hankd@pur-ee.UUCP (Hank Dietz) (07/25/89)
I have thought about this "copylefting" business quite a bit lately. T. Parr and I are on the verge of releasing Antlr (an extended-BNF LL(1) compiler-compiler which is part of his MS thesis), so it has been more than a theoretical concern.... When I first saw the info on copylefting I thought to myself, "Well, that should stop the scum from selling what we did for free." However, I have since seen that "the scum" don't really care if it is copylefted -- which only makes sense, because none of them ever seemed to care about copyrights before. They know that prosecution is mostly an empty threat. I have also seen that copylefts send a shudder down the spine of nearly every major corporation. Since they fear that they'll accidentally lose rights to some vital piece of in-house developed software by some nieve employee using a copylefted tool, many companies simply refuse to touch copylefted software... they will not even allow it on their machines. The propagation of copyleft rights via use of copylefted tools is an absurd power trip: imagine a fellow inventing a new type of nail and then claiming ownership/control of all houses built using at least one such nail because the assemblage of the house is fundamentally the use of the nails, not a function of the lumber used. This certainly violates the spirit of pure academic research as I see it; all we're supposed to want is proper credit for our good work, not a claim on somebody else's efforts. Hence, I've slowly become convinced that true "public domain" (PD) is the only way to go -- perhaps augmenting this by establishing a user group at the time of release. Companies are a very large and important user community, and I don't want to exclude them from using our software. Further, if we publicize the PD software enough, few people will pay big money for a commercial rip-off of it (maybe somebody else will get rich off of our work, but in some sense that's what academia is all about ;-). Anyway, here's what I think we will be doing: a pure PD release with a credit/modification history which must be maintained, but which carries no further rights. In other words, people would be free to sell modified versions of our software... but they must acknowledge the lineage of their product in some way. I'm not quite sure of the details... but it shouldn't be anything that would frighten away any potential users. And that's what this is all about. -hankd@ecn.purdue.edu
mwm@eris.berkeley.edu (Mike (I'll think of something yet) Meyer) (07/25/89)
In article <12344@pur-ee.UUCP> hankd@pur-ee.UUCP (Hank Dietz) writes:
<The propagation of copyleft rights via use of copylefted tools is an absurd
<power trip: imagine a fellow inventing a new type of nail and then claiming
<ownership/control of all houses built using at least one such nail because
<the assemblage of the house is fundamentally the use of the nails, not a
<function of the lumber used.
That's an absurd analogy. Try this one for size: a fellow invents a
new type of hammer, and allows anyone to use it to build houses so
long as they give away the plans for the house.
<This certainly violates the spirit of pure
<academic research as I see it; all we're supposed to want is proper credit
<for our good work, not a claim on somebody else's efforts.
You should take a a carefull look at copyleft as you've obviously
misinterpreted the entire thing. In any case, copyleft is much more in
the academic spirit than "publishing" software through the normal
channels; it makes others share the spirit if they use your tools and
want recognition for what they've done.
<Further, if we publicize the PD software enough, few people will pay big
<money for a commercial rip-off of it (maybe somebody else will get rich off
<of our work, but in some sense that's what academia is all about ;-).
Um, who's going to pay for that publicity?
<Anyway, here's what I think we will be doing: a pure PD release with a
<credit/modification history which must be maintained, but which carries no
<further rights.
Nice idea. To bad you can't do that. If you release something into the
public domain, then the public can do whatever they want with it -
compile & sell it, modify it in anyway they want and copyright that,
etc.
If you wish to prevent people from doing something to a piece of
software (say, removing your name & adding their own, or fixing bugs
and/or making enhancements and not telling you about them), you _must_
copyright the software. You can then grant the rights to copy it with
whatever conditions you want - for instatce, that it include a
modification history. Of course, you're now requireing others to do
work if they want to let others use that work. Isn't that nasty of
you?
<In other words, people would be free to sell modified
<versions of our software... but they must acknowledge the lineage of their
<product in some way.
Sounds like a copyleft to me. There, people are required to make
_original_ source of the copylefted software, or any derivative work,
to anyone they give binaries. They are also prohibited from forbidding
others to give away that software. They even require that
modifications be clearly marked (less work than your modification
history).
<mike
--
Tell me how d'you get to be Mike Meyer
As beautiful as that? mwm@berkeley.edu
How did you get your mind ucbvax!mwm
To tilt like your hat? mwm@ucbjade.BITNETlee@uhccux.uhcc.hawaii.edu (Greg Lee) (07/25/89)
From article <12344@pur-ee.UUCP>, by hankd@pur-ee.UUCP (Hank Dietz): " ... " The propagation of copyleft rights via use of copylefted tools is an absurd " power trip: ... This certainly violates the spirit of pure " academic research as I see it; all we're supposed to want is proper credit " for our good work, not a claim on somebody else's efforts. Yes. And software developed with public monies ought to be practically available to profit-making companies as well as to others. I wish Stallman all the best in his crusade, but since I have my hand in the public till, I don't think it would be proper for me to use a copyleft. " ... " Anyway, here's what I think we will be doing: a pure PD release with a " credit/modification history which must be maintained, but which carries no " further rights. As a practical matter I think it is reasonable for you to expect the courtesy of keeping your names associated with the code. It would not be a right, though, if I understand what PD means. Greg, lee@uhccux.uhcc.hawaii.edu
pardo@june.cs.washington.edu (David Keppel) (07/25/89)
>[Ongoing discussion of copyleft vs. public-domain (PD).]
A copyleft requires (most :-) people who redistribute your program to
redistribute the source. I believe (personally, IMHO) that this has
at least two important `academic' advantages over public-domain
software.
* There is a strong `sharing' incentive. If I find a bug, I share it.
If you find a bug, you share it. Bugs get fixed. With PD code, a
`branch' that goes proprietary will stop propigating bug fixes back
to the main tree.
* The copyleft is propigated with the code. If the copyleft includes
some kind of credit to the original authors, then the Great
Inspiration (tm) of users along the way will get passed back to the
creator. Aside from the ``warm fuzzy'' feeling of hearing from
people, you also get good ideas.
Finally, remember that GNU's copyleft has a very specific set of
goals. If you wish, you can `copydown' your programs in a way that is
more or less restrictive, as you please.
;-D on ( Parameter passing by copy-in copy-left ) Pardo
--
pardo@cs.washington.edu
{rutgers,cornell,ucsd,ubc-cs,tektronix}!uw-beaver!june!pardohankd@pur-ee.UUCP (Hank Dietz) (07/26/89)
In article <26633@agate.BERKELEY.EDU> mwm@eris.berkeley.edu (Mike (I'll think of something yet) Meyer) writes: >In article <12344@pur-ee.UUCP> hankd@pur-ee.UUCP (Hank Dietz) writes: .... I'd just like to clarify that public domain (PD) software DOES carry a copyright. However, the PD copyright notice explicitly grants all rights to the public at large -- except the right to be identified as the creator of the original software.... This is exactly how research results published in journals are treated: the text is copyrighted, the authors must be given due credit for their work, but everyone is free to build upon that work. This certainly isn't all a copyleft implies. Copyleft implies a profound form of transitive ownership rights, and it is this aspect I find disturbing. Someone who modifies your PD code shouldn't have any obligation to you other than to acknowledge, by appropriate citation, your prior work. Of course, this is just my opinion. -hankd
chuq@Apple.COM (Chuq Von Rospach) (07/26/89)
>I'd just like to clarify that public domain (PD) software DOES carry a >copyright. However, the PD copyright notice explicitly grants all rights to >the public at large -- except the right to be identified as the creator of >the original software.... Unfortunately, if you're going to clarify, try to do so correctly. Public domain has a very specific meaning: there is no restrictions to the rightss of the work. In other words, no copyrights. You can create a work that has a copyright that allows complete redistribution (or redistribution with few limitations) but that work is not public domain. Chuq Von Rospach =|= Editor,OtherRealms =|= Member SFWA/ASFA chuq@apple.com =|= CI$: 73317,635 =|= AppleLink: CHUQ [This is myself speaking. No company can control my thoughts.]
henry@garp.mit.edu (Henry Mensch) (07/26/89)
lee@uhccux.uhcc.hawaii.edu (Greg Lee) wrote:
->Yes. And software developed with public monies ought to be practically
->available to profit-making companies as well as to others. I wish
->Stallman all the best in his crusade, but since I have my hand in the
->public till, I don't think it would be proper for me to use a copyleft.
why not? why can't something done with public funds benefit
*everyone*? do not profit-making companies contribute to said public
till? why the artificial distinction?
# Henry Mensch / <henry@garp.mit.edu> / E40-379 MIT, Cambridge, MA
# {decvax,harvard,mit-eddie}!garp!henry / <henry@sics.bu.oz.au>lee@uhccux.uhcc.hawaii.edu (Greg Lee) (07/26/89)
From article <358@mit-amt.MEDIA.MIT.EDU>, by henry@garp.mit.edu (Henry Mensch): " ->Stallman all the best in his crusade, but since I have my hand in the " ->public till, I don't think it would be proper for me to use a copyleft. " " why not? The University of Hawaii is a state university. A purpose of the state in supporting the university is to benefit commerce and industry in Hawaii. Copylefted code is effectively unavailable to commercial enterprise, and so if an employee of the university employs a copyleft, he acts contrary to this purpose. " why can't something done with public funds benefit *everyone*? It can. Software placed in the public domain does (if it's of use). " do not profit-making companies contribute to said public till? Indeed they do. All the more reason to make the work of public employees available to them. I'm beginning to wonder whether you read what I wrote. " why the artificial distinction? Between what? Greg, lee@uhccux.uhcc.hawaii.edu
peter@ficc.uu.net (Peter da Silva) (07/26/89)
In article <26633@agate.BERKELEY.EDU>, mwm@eris.berkeley.edu (Mike (I'll think of something yet) Meyer) writes: > That's an absurd analogy. Try this one for size: a fellow invents a > new type of hammer, and allows anyone to use it to build houses so > long as they give away the plans for the house. Since there are people who make their living designing and selling plans for houses, this is a good analogy. It's an argument against the FSF copyleft, though. > <Further, if we publicize the PD software enough, few people will [ buy the commercial ripoff ] > Um, who's going to pay for that publicity? The same people who, through the goodness of their hearts, will pay for the lawyers to protect copylefts and for the development of copylefted software. Oh, and computer and trade magazines. i.e., there are MORE resources available to publicise PD software than to protect FSFware. -- Peter da Silva, Xenix Support, Ferranti International Controls Corporation. Business: peter@ficc.uu.net, +1 713 274 5180. | "...helping make the world Personal: peter@sugar.hackercorp.com. `-_-' | a quote-free zone..." Quote: Have you hugged your wolf today? 'U` | -- hjm@cernvax.cern.ch
oz@yunexus.UUCP (Ozan Yigit) (07/27/89)
In article <12344@pur-ee.UUCP> hankd@pur-ee.UUCP (Hank Dietz) writes: > >Anyway, here's what I think we will be doing: a pure PD release with a >credit/modification history which must be maintained, but which carries no >further rights. In other words, people would be free to sell modified >versions of our software... but they must acknowledge the lineage of their >product in some way. I'm not quite sure of the details... but it shouldn't >be anything that would frighten away any potential users. And that's what >this is all about. A Pure PD release would not guarantee that all derivatives or the modified versions carry the credit/modification history. In fact, the only thing it DOES guarantee is that the *original* is ALWAYS PD, and it cannot, in any way, be restricted, as *everyone* has equal rights to the original. I have no comment on Copyleft. I do not use copyleft, and try to avoid anything copylefted. I believe what you probably need is something "QUASI-PD" like released BSD stuff, which grants all rights (as in PD) *except* the right (!) to misrepresent the software in any way, and the right to remove the authorship notice. See Berkeley "Released" sources for a copyright that is probably quite close to what you want. For example: | | Copyright (c) 2000 Midnight-Oil Ltd. | All rights reserved. | | redistribution and use for any purpose whatsoever in source and | binary forms are permitted provided that the above copyright no- | tice and this paragraph are retained in all documentation and | source materials, blah blah blah ... | oz -- They are like the Zen students who, Usenet: oz@nexus.yorku.ca when the master points at the moon, ......!uunet!utai!yunexus!oz continue to stare at his finger.... Bitnet: oz@[yulibra|yuyetti] P. da Silva Phonet: +1 416 736-5257x3976
jeff@aiai.uucp (Jeff Dalton) (07/28/89)
lee@uhccux.uhcc.hawaii.edu (Greg Lee) wrote: >->Yes. And software developed with public monies ought to be practically >->available to profit-making companies as well as to others. I wish >->Stallman all the best in his crusade, but since I have my hand in the >->public till, I don't think it would be proper for me to use a copyleft. Ok, so you use public domain or else a copyright that gives away most of the rights. But some people seem to think it's wrong for FSF to take that software and copyleft their improvements. Why is it ok for a profit-making company to keep their improvements to public domain software completely secret and not ok for FSF to put on their kind of restriction?
lee@uhccux.uhcc.hawaii.edu (Greg Lee) (07/29/89)
From article <634@skye.ed.ac.uk>, by jeff@aiai.uucp (Jeff Dalton): >... >But some people seem to think it's wrong for FSF to take that software >and copyleft their improvements. Why is it ok for a profit-making >company to keep their improvements to public domain software completely >secret and not ok for FSF to put on their kind of restriction? Taking the FSF view, it is of course not ok to keep such secrets (assuming the company sells the software). I don't see anyone saying that the FSF restrictions are unethical. One might wonder whether they best serve FSF's stated goals. If, for instance, C programs derived by running bison could remain proprietary, bison would perhaps be more widely used, more widely tested, and more people could contribute to its development. Greg, lee@uhccux.uhcc.hawaii.edu
jejones@mcrware.UUCP (James Jones) (07/31/89)
In article <4449@uhccux.uhcc.hawaii.edu> lee@uhccux.uhcc.hawaii.edu (Greg Lee) writes: >If, for instance, C programs derived by running >bison could remain proprietary, bison would perhaps be more >widely used, more widely tested, and more people could contribute >to its development. Exactly. Does anyone know what happened to TRAC (which is a trademark of some company), which Calvin Mooers well and thoroughly protected via copyright? James Jones
preston@titan.rice.edu (Preston Briggs) (08/01/89)
In article <1308@mcrware.UUCP> jejones@mcrware.UUCP (James Jones) writes: >Exactly. Does anyone know what happened to TRAC (which is a trademark >of some company), which Calvin Mooers well and thoroughly protected via >copyright? > > James Jones Ted Nelson's "Computer Lib/Dream Machines" is out in a new edition. In it he brings us slightly up-to-date, but I can't remember many details. TRAC is/was a trademark of Rockford Research. Mooers has (or is with) a new company now, and has a new variation of the langauge too. Anybody else know something more concrete? TRAC was pretty cool, especially for the old 8-bit micros with only 64K of memory. A fairly complete and efficient interpreter would fit in about 2.5K on a 6809. I expect an interpreter on a modern system would be really fast and fun. I always though the system needed some way to control name-scoping though. Regards, Preston Briggs
chasm@attctc.Dallas.TX.US (Charles Marslett) (08/03/89)
In article <634@skye.ed.ac.uk>, jeff@aiai.uucp (Jeff Dalton) writes: > lee@uhccux.uhcc.hawaii.edu (Greg Lee) wrote: > >->Yes. And software developed with public monies ought to be practically > >->available to profit-making companies as well as to others. I wish > >->Stallman all the best in his crusade, but since I have my hand in the > >->public till, I don't think it would be proper for me to use a copyleft. Greg's point here is that he feels he cannot restrict his code with a copyleft. Please, people, the point is that the copyleft IS a copyright [in spite of the fancy word games]. It does restrict the use of the software to a subset of those who would otherwise use it. I do not complain that Microsoft does not distribute the source to OS/2 (I don't know anyone who would care, but I digress). I also do not complain that FSF restricts the use of the libraries it distributes with GCC (here I care, but it is their right to restrict access and use of their code). > Ok, so you use public domain or else a copyright that gives away > most of the rights. That's right, or I use commercial software that gives me full rights to sell the results of my work (that is, I use AT&T yacc rather than bison). > But some people seem to think it's wrong for FSF to take that software > and copyleft their improvements. Why is it ok for a profit-making > company to keep their improvements to public domain software completely > secret and not ok for FSF to put on their kind of restriction? The note by Greg Lee explicitly noted that it is NOT ok for him to keep his improvements from anyone contributing to the government of Hawaii (I think). BTW, the FSF license does not prohibit keeping improvements secret -- it prohibits their distribution without full source, a completely different animal. Like its commercial equivalent (the shareware license), a copyleft is very hard to enforce, but the ethical individual should abide by the author's wishes. And in some cases this means that a tool (even a "free" tool) is not accessable. Charles Marslett
chase@Ozona.orc.olivetti.com (David Chase) (08/04/89)
In article chasm@attctc.Dallas.TX.US (Charles Marslett) writes: >Like its commercial equivalent (the shareware license), a copyleft is very >hard to enforce, but the ethical individual should abide by the author's >wishes. And in some cases this means that a tool (even a "free" tool) is >not accessable. True, but one point (or implication of this) is missed. Note that the GNU "license" has not been tested in court. If I do some work derived from GNU software (and distribute it) thinking that the GNU license protects me from getting ripped off in some unethical way (or from being held liable in some suit), I'd end up really really sad if that weren't the case. That is, I know that I'm ethical, but do I have any guarantee that (a) all users of my code will be ethical and (b) I can do anything about it if they aren't? If I could require that users sign a license before getting a copy, I'd be in much better shape; this shrink-wrap stuff sure cuts down on the time and hassle, but what if it carries no legal force? What if it only carries legal force if I spend a zillion dollars on lawyers to haul someone into court? In this uncertain situation, of course, the ethical thing (if you dislike legal uncertainty) is to *not* derive interesting things from GNU software; I'll cheerfully make incremental improvements and send them back to FSF, but I won't commit massive amounts of time or money to a project, because I might end up with *no* control over its use. Understand, too, that what the FSF lawyer says means n-o-t-h-i-n-g to me. Since when does anyone take advice from someone else's lawyer? Not this fool, no sir. It is possible to do standard licensing in a do-it-yourself fashion, yes, but the GNU copyright/license is most definitely not standard. It takes more time to get things signed, and it might even cost more money to handle the paperwork, but a file full of signatures on licenses would be extremely useful if it were ever necessary to haul someone into court. The "copyleft" need not be hard to enforce. Understand, please, that I'm not bashing Stallman or the FSF or their goals. I *am* bashing (1) their casual approach to licensing and (2) all the people who claim that people reluctant to base work on GNU software are greedy software-hoarding idiots. David