Re: TeX Licenses & teTeX (Was: Re: forwarded message from Jeff Licquia)
Boris Veytsman <borisv@lk.net> writes:
> Now you are tempted to consider Knuth to be ignorant -- worse, to be
> an evidently stupid person who does not realize he is ignorant and
> pontificates about things he has no business to talk about. Well,
> since it is established that Knuth is a great talent or maybe even a
> genius, this conclusion is highly improbable. If I were you, I'd
> consider a much more probable interpretation. Namely, that Knuth
> perfectly KNOWS what he is talking about -- but his thoughts are too
> profound to be understood without some work.
Why must you toss words into my mouth that I manifestly didn't say?
Suffice it to say that you mischaracterize me, and your little bits of
psychological advice are not welcome.
I think you are a little too bit overawed by the "great man".
> As I understand it, Knuth put in the public domain the *code* of TeX,
> Metafont, CM fonts etc. Fragments of his code, his creative ideas and
> insights are freely used in many derived works. However -- and here is
> the most profound idea -- Knuth realizes, that TeX is not a sum of
> pieces of code, it is an entity, a whole. This entity is what Knuth
> did not want to put in PD. He wanted to retain the full rights on this
> entity.
Blech. You yourself don't understand the law either. No big
sin--most people don't.
The actual license on tex.web prohibits this activity, unless it is
indeed in the public domain. But if it is in the public domain, then
it's not true that "changing it is not allowed". It is *NOT POSSIBLE*
to simultaneously assert that tex.web is in the public domain and also
assert that changing it is legally prohibited. They are INCONSISTENT.
> The proof that Knuth perfectly understood what he was doing is in the
> fact that he trademarked the name TeX -- something that LaTeX3 should
> do with respect to LaTeX. For you this trademarking and putting TeX in
> public domain may seem contradictory -- for me it makes sense.
I *never* said this was contradictory. You have a *dismal* record at
even getting my own words right, it's no *wonder* that the
complexities of the copyright law escape you.
Which is no big sin. But then, *please*, I *beg* you, STOP trying to
"help us", because you're only serving to muck the works up.
> Like White Knight from "Through the Looking Glass", Knuth knows the
> difference between the song, its name, how the song is called and how
> the name is called. He distinguishes TeX and TeX code. White Knight
> might seem to be stupid -- but do not be fooled. He is just profound.
Indeed, he distinguishes TeX and TeX code. Here is what the *actual
legal licenses* say:
1) tex.web cannot be changed. I believe his statements that it's in
the public domain are erroneous, because copyright law generally is
rather jealous about protecting author's rights. So if he
simultaneously tries to limit something that can only be limited by
copyright, and also asserts that it's in the public domain, the law
probably discounts the latter, and for safety, so should we. So
tex.web is not in the public domain, and it cannot be modified.
2) Patches to tex.web are just fine, you can patch in however you
like, provided the patch is distributed as a separate file. It's
worth knowing that this is not by any means automatic--copyright
law does *not* generally say that people have this right, but in
the case of tex.web, it has been explicitly granted, so we're fine.
3) The name "TeX" is a trademark. As such, you cannot use it unless
you are the trademark owner or you have the owner's permission.
However, there are many exceptions in law, and one of them is the
"functional element" test. You may freely use a trademark if doing
so is a genuine functional element of some process or mechanism.
Since computer command names on GNU/Linux systems are functional
elements, you can call any program you like "tex" without violating
the trademark. However, banner messages are not functional
elements, so you probably cannot have a banner message say "This is
TeX" unless you conform to the owner's restrictions on the use of
that trademark.
4) The requirements for calling something "TeX"--for using the
trademark--are that you must be happy with the system as an TeX
installation, and that it must pass the trip test. Note that the
trip test uses its own font metric files. Note that there is no
requirement that the system in question have any particular
relationship to tex.web.
5) Taking bits of code out of tex.web and using them in other projects
is actually *not* allowed by the actual license on tex.web. This
is annoying in the extreme, but we're stuck with it, because it's
the condition on the code.
6) *NOBODY* installs unmodified tex.web (that is, without patching it)
as a TeX system. So the claim that "any modification requires
renaming" is manifestly not true.
7) The rules for mf.web and the trademark "METAFONT" are parallel to
the rules for tex.web and the trademark "TeX", but they make no
reference to each other, and function as independent systems,
legally speaking.
8) The Computer Modern trademark has a more restrictive use that the
METAFONT and TeX trademarks, and can refer only to the specific
files unmodified that Knuth mentions. Again, however, trademark
restrictions do not apply to functional elements, and therefore
something may be called cmr10.tfm without any penalty.
9) cmr10.mf comes with a very restrictive notice: the code may not be
changed at all without changing the filename. This might seem to
violate DFSG-4, since there is no permission to use patches.
However, the METAFONT language is flexible, and so it's possible to
get the same effect as using patches, so Computer Modern just
barely eeks into DFSG. The restrictions on the name "cmr10.mf" are
thus as follows: you may name something cmr10.mf if EITHER it
contains no part of the Knuth cmr10.mf file, OR if it is exactly
the Knuth cmr10.mf file. If you meet the former condition, then
since you are not copying Knuth's cmr10.mf, you are not subject to
its copyright restrictions, and since filenames are functional
elements, you are not subject to the trademark restriction.
10) The rules on cmr10.mf are entirely separate from the TeX rules.
Nothing in the TeX rules refers to Computer Modern fonts; the
actual test (trip) does not in any way refer to the computer
modern fonts.
11) The rules on cmr10.mf in no way are rules on the names "cmr10.tfm"
or various names for bitmap fonts. If I change cmr10.mf, I must
name it something else (say, "tb-cmr10.mf"). But there is no
restriction placed on what I can do with the result. Including,
for example, compiling it with METAFONT, and renaming the output
"cmr10.tfm" if I like.
12) Taking advantage of these permissions in the way Knuth wishes we
wouldn't would generally be an anti-social thing. For this
reason, social pressure is appropriately used to get people not
to--as is in fact Knuth's actual practice, supporting my
interpretation. However, that these rights exist is essential for
us--not because we intend to exercise them, but because we lack
the arrogance to insist that they will never be important, and
because we have seen them turn out to be important for other
software again and again.
Now, if you want to contest these, which are founded on the
*actual*licenses* on the *actual*files*, then that's great. But
appealing to "what Knuth wants" won't do. I want world peace, but the
copyrights I've placed on my packages don't achieve it. I want people
not to lie to each other, but the copyrights on my packages don't
restrict lying at all. What matters is the actual license texts.
Please quote *those*.
If you want to appeal to some mystical Gestalt that the Great Knuth
has conceived, with his inimitable brilliance to awe all the copyright
attorneys of the world, then there's going to be a problem, because
the law already works a certain way, and no matter how clever some new
scheme might be, it doesn't just spring into being because the Great
Knuth wants it that way.
Thomas
Reply to: