
0 1 255 {1 index exch /.notdef put}

If it doesn't, it probably is a Fontographer generated font, which means that
the encoding in the original font looks like this:

/Encoding 255 array

256 assignments, including the necessary /.notdefs needed to fill out
the 256 slots

You can solve your problems either by (a) fixing the original font or
(b) getting the latest snapshot of tetex, which uses Han The Thanh's
writet1 module to handle font subsetting (instead of t1part).

Tom

def (or readonly def)
On Thu, 20 Jan 2000, Karsten Tinnefeld wrote:

 Who is the one who currently maintains dvips? I have some trouble with 
 partially encoded fonts with non-standard encoding and late postscript 
 interpreters in recent hp printers.
 
 unfortunately, I have no idea who is the current maintainer. ctan has 
 dvips versions up to 5.67, dvipsk to 5.78a, and 
 http://www.radicaleye.com/dvipsalpha.html talks of 5.83 as being 
 current. However, tetex 1.0.6 comes with 5.86.
 
 Below the description of the problem:
 I have written a short test text which uses a few characters of an 
 expert font. When running dvips on the resulting dvi file with partial 
 font encoding on, it shortens the definition of the Encoding vector to
 
 /Encoding 256 array
 dup 58 /ff put
 dup 63 /ffi put
 dup 64 /ffl put
 readonly def
 
 . Our postscript printers hp 2000n and clj 4050 complain and do not 
 print: "Error: typecheck; OffendingCommand: definefont". They are 
 right, since the postscript specs say "All unused posistions must
 be filled with the name .notdef"
 
 When inserting the following line after "... array", the error vanishs:
   0 1 255 {1 index exch /.notdef put} for
 
 Ideas or references? Thank you, greetings,
 



Re: dvips(k) state and maintainer

2000-01-20 Thread Tom Kacvinsky

That should be

/Encoding 256 array

Sorry about the mix up...

Tom

 
 /Encoding 255 array
 0 1 255 {1 index exch /.notdef put}