oz@yetti.UUCP (Ozan Yigit) (09/26/87)
I have been following, with great interest, the recent chatter surrounding the C implementation (Ctex) of tex. Many institutions may have now discovered that the copies of Ctex obtained from Texas A&M is not (somehow) legal, as indicated by Tomas Rokiki, who has a copyright notice on Ctex sources. I would like to suggest that there has possibly been an abuse of the US copyright laws, and both Ctex and CommonTex cannot, in fact, carry anyone else's copyright except that of Don Knuth, and hence, carry the same distribution rights/restrictions [if any] of the original TeX. Why: If I am not mistaken, US copyright laws suggest that a translation of an originally copyrighted work carries the same copyright of the original - in other words, mechanical and/or hand translation does not obliterate the original copyright. Here is the header of tex.web: % This program is copyright (C) 1982 by D. E. Knuth; all rights are reserved. % Copying of this file is authorized only if (1) you are D. E. Knuth, or if % (2) you make absolutely no changes to your copy. (The WEB system provides % for alterations via an auxiliary file; the master file should stay intact.) One lawyer, Jordan J. Breslow, did address the issue of "derivative works" in his most enlightening paper [may be found under the doc directory of news 2.11 distribution]. He indicates that the translation of, say a copyrighted COBOL program to BASIC would be the infringement of the copyright owner's exclusive right to make derivative works. Is there something missing ?? Did Don Knuth drop his copyright on TeX ?? Now, it is possible that there exists a license arrangement between Tomas Rokiki and Don Knuth, which allows Tomas to claim copyright on Ctex. If such is the case, I would like to know about it, so that I can understand a request to flush all binaries and sources of Ctex obtained from Texas A&M. Same issue, of course, applies to the "copyleft" notice of Pat Monardo, for the CommonTex implementation. NOTE: This article is not meant to insult the authors of Ctex and CommonTex, and does not in any way meant to invalidate the appreciation they deserve for the substential amount of work to generate C versions of TeX. It is merely meant to encourage a discussion and/or clarification of the legal statues of such implementations, and to hopefully produce a better understanding of the copyright issues with regards to "derivative" works. Hence, any flames should be directed to /dev/null. oz
allbery@ncoast.UUCP (09/28/87)
As quoted from <167@yetti.UUCP> by oz@yetti.UUCP (Ozan Yigit): +--------------- | % This program is copyright (C) 1982 by D. E. Knuth; all rights are reserved. | % Copying of this file is authorized only if (1) you are D. E. Knuth, or if | % (2) you make absolutely no changes to your copy. (The WEB system provides | % for alterations via an auxiliary file; the master file should stay intact.) | | One lawyer, Jordan J. Breslow, did address the issue of "derivative works" | in his most enlightening paper [may be found under the doc directory of news | 2.11 distribution]. He indicates that the translation of, say a copyrighted | COBOL program to BASIC would be the infringement of the copyright owner's | exclusive right to make derivative works. +--------------- As I read this, this means that I cannot get a copy of TeX from the U. of Washington (presumably the original Pascal version straqight from Don Knuth) and run it through p2c, even for my own use. Is this correct? Must I either (a) try to find a Pascal compiler for my 3B1 (ha ha good luck sucker) or (b) be stuck with nroff for the rest of my life (AAAAAARRRRGGHHHHHHHHHH!!!!!!!!!!!!)? Or is Knuth willing to "authorize" a C translation and right to port (i.e. from the inevitable BSD-only first version to my System V)? Or am I just plain screwed? -- Brandon S. Allbery, moderator of comp.sources.misc {{harvard,mit-eddie}!necntc,well!hoptoad,sun!mandrill!hal}!ncoast!allbery ARPA: necntc!ncoast!allbery@harvard.harvard.edu Fido: 157/502 MCI: BALLBERY <<ncoast Public Access UNIX: +1 216 781 6201 24hrs. 300/1200/2400 baud>> "Mummy, what's an opinion?"
weening@Gang-of-Four.Stanford.EDU (Joe Weening) (09/29/87)
[I've added comp.text to the distribution of this message, and removed
comp.sys.atari.st.]
Yesterday I showed Don Knuth a copy of Brandon Allbery's message, and he
provided the following reply:
I intend to exercise my right of copyright only (and vigorously)
to the extent of denying the right to use the name TeX unless the
TRIP test has been passed and there are no extensions to the
language. The bit about not changing the file is for the WEB
version, not a derivative in another language ... but any changes
must preserve the TRIP-test validation.
This didn't completely answer the question as I understood it, so we had
a further exchange of messages as follows. (Text in brackets is mine.)
[JSW:]
Brandon Allbery asked whether he was prevented from running the
Pascal source of TeX through a Pascal-to-C generator. My
interpretation of this is that it is OK, since Pascal-to-C plus a
C compiler is equivalent to a Pascal compiler. As long as the
result passes the Trip test, it can use the name TeX. Do you
agree?
[DEK:]
Yes. (I don't believe there exists a valid Pascal-to-C translator
in the world, so the TRIP part is very important! Even Lyle
[Ramshaw]'s Pascal-to-Mesa translator had lotsa problems, and I
know DRF [David Fuchs] had to do a lot of handwork to get around
bugs in the translators he used.) (Also, TeX has found at least
one bug in every Pascal compiler it's been run on, I think, and at
least two in every C compiler...! So the code sometimes must be
patched to get around these system bugs.
The code is in the public domain. Anybody can copy it and change
it in any way if they don't call it TeX. If they want to call it
TeX, it's OK if they rewrite it in C or Ada or LISP or whatever,
as long as they don't add any extensions and as long as they pass
the TRIP test. My note in the program says that I don't want the
WEB files to be changed by anybody, because WEB has a better way
to make changes. As you know, the master WEB file still gets
updated sporadically...
The following is my opinion based on the above and experience with TeX,
so don't take it as being as "official" as the above comments.
(1) Don Knuth's main concern is that the TeX name not be applied to
anything that differs in a noticeable way from the WEB implementation.
Any program that passes the TRIP test and does not extend the language
may call itself TeX, but nothing else may.
(2) His WEB version of TeX is public domain. Now it seems that this
is contradicted by the copyright, but he clearly has chosen not to
exercise all of his rights. (I believe you can find published
statements similar to those above in TUGboat, the newsletter of the
TeX User's Group, but I haven't checked.)
(3) In addition to waiving the right to make copies of the WEB code,
Knuth has waived the right to make derivative works. In fact, I'm
sure that he encourages translation of TeX into any language if it
makes it possible to run on new systems, or improves its performance.
Many people have already done this. If you use WEB, you should use
the change-file mechanism rather than editing the main source file.
In summary, if you make changes in order to get TeX to work correctly,
you're in no trouble. And if you want to use the WEB code for TeX or
pieces of it in other programs, not called TeX, you're also free to do
so. Just don't use the name TeX unless your version passes the test.
Joe Weening
Internet: JSW@SAIL.Stanford.EDUrandy@peora.UUCP (09/29/87)
In article <4760@ncoast.UUCP>, allbery@ncoast.UUCP (Brandon Allbery) writes:
- As quoted from <167@yetti.UUCP> by oz@yetti.UUCP (Ozan Yigit):
-|% This program is copyright (C) 1982 by D. E. Knuth; all rights are reserved.
-|% Copying of this file is authorized only if (1) you are D. E. Knuth, or if
-|% (2) you make absolutely no changes to your copy.
...
-| One lawyer, Jordan J. Breslow, did address the issue of "derivative works"
...
-| He indicates that the translation of, say a copyrighted
-| COBOL program to BASIC would be the infringement of the copyright owner's
-| exclusive right to make derivative works.
-
- As I read this, this means that I cannot get a copy of TeX from the U. of
- Washington (presumably the original Pascal version straqight from Don Knuth)
- and run it through p2c, even for my own use.
Does this mean a compiler violates copyright every time it generates
an intermediate (or assembly) language? Does it matter is the intermediate
is stored on disk? I think that using p2c could be viewed just as another
"compiling" step. Now, if you started modifying your C source you'd
probably be creating a derivative work.chongo@amdahl.amdahl.com (Landon Curt Noll) (09/30/87)
As quoted from <167@yetti.UUCP> by oz@yetti.UUCP (Ozan Yigit): >+--------------- >| % This program is copyright (C) 1982 by D. E. Knuth; all rights are reserved >| % Copying of this file is authorized only if (1) you are D. E. Knuth, or if >| % (2) you make absolutely no changes to your copy. (The WEB system provides >| % for alterations via an auxiliary file; the master file should stay intact. How do they expect to enforce the 'no changes' section of this copyright? Are they going to checksum each version that they find? Are they going to drag someone into court because their Pascal compiler makes an altered copy of TeX by rewriting it in assembler? Do they really think that someone won't edit the source to get it to compile on their compiler? I can understand how they can prevent you from making and/or distributing copies. I can also understand that they can prevent you from distributing alter sources or derived sources. But a 'no changes' to your copy clause? The copyright gets even more BOGUS when you consider that Knuth claims that TeX is in the public domain! No, I don't have a copy of TeX, nor do I plan to as long as it has bogus copyright restrictions on it. chongo <maybe i'll take a MIX version of TeX :-}> /\oo/\ -- [views above shouldn't be viewed as Amdahl views, or as views from Amdahl, or as Amdahl views views, or as views by Mr. Amdahl, or as views from his house]
rokicki@rocky.UUCP (09/30/87)
> % This program is copyright (C) 1982 by D. E. Knuth; all rights are reserved > % Copying of this file is authorized only if (1) you are D. E. Knuth, or if > % (2) you make absolutely no changes to your copy. (The WEB system provides > % for alterations via an auxiliary file; the master file should stay intact. > > How do they expect to enforce the 'no changes' section of this copyright? > Are they going to checksum each version that they find? Are they going > to drag someone into court because their Pascal compiler makes an altered > copy of TeX by rewriting it in assembler? Do they really think that someone > won't edit the source to get it to compile on their compiler? Please, give Don a bit more credit than that. First of all, the Pascal compiler's transformation into assembly creates a file which is not a copy of the original, and it is also a necessary step to get the program to work on a machine. Secondly, as the message above states, the WEB system allows you to make any necessary changes in a small auxillary file. The WEB system's macro processor allows global changes to be stated once, and the program is written so as to isolate those system dependent changes. This also allows you to update your version of TeX simply by getting the new master copy and using the same auxillary file with your local changes. And, no, nobody is going to checksum each version they find. Rather, no one may distribute a version of TeX that does not fit the following three criteria: 1. It must pass a standard validation suite called `trip', with identical results to the master copy. This test suite exercises all of the lines in the code except a few seriously fatal error messages, including the limitations of the program. 2. It work reasonably well and identicallly on `normal' documents (which trip certainly is not.) 3. The implementer must be happy with his implementation. Only (1) and (2) of the above can be tested, but if a version is found that violates either of these, and is called TeX, its distribution will be stopped. -tom
allbery@ncoast.UUCP (10/03/87)
As quoted from <2628@peora.ccur.com> by randy@peora.UUCP: +--------------- | In article <4760@ncoast.UUCP>, allbery@ncoast.UUCP (Brandon Allbery) writes: | - As quoted from <167@yetti.UUCP> by oz@yetti.UUCP (Ozan Yigit): | -| He indicates that the translation of, say a copyrighted | -| COBOL program to BASIC would be the infringement of the copyright owner's | -| exclusive right to make derivative works. | - | - As I read this, this means that I cannot get a copy of TeX from the U. of | - Washington (presumably the original Pascal version straight from Don Knuth) | - and run it through p2c, even for my own use. | | Does this mean a compiler violates copyright every time it generates | an intermediate (or assembly) language? Does it matter is the intermediate | is stored on disk? I think that using p2c could be viewed just as another | "compiling" step. Now, if you started modifying your C source you'd | probably be creating a derivative work. +--------------- You chopped my comment about porting. If I fix a BSD-ism (if I find one), I have violated copyright. Remember, some of us use and LIKE System V. -- Brandon S. Allbery, moderator of comp.sources.misc {{harvard,mit-eddie}!necntc,well!hoptoad,sun!mandrill!hal}!ncoast!allbery ARPA: necntc!ncoast!allbery@harvard.harvard.edu Fido: 157/502 MCI: BALLBERY <<ncoast Public Access UNIX: +1 216 781 6201 24hrs. 300/1200/2400 baud>> "...he calls _that_ a `little adventure'?!" - Cmdr. Ryker