[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: partial downloading with dvips
- To: bkph@ai.mit.edu
- Subject: Re: partial downloading with dvips
- From: Werner LEMBERG <sx0005@sx2.HRZ.Uni-Dortmund.DE>
- Date: Thu, 24 Aug 2000 22:01:21 +0200 (CEST)
- Cc: tjk@ams.org, tex-fonts@math.utah.edu, tex-k@tug.org, schlegel@vossnet.de, yannis@fluxus-virus.com, handa@etl.go.jp, thep@links.nectec.or.th, surapan@nectec.or.th, virach@nwg.nectec.or.th, tetex-pretest@informatik.uni-hannover.de
- In-Reply-To: <4.2.2.20000823083718.01f96610@ai.mit.edu>
- References: <20000822.081447.116087348.sx0005@sx2.hrz.uni-dortmund.de><Pine.GSO.4.10.10008221819280.21367-100000@sun06.ams.org><4.2.2.20000823083718.01f96610@ai.mit.edu>
- Reply-To: Werner LEMBERG <wl@gnu.org>
[About problems with partial downloading of the Norasi PS fonts with
dvips.]
> I believe the problem may be that true Adobe PS interpreters used in
> some popular Apple printers made during a certain period cannot
> handle more than two elements in the StemSnapH and StemSnapV lists.
> This font has:
>
> /StemSnapH [ 19 21 28 33 52 ] ND
> /StemSnapV [ 17 18 19 36 39 45 52 129 212 ] ND
>
> There is code one can add to test for printers of this type and then
> use shortened lists, but this font does not use this (tedious)
> refinement (there are other problems with this font such as many
> reversed contours, and truly chaotic "auto-hinting", but that should
> not affect the printer or DVIPS except in quality of output).
The reason of failure is indeed a bug in Thanh's writet1.c module,
used by both pdfTeX and dvips. He writes:
This is a known problem with fonts where the encoding vector is not
ended with `readonly def' but only `def'.
[He also sent me a new version of writet1.c which is hopefully in the
next teTeX beta.]
For the moment, I will add `readonly' to all encoding vectors in the
Norasi fonts. Virach, Theppitak, Surapant, please fix this for the
sake of backwards compatibility.
Werner