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

Re: TeX Licenses & teTeX (Was: Re: forwarded message from Jeff Licquia)

On Mon, Aug 05, 2002 at 09:22:24PM +0200, Frank Mittelbach wrote:
> no he was talking about these three, but that is the article published in
> tugboat and i don't see that it states at any point that it supersedes
> anything put on individual files by him. It wasn't written as a license
> statement it was written as a statement that he doesn't want to change those
> source files any further.

Fine; this doesn't leave Debian in any worse a position than the
contents of the original files do.

>  > Is a copyright holder allowed to re-license his work or clarify his
>  > licensing statements?  Is Knuth not allowed to come back to tex.web
>  > after 12 years and clarify his intent as he did in comp.text.tex?
> i don't know. I think he should be allowed so you last post seem to indicate
> that you think otherwise.

Hogwash.  Of course a copyright holder is permitted to revisit his
licensing decisions in general.  However, in free software we insist
that licenses be perpetual unless terminated for non-compliance.  In
other words, you can make a DFSG-free license non-DFSG-free simply by
adding terms like "the copyright holder can terminate this license for
any reason at any time" or "this license and your rights under it
terminate on February 5, 2003".

No one is challenging Knuth's right to license TeX, METAFONT or the
Computer Modern fonts however he pleases.  However, not all possible
licensing decisions he could make will be DSFG-free.  I would think that
this statement is uncontroversial.

> anyway, that statement wasn't made as an answer to "dear Prof Knuth,
> you licensing statements on files X YZ are strange could you please
> clarify them" it was made because people wanted him to extend TeX.

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

> no my contention isn't that. my contention is that you have to take that
> article plus those copyright statements and try to interpret them as he (Don)
> intended them to be interpreted. As for plain.tex that might in fact mean that
> one has to ask him to change it to something else.

You just told me that the Tugboat article had nothing to do with
licensing, but with people wanting to extend TeX.  You are asking me to
narrow my context, not expand it.  A narrow reading of the terms you
have taken time to quote from TeX itself lead me to conclude that TeX is
not DFSG-free.

If I were to read the Tugboat article as being relevant to licensing, I
would come to a different conclusion, but you've just told me not to do

"i don't see that it states at any point that it supersedes
anything put on individual files by him. It wasn't written as a license

"that statement wasn't made as an answer to "dear Prof Knuth, you
licensing statements on files X YZ are strange could you please clarify

"you have to take that article plus those copyright statements and try
to interpret them"

Er, uh, well, two consistent statements out of three isn't bad, I
suppose.  Henceforth I'll err on the side of caution and pretend like
the Tugboat article doesn't exist, which leaves us with the restrictive
and even self-contradictory license terms on TeX which you already

>  > Is Professor Knuth claiming a poor man's trademark on the strings "cmr10"
>  > or "cmr10.tfm"?
>  > 
>  > In other words, will Professor Knuth regard it as infrigment of his
>  > copyright of someone uses the filenames "cmr10" or "crm10.tfm" for
>  > something even if they aren't actually derived from his Computer Modern
>  > fonts, but rather original works?
> i doubt it, as long as a TeX program that would say
>  \font\foo=cmr10
> wouldn't pick up such fonts (or nofonts). but to be sure you would have to ask
> him.

>  > Will Professor Knuth regard it as infrigment of his copyright of someone
>  > uses the filenames "cmr10" or "crm10.tfm" for something that isn't a
>  > font at all?
> same answer.

Well, if Knuth or AMS aren't going to use trademark law to stop people
from coming up with original works and storing them on their computers
with the name "cmr10.tfm" -- which would violate DFSG 9 -- then to avoid
the scenario you describe above "as long as a TeX program...wouldn't
pick up such fonts" would be to forbid modification of TeX such that it
could not accidently use anything but Knuth-blessed cmr10 when the
"\font\foo=cmr10" directive is interpreted.  This is probably not only
practically impossible, but would violate DFSG 3.  Nothing the DFSG
permits a license holder to place restrictions on the aims, goals, or
ends of a modification to a work.

>  > No, you're saying the actual licenses are right there in the files in
>  > black and white, and that Debian cannot rely on Knuth's statements in
>  > comp.text.tex to interpret them.
> no

You're saying that very thing, and in the very message to which I am

"i don't see that it states at any point that it supersedes
anything put on individual files by him. It wasn't written as a license

"that statement wasn't made as an answer to "dear Prof Knuth, you
licensing statements on files X YZ are strange could you please clarify

If nothing in the Tugboat article supersedes what's in the individual
files, then all I need to conduct a DFSG analysis are the contents of
the individual files.  Apparently, nothing Knuth has ever said since
substantively modifies the licensing statements therein.

> i'm saying that my understanding of his intentions is that he wants to
> ensure that within a TeX system (ie program plus surroundings)  
>  \font\foo=cmr10
> refers to his CMR10 and 
>  \input plain
> to his plain.tex

In full generality, this is an impossible thing to assure.  As was
pointed out a very long time ago in this discussion, you cannot
effectively stop someone from creating a wholly original work called
"plain.tex" (which is thus itself not subject to Knuth's copyright).
You cannot stop someone else from taking this independent plain.tex and
putting it onto their computer where TeX will find it, without violating

Are you asking Debian to regard the following license as DFSG-free?

	Copyright 1996-2002 Software in the Public Interest, Inc.

	Permission is hereby granted, free of charge, to any person
	obtaining a copy of this software and associated documentation
	files (the "Software"), to deal in the Software without
	restriction, including without limitation the rights to use,
	copy, modify, merge, publish, distribute, sublicense, and/or
	sell copies of the Software, and to permit persons to whom the
	Software is furnished to do so, subject to the following

	1) The above copyright notice and this permission notice shall
	be included in all copies or substantial portions of the

	2) No file called "plain.tex" shall ever exist on a system where
	a TeX interpreter could possibly locate it.

>  > Not a legal restriction, at least not one with any precedent of which I
>  > am aware.
> it might well be that there is no legal precedent it might even be that it is
> not possible to ensure that legally (though I think it is).

Sure; people used to be able to legally sell themselves into slavery,
too; that didn't mean it was a morally justifiable activity.

> but be it as it may, I'm communicating (for the last time) what I
> understand his intentions have been (including that I don't think they
> got changed by that article you cited) and that precisely these
> intentions are what we try to capture with LPPL (except that we work
> hard on finding a way around that file renaming restriction if there
> is one). I also would guess (but that is my private assumption) that
> Don could be persuade to change from a hard file rename requirenment
> to some LPPL type licensing assuming there is an acceptable compromise
> found.

I'm willing to stipulate that the existing licensing on TeX, METAFONT,
and the Computer Modern fonts is not DFSG-free.  This is unfortunate.

>  > Note that there is a copyrighted work out there called "Microsoft Word".
>  > At least the last time I used it, which was many years ago, the "file
>  > name" that corresponded to this work was "MSWORD.EXE".  Microsoft has
>  > legal remedies available to them if someone packages a work of their own
>  > as "Microsoft Word", but to the best of my knowledge they can't attack
>  > someone just for calling a file "MSWORD.EXE".
> and there is no need for them, is there? forget the filerenaming requirement
> and stick with requiring that a work identifies itself to related works under
> a different name.

I am unwilling to accept as DFSG-free any license which places
restrictions on the implementation or interface of any computer system.
The only acceptable restrictions on modification are those which
preserve applicable legal notices, and which (within reason) identify
alterations of an original work *to human readers*.

No DFSG-free license is permitted to tell me what I may or may not name
a file, function, variable, method, class, or object, how I may perform
memory allocation, how many arguments I may pass to a function or
method, or what sequences of bits I am or am not permitted to transmit,
store, or receive.  The current DFSG doesn't permit any nonsense like
this and I will strongly oppose any effort to parse the DFSG such that
it is interpreted in such a way, or any effort to amend the DFSG such
that such restrictions are deemed acceptable.

> I don't insist anything, I only give you my understanding of what he means
> there and that it wasn't written to answer the question whether or not certain
> smaller works are allowed to identify themselves to the TeX system under their
> original name when in fact they are modified. 

Any work must be allowed to identify itself "to the TeX system" however
it pleases.  How it identifies itself to human beings is another matter.

> Beside that I would like to see your comments on my interpretation of
> your point 3.

You'll have to remind me which point 3 you're referring to here.

>  > There is not just what at "Don's heart".  There *are* practical
>  > limitations on what one can accomplish with a copyright license.
> there sure are, but are you now arguing that assuming, say, 
>  - the copyright notice on cmr10.mf is as Don wants it to be
>  - that it wouldn't serve the purpose of disallowing you to call a derived
>    file again cmr10.mf?
> it may or may not be DSFG-free, but that is not what i'm  asking here?

I don't understand what you are asking me.  I cannot speculate as to how
well the license terms inside cmr10.mf reflect Professor Knuth's

> i htink he is/was interested in both and that is whay trademarks have been
> picked up for the "collective works"

How the trademarks are used is up to the AMS and Addison-Wesley, it
would appear.  It would be quite worthwhile to determine what their
policies are regarding the usage of these trademarks.  They might or
might not impose any de facto DFSG problems on Debian.  In that respect
they are similar to patents.

>  > I *suspect* that Professor Knuth is not going to attempt to retract
>  > the copyright license (ambiguously stated as it is, with a very
>  > confusing reference to the "public domain") from someone who
>  > happened to use the string "cmr10" in a filename.  (Let's say
>  > Carnegie Mellon University developed a regular expression engine,
>  > and it saw 10 revisions -- should Knuth raise hell if they called
>  > it "CMR10" for "Carnegie Mellon Regex version 10"?)
> no, and i think you know it. nobody is after the CMR10 as a trademark as such
> but as an identification within a TeX system-

I don't think you can defend this namespace at a technical in any DFSG-free way.
The DFSG offers broad latitude to require that works not misrepresent
themselves to *human* consumers.

>  > I *hope*, though I am not sure, that Professor Knuth would not mind if
>  > someone modified the Computer Modern fonts and left the filenames the
>  > same but provided conspicuous notices that "THIS IS NOT DON KNUTH'S
>  > SAME" and distributed the fonts not as "Computer Modern", but as "Waylon
>  > Arbogast's Nifty Fonts".
> he would mind a lot (at least he did in the notso recent past) he would mind a
> lot the moment such fonts are set up to be inserted into a TeX system (or
> likely to be picked up by one). Unfort

Well, I'd say that's a problem.  It is a violation of DFSG 3, not to
mention the Free Software Foundation's freedoms 1 and 3, to place
restrictions on what changes can be made to a work of software at the
functional level.


>  > I suppose that if the TeX community shares your opinion, that will be
>  > necessary.
> I have no idea how to find out, you might ask at ctt, but if my opinion is a
> problem for you then eventually you have to ask Don himself, I guess.

Professor Knuth doesn't know me from Adam, so I would suggest that
someone serve as Debian's representative in any such action.  As the
package maintainer I would think C. M. Connelly should be involved in
such an appointment procedure.

G. Branden Robinson                |
Debian GNU/Linux                   |     Music is the brandy of the damned.
branden@debian.org                 |     -- George Bernard Shaw
http://people.debian.org/~branden/ |

Attachment: pgpiK6_ONiPfI.pgp
Description: PGP signature

Reply to: