cjk-list
[Top][All Lists]
Advanced

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

[cjk] On modifying non-compliant fonts Re: texlive svn write access Fw:


From: Hin-Tak Leung
Subject: [cjk] On modifying non-compliant fonts Re: texlive svn write access Fw: Re: revisiting ttf2tfm and dvipdfmx
Date: Thu, 18 Jul 2013 22:07:09 +0100
User-agent: Mozilla/5.0 (X11; Linux x86_64; rv:20.0) Gecko/20100101 Firefox/20.0 SeaMonkey/2.17.1

On modifying non-compliant fonts - or modifying a font for whatever reason, like adding a cmap to make it possible to view traditional chinese documents with simplified chinese glyphs -, besides the issue of know-how (not everybody should or would read the font format spec and hack around with a hex editor), there is also the issue with whether the result can be posted/shared.

There are fonts which have a GPL-like/compatible license - the Arphic fonts have half a dozen derivatives, differing in names and minor details and mostly incompatible and confusing with each other because of that; and the wcl fonts have derivatives which are "wrong" like having duplicate/redundant cmaps. Being legally allow to modify and redistribute a derivative isn't necessarily a good thing.

There are the restrictive ones, tied to platform, installation instance, or no printing/embedding.

There are also many in the middle - free use and/or re-distribution but only as-is. AFAIK, most of the government-issued reference fonts have this - the current and past Taiwanese ones and the Hong Hong ones; the IPA fonts from the Japanese government allows re-distribution of derivatives but have a "no derivative claiming the old name" clause (come to think of it, I believe the 4 Arphic fonts also have this clause - so the way cjk/texlive is bundling the Type1 derivatives might be violating the license).

That's a few dozen fonts there.

Those are useable, but you live with whatever non-compliantness or not-fit-for-purpose-ness, there is.

Anyway, editing a font with a hex editor as a proof-of-concept that it is not strictly spec-compliant, I am okay with that. But creating/maintaining a derivative as a general and/or long-term solution, I'd prefer not to go there.

FWIW, the "abort on unknown post table version" issue (instead of "treat unknown as 00030000 with a warning and continue") is wide-spread, and located at - by grep'ing through:

texk/web2c/pdftexdir/writettf.c
texk/web2c/pdftexdir/ttf2afm.c
texk/web2c/luatexdir/font/writettf.w
texk/xdvipdfmx/src/tt_post.c
texk/dvipdfmx/src/tt_post.c

I have certainly seen the 1st and last - and got around them by editing those 2 fonts *as a proof of concept*. I don't see why unknown post table version should cause an abort, given that just flipping one byte to v3 works. Obviously neither pdftex nor dvipdfmx are using or needing any information from the post table.

(None of the current government-issued fonts are among the 460 instances I mentioned - I just haven't got round to going through them yet).



reply via email to

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