Re: TeX Licenses & teTeX (Was: Re: forwarded message from Jeff Licquia)
Branden Robinson <firstname.lastname@example.org> writes:
> All right. It doesn't harm my analysis to presume that the message
> quoted by Ms. Connelly does not constitute a grant of license to any
> party. We're still in the position of needing the copyright notices and
> license terms inside TeX, METAFONT, and the Computer Modern font files
1) Frank Mittelbach is nowhere near the right person to ask for that.
As you ably point out, he's way way confused, muddled, and prone to
inconsistency. And, he has nothing to do with TeX, Metafont, or CM
2) In any case, I don't think much clarification is actually needed.
> I'm willing to stipulate that the existing licensing on TeX, METAFONT,
> and the Computer Modern fonts is not DFSG-free. This is unfortunate.
TeX comes with two relevant provisions, neither of which impact DFSG.
(The situation for METAFONT is exactly parallel to TeX.)
You cannot modify tex.web at all, but you are free to patch it with
what you want and distribute the results, including binaries made from
it. This is exactly the sort of thing that DFSG 4 had in mind,
indeed, TeX has been the classic example of the implementation of DFSG
A separate--entirely separate, physically and sensibly--provision is
about what you can legally call "TeX", which happens to be a
registered trademark. The intentions about the use of that trademark
have been published by Knuth; they at the least include the
requirements that you be happy with it as a TeX installation (which
implies some measure of compatibility), and also that it pass the trip
test. (There is also such a test for METAFONT as well.)
Now, it does not matter what the trademark usage requirements are.
They could say "nobody can call it TeX unless they are employees of
Stanford". They could say "you can only call it TeX if you stand on
your head and chant the Rig Veda". Regardless they are not copying
permissions on tex.web, nor could they be. And, it is a standing
principle of trademark law that it cannot cover the use of a trademark
as a functional element. So the use of the word "tex" as a command,
for example, can persist whether something passes the trip test or
not--as long as in non-functional contexts you don't call it TeX.
This is nothing special about TeX. You also can't call something
"Buick", "Windows", or "Rangers", in contents which might cause
confusion with the car, the software, or the sports team that usually
bear those trademarks. *No* such restriction could impact freeness,
in part because of the happy guarantee of trademark law that it cannot
impact functional elements.
So much for TeX and METAFONT. They are entirely free.
Computer Modern is a different matter. If the font files really
prohibit modifications, then that is unfortunate. Someone (perhaps
the tetex maintainer?) should ask Knuth for a change to his license.
For example, "you may modify these files, provided you do not call
them the Computer Modern fonts". Or perhaps a notification that the
files have been modified, akin to the GPL's similar requirement. If
there is no solution, then we would have to move them into non-free,
along with the other unmodifiable-but-otherwise-free things already
That leaves the stuff under the LaTeX public license. That's not
Knuth's stuff. And the license seems to be a hamfisted attempt to
achieve what Knuth has achieved for the TeX/METAFONT situation, but
with the unfortunate side-effect of losing DFSG-ness. Importantly,
TeX's trademark cannot cover functional elements, but the LaTeX
license *only* covers functional elements (filenames)!
So let's separate these three cases:
TeX/METAFONT: surely free.
Computer Modern Fonts: maybe not free, but we can ask Knuth.
LaTeX public license: still open for discussion.
Mittelbach and the other LaTeX people--please don't try and "help" us
understand TeX and CM licensing--so far, all you've done is post
half-truth and misinformation on those topics.