How To ???
John R. Strohm
Does anyone have IN ONE PLACE a list of the pieces I have to download and a
procedure for putting them together to end up with a running ITS on a
simulated KS10 or KLH10?
Same question for TOPS-10 (or TWENEX).
I already have the emulators. I need the OS and applications and an
"assembly procedure".
Thanks,
--John
Zane H. Healy
I've got a webpage that should point you to everything you'll need.
http://www.aracnet.com/~healyzh/pdp10emu.html
The page has pointers to emulators, software, and documentation (actually
the page has pointers to just about every piece of PDP-10 related info on
the web).
BTW, TWENEX isn't available, but TOPS-20 is. TOPS-20 V7 on KLH10 is quite
nice!
Zane
Douglas H. Quebbeman
news:b77ac...@enews1.newsguy.com...
I thought TWENEX was a nickname for TOPS-20 that
acknowledged its TENEX heritage...
??
John R. Strohm
TOPS-20 *IS* TWENEX.
TENEX was an early paging system for a KA-10 with aftermarket paging
hardware. BBN did the hardware and the OS.
DEC later snarfed a lot of the flavor of TENEX for TOPS-20, and it got
tagged with the monicker "TWENEX" by an unappreciative audience.
jmfb...@aol.com
<F07C912B8269410C.4EB6414A...@lp.airnews.net>,
The Monitor Installation Guide (MIG) will tell you how to install
a TOPS-10 system. This does not include lanugages that were sold
and shipped on separate tapes (e.g., COBOL and FORTRAN). These
products had their own installation guides; most of these instructions
told the site to
COPY REL:=*.REL
COPY SYS:=*.EXE
COPY UNV:=*.UNV
I don't recall where the RELs were expected. Our standards would
have them put in REL: but the language people occasionally had
standard-moments.
There was never any canned procedure _shipped_ to build everything.
There was no request for this since sites tended to do individual
builds. The procedures I developed for Release Engineering to
build the stuff that went on the CUSP tape may be buried on
iron oxide; the save set is missing two tapes.
/BAH
Subtract a hundred and four for e-mail.
John Saeger
maybe the source code? The currently available GPL'd program Maxima is
based on Macsyma code from the 1980's, I thought now that emulators are
available it might be fun to play with the original version. I also found
some original Macsyma manuals in our library and I'm wondering if it's
worth the trouble of copying or scanning them. If they're already
available I won't bother. I found Introductory Macsyma which has an intro
to ITS and Macsyma and has some pretty cool hand-drawn pictures. I also
found a two volume Macsyma reference manual for version 10.
John
Mark Crispin
> Does anybody know if runnable PDP-10 Macsyma is available anywhere? And
> maybe the source code?
The only known runnable PDP-10 Macsyma is the TOPS-20 port of Macsyma 304.
Unfortunately, it has a 1983 Symbolics copyright and trade secret notice
for "enhancements", meaning that it can't be distributed without the
permission of the current owner of the assets of Macsyma, Inc.
The ITS archive had a copy of MIT (not Symbolics) Macsyma source, but I
don't know if that is available or what needs to be done to whip it into
shape.
> The currently available GPL'd program Maxima is
> based on Macsyma code from the 1980's, I thought now that emulators are
> available it might be fun to play with the original version.
There seem to be three alternatives for TOPS-20 Macsyma:
1) get Maxima running under TOPS-20 Common Lisp (CLISP)
2) figure out how to get what exists of sources from the ITS archive into
a running Maclisp-based Macsyma on TOPS-20
3) get permission from the owner of the assets of Macsyma, Inc. to
distribute the old TOPS-20 Macsyma.
I have discussed this with several individuals, and the general concensus
is that (1) appears to be much easier than (2) or (3).
Since Maxima is under GPL, that may mean that a TOPS-20 Maxima would have
to be released individually, rather than bundled with a TOPS-20
distribution, because of the viral aspects of GPL. Is there a version of
Maxima that has a non-viral open source license (such as free-fork)?
-- Mark --
http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Paul Rubin
> Since Maxima is under GPL, that may mean that a TOPS-20 Maxima would have
> to be released individually, rather than bundled with a TOPS-20
> distribution, because of the viral aspects of GPL. Is there a version of
> Maxima that has a non-viral open source license (such as free-fork)?
The GPL does not prevent that kind of bundling, as long as TOPS-20
and Maxima are basically independent programs. It's like distributing
the Windows port of GCC.
Mark Crispin
> The GPL does not prevent that kind of bundling, as long as TOPS-20
> and Maxima are basically independent programs. It's like distributing
> the Windows port of GCC.
Clause 2(b) of the GPL is a problem. It requires that the GPL be applied
to the whole of any work that contains any part of any GPL software.
Clause 2 does exempt "mere aggregation of another unrelated program", but
does define "mere aggregation" and "unrelated". Since TOPS-20 Maxima
necessarily runs under TOPS-20, there clearly is a relationship; and thus
this exemption does not apply.
The GPL has no clause permitting waiver of any of its terms by permission
of the copyright holder. In other words, even if the copyright holder of
Maxima says that he doesn't mind if a TOPS-20 Maxima is distributed as
part of a TOPS-20 distribution, the GPL still prohibits doing so.
Some people claim that the GPL was poorly-written, and that these viral
characteristics are unintentional. I disagree. I believe that the GPL
was very carefully written to be viral.
Paul Rubin
> Clause 2(b) of the GPL is a problem. It requires that the GPL be applied
> to the whole of any work that contains any part of any GPL software.
>
> Clause 2 does exempt "mere aggregation of another unrelated program", but
> does define "mere aggregation" and "unrelated". Since TOPS-20 Maxima
> necessarily runs under TOPS-20, there clearly is a relationship; and thus
> this exemption does not apply.
By that logic, the FreeBSD port of GNU Emacs can't be included with
FreeBSD distributions, but in fact that is done all the time. I think
the GPL would be more of an issue if TOPS-20 depended on Maxima than
the other way.
> The GPL has no clause permitting waiver of any of its terms by
> permission of the copyright holder. In other words, even if the
> copyright holder of Maxima says that he doesn't mind if a TOPS-20
> Maxima is distributed as part of a TOPS-20 distribution, the GPL
> still prohibits doing so.
Actually, such a clause is present:
10. If you wish to incorporate parts of the Program into other
free programs whose distribution conditions are different, write
to the author to ask for permission.
But even if it wasn't present, of course the copyright holder can
still waive the license terms. Waiving just means deciding not to
enforce. Since nobody but the copyright holder has authority to
enforce the license, if the copyright holder decides not to enforce,
it's waived.
However, I think it's more satisfying on technical grounds to try to
dig up a Maclisp Macsyma from the pre-Symbolics days, than to backport
Maxima to TOPS-20 Common Lisp. YMMV.
Christopher C. Stacy
Mark> On Sat, 12 Apr 2003, John Saeger wrote:
>> Does anybody know if runnable PDP-10 Macsyma is available anywhere? And
>> maybe the source code?
Mark> The only known runnable PDP-10 Macsyma is the TOPS-20 port of Macsyma 304.
Mark> Unfortunately, it has a 1983 Symbolics copyright and trade secret notice
Mark> for "enhancements", meaning that it can't be distributed without the
Mark> permission of the current owner of the assets of Macsyma, Inc.
Mark> The ITS archive had a copy of MIT (not Symbolics) Macsyma
Mark> source, but I don't know if that is available or what needs to
Mark> be done to whip it into shape.
If anyone has a copy of those archives, I would suggest being
very careful about the legal implications of releasing them.
MIT sold Symbolics the exclusive rights to that very code.
Afterwards, the U.S. Department of Energy demanded rights to a
derivative work of that very same code. Various legal battles
ensued to resolve that predicament, and the Macsyma that was
eventually made available under license from DOE was _not_ the
MACLISP version. It was a subset that had been ported to NIL,
a long dead dialect of Lisp for the VAX. The "Maxima" program
is yet a third derivative of that, whose rights to GPL are from
an agreement with DOE, deriving from DOE's license with MIT.
The version of Macsyma from Symbolics was eventually rewritten
by them into Common Lisp. This is not, however, the basis of
the "Maxima" program, which was an independant effort operating
on a different fork of the code.
Also, the ITS version doubtless includes various works that
are owned by private individual contributors (rather than MIT
or Symbolics, and certainly not DOE). One should be careful
to identify the proper owner of each file (which could possibly
conflict with what the copyright notices say). Just because MIT
or somebody perhaps put an incorrect copyright notice on something
doesn't relieve any later third parties who want to copy those files.
Also, be aware that the name "Macsyma" is trademarked by Symbolics
(whose assets have been bought by various successor companies,
whose company names include the "Macsyma" or "Symbolics").
I did a little work on it about 9 years ago during the period
it was owned by "Macsyma, Inc.".
To use the MACLISP version of Macsyma, one should seek to secure
a legal agreement license from the current incarnation of Symbolics
(which probably still owns the exclusive license to that code),
and possibly also a second legal agreement from MIT, which has
entered into licensing contracts with other third parties (ie. DOE).
All that being said, I think it would be really cool if the original
MACSYMA that ran in MACLISP on ITS and TOPS-20 was available for the
PDP-10 emulators. Perhaps someone will sort out all the legalities.
However, the GPL'd version of MACSYMA, aka MAXIMA, is better than
the original program on the PDP-10, if only for its modern graphics.
MAXIMA is of course written in Common Lisp, not MACLISP.
If you just want a really good version of Macsyma, not having
to do with PDP-10s, you can currently purchase a copy of "Macsyma"
from Symbolics. It started as the original MACLISP version, then
was ported to Zetalisp (the Lisp Machine), then to ANSI Common Lisp.
Macsyma runs on Windows, and has lots more features than "Maxima",
and represents many years of additional commercial development.
I think the most recent incarnation of Symbolics is called
"New Symbolics Technologies" or something. They are based
in Northern Virginia. I could dig up the contact information.
ObHistory: The final product from Symbolics was a Lisp Machine
called "Open Genera" that ran in software: a complete emulation
of the Ivory processor chip that powered the workstations.
The emulator only runs on DEC Alphas (that being the fastest
thing around at the time, and the only 64-bit platform, anyway.)
This product is also still available, but you don't need it to
run Macsyma. Their PC version of Macsyma is much better,
having had extensive development post-LispM era.
Paul Rubin
> MIT sold Symbolics the exclusive rights to that very code.
> Afterwards, the U.S. Department of Energy demanded rights to a
> derivative work of that very same code. Various legal battles
> ensued to resolve that predicament, and the Macsyma that was
> eventually made available under license from DOE was _not_ the
> MACLISP version. It was a subset that had been ported to NIL,
> a long dead dialect of Lisp for the VAX. The "Maxima" program
> is yet a third derivative of that, whose rights to GPL are from
> an agreement with DOE, deriving from DOE's license with MIT.
>
> The version of Macsyma from Symbolics was eventually rewritten
> by them into Common Lisp. This is not, however, the basis of
> the "Maxima" program, which was an independant effort operating
> on a different fork of the code.
Thanks for that history. Where does Vaxima fit into it? That's the
version that ran under Franz Lisp on the Vax and later was ported to
Common Lisp, if I have it right. I had thought that it was directly
adapted from a Maclisp Macsyma that had been obtained from the DoE and
that the GPL'd Maxima (done at UT Austin unless I'm confusing
different versions with each other) descended from it, but I'm no
expert.
Mark Crispin
> If anyone has a copy of those archives, I would suggest being
> very careful about the legal implications of releasing them.
Precisely. The fact that "so-and-so does it and hasn't been prosecuted"
does not translate into "you can do it and not be prosecuted."
As far as I can determine, all versions of Macsyma are encumbered in one
way or another. Maxima is probably the best of a bad lot; however the GPL
virus is a problem that will probably preclude it from ever being part of
a pre-built TOPS-20 system.
Mark Crispin
> By that logic, the FreeBSD port of GNU Emacs can't be included with
> FreeBSD distributions, but in fact that is done all the time.
Perhaps the copyright owner of GNU emacs simply hasn't sued FreeBSD yet.
Perhaps that will yet happen.
> Since nobody but the copyright holder has authority to
> enforce the license, if the copyright holder decides not to enforce,
> it's waived.
That's not so clear with the GPL. There is reason to believe that third
parties can sue to enforce its terms even if the copyright holder chooses
not to. Nothing in the GPL says that failure to enforce constitutes
waiver. Do you want to be the defendant of a lawsuit for an indiscretion
with GPL software that you committed 20 years earlier?
The GPL is written in a very particular way. I for one do not desire to
test those particular aspects in court.
The only safe assumption is that if you distribute something that is not
GPL (or can not be placed under GPL), you had better not have anything in
that distribution which is GPL.
You also have to be careful if you are the author of non-GPL open source
software. If your non-GPL license does not resist the GPL virus, the GPL
can be used to seize control of your software.
In the case of Maxima, probably the best thing would be for someone to
port it to run under TOPS-20 Common Lisp, and then make a DUMPER tape
image that can be loaded onto other TOPS-20 systems. Since the DUMPER
tape image only contains Maxima, that will restrain the GPL virus.
If the successor to Symbolics is willing to license redistribution of the
old TOPS-20 Symbolics Macsyma, that would be the easiest as far as
licensing. MIT would not be in the picture, because as a licensee of the
successor of Symbolics one would be an assign of MIT's license to
Symbolics.
Paul Rubin
> > Since nobody but the copyright holder has authority to
> > enforce the license, if the copyright holder decides not to enforce,
> > it's waived.
>
> That's not so clear with the GPL. There is reason to believe that third
> parties can sue to enforce its terms even if the copyright holder chooses
> not to.
What reason? Nobody else has standing.
> Nothing in the GPL says that failure to enforce constitutes waiver.
That's correct--waiver would be an explicit promise not to enforce.
The GPL certainly doesn't prevent a copyright holder from making such
a promise. One has to be careful, though, if the code contains
contributions from third parties, since each contributor would have to
make the same promise or otherwise assign their rights.
> In the case of Maxima, probably the best thing would be for someone to
> port it to run under TOPS-20 Common Lisp, and then make a DUMPER tape
> image that can be loaded onto other TOPS-20 systems. Since the DUMPER
> tape image only contains Maxima, that will restrain the GPL virus.
That sounds like a safe and clean approach and should be perfectly fine.
> If the successor to Symbolics is willing to license redistribution of the
> old TOPS-20 Symbolics Macsyma, that would be the easiest as far as
> licensing. MIT would not be in the picture, because as a licensee of the
> successor of Symbolics one would be an assign of MIT's license to
> Symbolics.
That sounds good too, though maybe harder to accomplish, unless the
Symbolics folks have softened up a lot since the Vaxima days.
Peter da Silva
Mark Crispin <M...@CAC.Washington.EDU> wrote:
>Since Maxima is under GPL, that may mean that a TOPS-20 Maxima would have
>to be released individually, rather than bundled with a TOPS-20
>distribution, because of the viral aspects of GPL.
Much as i dislike the GPL myself, if *Microsoft* is willing to bundle
GPLed software (including GCC!) with Windows Services for UNIX I think
you'd be safe here.
--
Rev. Peter da Silva, ULC. 29.6852N 95.5770W WWFD?
"Be conservative in what you generate, and liberal in what you accept"
-- Matthew 10:16 (l.trans)
Peter da Silva
Mark Crispin <m...@CAC.Washington.EDU> wrote:
>Clause 2 does exempt "mere aggregation of another unrelated program", but
>does define "mere aggregation" and "unrelated". Since TOPS-20 Maxima
>necessarily runs under TOPS-20, there clearly is a relationship; and thus
>this exemption does not apply.
The FSF says that linking with libraries that are part of an
operating system is normal use of the operating system. Hence use
of this libraries is not considered derivative work.
>The GPL has no clause permitting waiver of any of its terms by permission
>of the copyright holder.
10. If you wish to incorporate parts of the Program into other
free programs whose distribution conditions are different, write
to the author to ask for permission.
In addition, the copyright holder can release the work under a modified
version of the license with any term they want. There are many programs
released under modified forms of the GPL, or under multiple licenses.
>Some people claim that the GPL was poorly-written, and that these viral
>characteristics are unintentional. I disagree. I believe that the GPL
>was very carefully written to be viral.
I agree that it was deliberately written to be viral. I don't believe that
it's so viral as to engulf the OS it's running on.
Peter da Silva
Mark Crispin <M...@CAC.Washington.EDU> wrote:
>software. If your non-GPL license does not resist the GPL virus, the GPL
>can be used to seize control of your software.
It can be used to force redistribution of the version that was distributed
under GPL. It can't do anything to a version you've distributed under other
terms.
Mark Crispin
> > That's not so clear with the GPL. There is reason to believe that third
> > parties can sue to enforce its terms even if the copyright holder chooses
> > not to.
> What reason? Nobody else has standing.
Don't be so sure about that. The way that the GPL is worded can be
construed to give standing to any licensee who alleges that his rights
under the GPL were infringed.
It looks like it doesn't even have to be someone who received a copy of
your distribution. Note the use of phrases such as "all third parties".
For that matter, I don't think that people residing in the US can legally
distribute GPL version 2 software at all.
It's your choice if you wish to distribute GPL software in the hope that
nobody decides to exploit that fact to sue you. That does not invalidate
the choice of others who have read the GPL carefully, shuddered at the
implications, and decided that they will not distribute any GPL software.
You can claim that these aren't real problems until you're blue in the
face. I don't think that you have sufficient assets to indemnify me in
the event that you are wrong. I don't think that you can find an insurer
who'd underwrite such indemnification either.
Mark Crispin
> Much as i dislike the GPL myself, if *Microsoft* is willing to bundle
> GPLed software (including GCC!) with Windows Services for UNIX I think
> you'd be safe here.
Microsoft can fight a protracted war in court until the other side is
bankrupted by the legal fees and forced to settle on Microsoft's terms.
Rules that apply to you and me do not apply to mega-corporations.
John R. Strohm
news:Pine.WNT.4.55.03...@Shimo-Tomobiki.Panda.COM...
> On Mon, 14 Apr 2003, Christopher C. Stacy wrote:
> > If anyone has a copy of those archives, I would suggest being
> > very careful about the legal implications of releasing them.
>
> Precisely. The fact that "so-and-so does it and hasn't been prosecuted"
> does not translate into "you can do it and not be prosecuted."
>
> As far as I can determine, all versions of Macsyma are encumbered in one
> way or another. Maxima is probably the best of a bad lot; however the GPL
> virus is a problem that will probably preclude it from ever being part of
> a pre-built TOPS-20 system.
I have queried the Free Software Foundation about this. It is their
license, after all.
John R. Strohm
news:Pine.LNX.4.55.03...@shiva1.cac.washington.edu...
> For that matter, I don't think that people residing in the US can legally
> distribute GPL version 2 software at all.
The Version 2 GPL specifically grants unlimited license to distribute,
provided that certain requirements are satisfied.
Which specific requirement(s) of the GPL do you believe that people residing
in the US cannot satisfy?
In detail, please.
John R. Strohm
news:Pine.LNX.4.55.03...@shiva1.cac.washington.edu...
According to the University of Washington web page, UW has a law school.
Four of the law school professors list specializations in copyright law; two
of those four list specializations in intellectual property, copyright,
patent, and trademark law. Of the two, Professor Townsend
http://www.law.washington.edu/LawSchool/faculty/Profiles/Townsend.html would
seem to be uniquely qualified to take a serious look at the GPL and address
your concerns.
Might I respectfully suggest you give him a call?
In the meantime, you might ponder the following quote from the FSF webpage,
found at http://www.fsf.org/licenses/gpl-violation.html: "Note that the GPL,
and other copyleft licenses, are copyright licenses. This means that only
the copyright holders are empowered to act against violations. The FSF acts
on all GPL violations reported on FSF copyrighted code, and we offer
assistance to any other copyright holder who wishes to do the same."
Mark Crispin
> According to the University of Washington web page, UW has a law school.
> Four of the law school professors list specializations in copyright law; two
> of those four list specializations in intellectual property, copyright,
> patent, and trademark law. Of the two, Professor Townsend
>
Do you really expect to be taken seriously when you make stupid, illogical
comments like that?
Do you think that professors are in the business of offering free legal
advice? Do you have the first clue as to what "legal advice", much less
an "attorney-client relationship" constitutes?
Didn't think so.
> In the meantime, you might ponder the following quote from the FSF webpage
A quote on an FSF web page does not constitute any sort of
legally-enforcable statement.
Mark Crispin
> The Version 2 GPL specifically grants unlimited license to distribute,
> provided that certain requirements are satisfied.
>
> Which specific requirement(s) of the GPL do you believe that people residing
> in the US cannot satisfy?
Consider the meaning of "any third party" and the implication of that
vis-a-vis US export law.
Now read GPLv2 term 7.
Mark Crispin
> The FSF says that linking with libraries that are part of an
> operating system is normal use of the operating system. Hence use
> of this libraries is not considered derivative work.
Apples and oranges.
I am referring to the inclusion of a <MAXIMA> directory and a
<SUBSYS>MAXIMA.EXE on a prebuilt TOPS-20 filesystem. That is clearly not
"linking with libraries".
> >The GPL has no clause permitting waiver of any of its terms by permission
> >of the copyright holder.
> 10. If you wish to incorporate parts of the Program into other
> free programs whose distribution conditions are different, write
> to the author to ask for permission.
That's not waiver of terms.
Furthermore, this gets you into the business of the definition of "other
free programs." FSF has position papers on what they define as "free"; I
suggest that you read them.
> In addition, the copyright holder can release the work under a modified
> version of the license with any term they want. There are many programs
> released under modified forms of the GPL, or under multiple licenses.
Not legally. The GPL itself is copyrighted. From the GPL:
Everyone is permitted to copy and distribute verbatim copies
of this license document, but changing it is not allowed.
> I agree that it was deliberately written to be viral. I don't believe that
> it's so viral as to engulf the OS it's running on.
It will, however, engulf anything that bundles it. Maxima can not be
bundled as part of a TOPS-20 distribution because of GPL. It can only be
distributed separately.
I want something that I can bundle.
Mark Crispin
> >You also have to be careful if you are the author of non-GPL open source
> >software. If your non-GPL license does not resist the GPL virus, the GPL
> >can be used to seize control of your software.
> It can be used to force redistribution of the version that was distributed
> under GPL. It can't do anything to a version you've distributed under other
> terms.
Consider a necessary and obvious modification (extension, bugfix, or what
have you) that is made by someone else and placed under GPL. Unless you
have superior rights, you can not do the same thing (or something
substantially similar) in the non-GPL version.
It's just like software patents.
Christopher C. Stacy
Paul> cst...@dtpq.com (Christopher C. Stacy) writes:
>> MIT sold Symbolics the exclusive rights to that very code.
>> Afterwards, the U.S. Department of Energy demanded rights to a
>> derivative work of that very same code. Various legal battles
>> ensued to resolve that predicament, and the Macsyma that was
>> eventually made available under license from DOE was _not_ the
>> MACLISP version. It was a subset that had been ported to NIL,
>> a long dead dialect of Lisp for the VAX. The "Maxima" program
>> is yet a third derivative of that, whose rights to GPL are from
>> an agreement with DOE, deriving from DOE's license with MIT.
>>
>> The version of Macsyma from Symbolics was eventually rewritten
>> by them into Common Lisp. This is not, however, the basis of
>> the "Maxima" program, which was an independant effort operating
>> on a different fork of the code.
Paul> Thanks for that history. Where does Vaxima fit into it? That's the
Paul> version that ran under Franz Lisp on the Vax and later was ported to
Paul> Common Lisp, if I have it right. I had thought that it was directly
Paul> adapted from a Maclisp Macsyma that had been obtained from the DoE and
Paul> that the GPL'd Maxima (done at UT Austin unless I'm confusing
Paul> different versions with each other) descended from it, but I'm no expert.
I think you're right, except that Vaxima was derived from the NIL
version, not the MACLISP version, so it's one more removed from that.
The DOE release was sanitized in some fashion, but I wasn't looking
at the sources when all that stuff happened, so I don't know the
precise relationships of the file contents in the different forks
around the times of: the original LispM port, NIL, Symbolics licensing
and after that DOE and Franz Vaxima. (I only saw the original ITS sources,
and then many years later, the Symbolics-cum-Macsyma-Inc sources.)
Peter da Silva
Mark Crispin <m...@CAC.Washington.EDU> wrote:
>Consider a necessary and obvious modification (extension, bugfix, or what
>have you) that is made by someone else and placed under GPL. Unless you
>have superior rights, you can not do the same thing (or something
>substantially similar) in the non-GPL version.
I suspect you are confusing copyrights with patents and trademarks.
Peter da Silva
Mark Crispin <m...@CAC.Washington.EDU> wrote:
>I am referring to the inclusion of a <MAXIMA> directory and a
><SUBSYS>MAXIMA.EXE on a prebuilt TOPS-20 filesystem. That is clearly not
>"linking with libraries".
No, that's aggregation.
You can clearly include an empty <MAXIMA> directory and a <SUBSYS>MAXIMA.EXE
that built and installed the package from source, and then replaced itself
with the actual program, without violating copyright.
>Not legally. The GPL itself is copyrighted. From the GPL:
>
> Everyone is permitted to copy and distribute verbatim copies
> of this license document, but changing it is not allowed.
OK, then they can release it under the GPL and the BSD license. Sheesh.
Zane H. Healy
> I thought TWENEX was a nickname for TOPS-20 that
> acknowledged its TENEX heritage...
Let's chock that up to my lack of sleep of late. I read (and typed) TWENEX,
but was thinking of TENEX.
Zane
Paul Rubin
> > > That's not so clear with the GPL. There is reason to believe that third
> > > parties can sue to enforce its terms even if the copyright holder chooses
> > > not to.
> > What reason? Nobody else has standing.
>
> Don't be so sure about that. The way that the GPL is worded can be
> construed to give standing to any licensee who alleges that his rights
> under the GPL were infringed.
I don't think it can do that, at least enforceably. Remember that the
enforcement mechanism of the GPL is copyright. Only the copyright holder
has standing to bring a copyright infringement suit.
> It's your choice if you wish to distribute GPL software in the hope that
> nobody decides to exploit that fact to sue you. That does not invalidate
> the choice of others who have read the GPL carefully, shuddered at the
> implications, and decided that they will not distribute any GPL software.
I do distribute GPL stuff, but so far only either by itself, or in
combination with other GPL stuff or GPL-compatible stuff. While I do
have some worries, they're generally not GPL-related.
> You can claim that these aren't real problems until you're blue in the
> face. I don't think that you have sufficient assets to indemnify me in
> the event that you are wrong. I don't think that you can find an insurer
> who'd underwrite such indemnification either.
Just the same, there are any number of companies shipping products
with embedded Linux kernels. There's also the Caldera and Lindows OS
distributions which use Linux kernels and GNU utilities but also
bundle with proprietary applications. I think there are even some Red
Hat distributions like that. I think those companies have all managed
to find insurers.
John R. Strohm
> Furthermore, this gets you into the business of the definition of "other
> free programs." FSF has position papers on what they define as "free"; I
> suggest that you read them.
I have. Quite a while back. I consider Stallman's position reasonable.
If what you want to do boils down to incorporating free programs into
non-free programs, with the result being anything other than free, I expect
that you will get very little sympathy, or help, from the Free Software
Foundation.
Jim Gosling and Unipress taught Richard Stallman a painful lesson on that
score. The GPL is a direct result of that lesson.
So.
What precisely is it that you want to do?
Mark Crispin
> What precisely is it that you want to do?
You haven't been paying attention. I would like to bundle some version of
TOPS-20 Macsyma in a pre-built TOPS-20 filesystem.
A pre-built TOPS-20 filesystem has similar characteristics to "a volume of
a storage or distribution medium" but is not the same thing. Thus, it is
not exempt from the GPL virus under the aggregation clause.
Perhaps if virtual filesystems had been considered when the GPL was
written, they would have been exempted too. But they weren't and they
aren't.
Glen Herrmannsfeldt
> On Thu, 17 Apr 2003, John R. Strohm wrote:
> > What precisely is it that you want to do?
>
> You haven't been paying attention. I would like to bundle some version of
> TOPS-20 Macsyma in a pre-built TOPS-20 filesystem.
>
> A pre-built TOPS-20 filesystem has similar characteristics to "a volume of
> a storage or distribution medium" but is not the same thing. Thus, it is
> not exempt from the GPL virus under the aggregation clause.
>
> Perhaps if virtual filesystems had been considered when the GPL was
> written, they would have been exempted too. But they weren't and they
> aren't.
OK, say I agree that virtual file systems should have been exempted but
weren't.
Is it true that real file systems are exempt? Is putting two files on the
same CD considered "mere aggregation"?
Now, I might consider ISO9660 to be a virtual file system. It is not (as
far as I know) the native file system of any OS, but if a virtual file
system readable by many OS.
If a TOPS-20 file system is written onto a CD, not as a file inside an
ISO9660 file system, but directly onto a CD, is that, in terms of mere
aggregation, any different than in ISO9660?
--------------------------------------------------------
OK, now consider something completely different. In a virtual file system,
(that I know) the entire file system is contained in one file on a real file
system. How about, though, a virtual virtual file system where each file of
the virtual file system is also a file in the real file system?
Consider what NFS does. It takes the files in one file system and, using
special software, makes them appear as files in another OS, even having the
properties that are expected on that OS. If one had NFS for TOPS20 one
could, for example, put a CD on a unix system, NFS mount that CD to the
TOPS-20 system, so that what was mere aggregation on the CD now looks like a
complete file system to TOPS-20.
Now, leave out the NFS step above, but just write device driver for either
the guest OS (so that TOPS-20 could read ISO9660, for example), or as part
of the emulator (so it would make an ISO9660 file system look like a TOPS-20
file system to the guest).
(As an example, this is what OS/2 did in its DOS emulation. There is a DOS
device driver that allows emulated DOS to read files from an OS/2 file
system, by passing the commands to OS/2 to process, and passing the results
back.)
---------------------------------------------------------------
Now, considering that I am not a lawyer, and haven't read the GPL for some
years, and don't know the legal definition of "mere aggregation," I might
say that if "mere aggregation" is allowed for files on a CD then it should
be allowed for files on a TOPS-20 file system.
It seems to me that the only difference is that ISO9660 is more popular than
TOPS-20.
-- glen
Mark Crispin
>[snip]
Indeed. You can argue it either way. That's why it's a problem.
Glen Herrmannsfeldt
> On Thu, 17 Apr 2003, Glen Herrmannsfeldt wrote:
> >[snip]
>
> Indeed. You can argue it either way. That's why it's a problem.
A Google search with the quoted string "mere aggregation" comes up with
16000 definitions.
It does seem that they specifically mention "media or file system" as means
of "mere aggregation."
Then they confuse it by saying that if they run in the same address space,
that they are definitely not "mere aggregation". Well, I know plenty of OS
that run multiple separate programs in the same address space. On machines
running those OS, everyone would have to be very careful what programs
other users were running, in order not to violate the GPL.
So I come up with my own definition. If the bit string of the supposed
"mere aggregation" contains the individual programs as substrings, then it
is "mere aggregation". A simple substring operation could separate the
two. This eliminates all compressed archive formats, for example, but not
uncompressed tar files. The FSF people make a strong claim that linking,
as for example the unix ld command, is definitely not mere aggregation. Yet
I know of some systems where the result of a link command is just the
concatenation of the individual files. (That is, the unix cat command would
do it.) There is another problem, in that one could say that the bits
actually written on a CD are encoded with various error correcting codes,
and so cannot be separated anymore. However, I believe it is well
understood that a CD is a common distribution method for "mere
aggregations".
So, my result is that an ucompressed virtual file system is legal, but a
compressed one is not!
-- glen
vi...@parcelfarce.linux.theplanet.co.uk
Glen Herrmannsfeldt <g...@ugcs.caltech.edu> wrote:
>So, my result is that an ucompressed virtual file system is legal, but a
>compressed one is not!
That depends. If you can't take an fs image and replace the files in
question, I would expect major objections even if files were substrings
of the image.
If, OTOH, you provide tools that allow
a) turn a collection of named files into valid fs image
b) turn fs image into collection of named files
and these tools are not patent-encumbered, etc., the common interpretation
appears to be that "mere aggregation" still applies.
If somebody tried to sue in situation above, I would expect a *lot*
of seriously upset companies with money to burn and strong reasons to
prevent establishing such precedents.
John R. Strohm
news:_7Rna.507669$L1.145710@sccrnsc02...
I see considerable confusion cropping up here. The confusion appears to
stem from the belief that the GPL is written in English. This belief is
erroneous: the GPL is written in legalese, a human programming language that
borrows some words, syntax, and base semantics from English, but changes the
definition of those words in ways not at all obvious to persons not in the
legal profession (which includes me). In particular, "mere aggregation" is
a legal term, with a specific legal meaning, that may or may not mean what
you think. The following paragraphs, extracted from a 53-year-old U.S.
Supreme Court ruling, may help to illustrate the problem.
"While this Court has sustained combination patents, 4 it never has ventured
to give a precise and comprehensive definition of the test to be applied in
such cases. The voluminous literature which the subject has excited
discloses no such test. 5 It is agreed that the key to patentability [340
U.S. 147, 151] of a mechanical device that brings old factors into
cooperation is presence or lack of invention. In course of time the
profession came to employ the term "combination" to imply its presence and
the term "aggregation" to signify its absence, thus making antonyms in legal
art of words which in ordinary speech are more nearly synonyms. However
useful as words of art to denote in short form that an assembly of units has
failed or has met the examination for invention, their employment as tests
to determine invention results in nothing but confusion. The concept of
invention is inherently elusive when applied to combination of old elements.
This, together with the imprecision of our language, have counselled courts
and text writers to be cautious in affirmative definitions or rules on the
subject. 6
The negative rule accrued from many litigations was condensed about as
precisely as the subject permits in Lincoln Engineering Co. v.
Stewart-Warner Corp., 303 U.S. 545, 549 : "The mere aggregation of a number
of old parts or elements which, in the aggregation, perform or produce no
new or different function or operation than that theretofore performed or
produced by them, is not patentable invention." To the same end is Toledo
[340 U.S. 147, 152] Pressed Steel Co. v. Standard Parts, Inc., 307 U.S.
350 , and Cuno Engineering Corp. v. Automatic Devices Corp., 314 U.S. 84 .
The conjunction or concert of known elements must contribute something; only
when the whole in some way exceeds the sum of its parts is the accumulation
of old devices patentable. Elements may, of course, especially in chemistry
or electronics, take on some new quality or function from being brought into
concert, but this is not a usual result of uniting elements old in
mechanics. This case is wanting in any unusual or surprising consequences
from the unification of the elements here concerned, and there is nothing to
indicate that the lower courts scrutinized the claims in the light of this
rather severe test."
--from "A. & P. TEA CO. v. SUPERMARKET CORP.,
340 U.S. 147 (1950)"
In other words, your personal definition of "mere aggregation" is not
useful.
These two paragraphs from the Court suggest a simple test. If the non-GPL
product would work perfectly well without the inclusion of the GPL work,
*AND* without the inclusion of some equivalent work, then the union of their
two parts is almost certainly a "mere aggregation", and, as such, would not
invoke what is commonly called the "viral" property of the GPL. TOPS-20
works perfectly well without Maxima, even if a <MAXIMA> directory has been
set aside, and a (placeholder) <SUBSYS>MAXIMA.EXE included. Furthermore, if
such a TOPS-20 were distributed, the act of removing Maxima would not affect
TOPS-20 in the slightest. On the other hand, if a product were produced
that REQUIRED Maxima, or some similar symbolic mathematics program, in order
to function (say, something like MathCad, that both formats mathematical
expressions and equations for typesetting *and* allows the user to do
mathematical operations on the formatted expressions and equations), then a
distribution of that product with Maxima would probably NOT be a "mere
aggregation". Under this test, which admittedly has not received any kind
of legal scrutiny, it is immediately obvious that Mark Crispin's prebuilt
TOPS-20, with Maxima included, is a "mere aggregation".
Glen Herrmannsfeldt
"John R. Strohm" <str...@airmail.net> wrote in message
news:5511515E6880A8C9.7C069F39...@lp.airnews.net...
(snip)
(snip of examples from patent law)
> These two paragraphs from the Court suggest a simple test. If the non-GPL
> product would work perfectly well without the inclusion of the GPL work,
> *AND* without the inclusion of some equivalent work, then the union of
their
> two parts is almost certainly a "mere aggregation", and, as such, would
not
> invoke what is commonly called the "viral" property of the GPL. TOPS-20
> works perfectly well without Maxima, even if a <MAXIMA> directory has been
> set aside, and a (placeholder) <SUBSYS>MAXIMA.EXE included. Furthermore,
if
> such a TOPS-20 were distributed, the act of removing Maxima would not
affect
> TOPS-20 in the slightest. On the other hand, if a product were produced
> that REQUIRED Maxima, or some similar symbolic mathematics program, in
order
> to function (say, something like MathCad, that both formats mathematical
> expressions and equations for typesetting *and* allows the user to do
> mathematical operations on the formatted expressions and equations), then
a
> distribution of that product with Maxima would probably NOT be a "mere
> aggregation". Under this test, which admittedly has not received any kind
> of legal scrutiny, it is immediately obvious that Mark Crispin's prebuilt
> TOPS-20, with Maxima included, is a "mere aggregation".
As far as I know there are differences between patent and copyright law that
may apply here, but...
I think the biggest difference from our definitions is physical vs. logical
coupling.
My definition was meant to express a physical connection. They are, or are
not easily separated. It looks to me like yours is a logical connection. A
requires, or does not require, B to work. So I make two examples:
1) Program A requires members of GPL'ed library B to work. I supply two
separate CD's, which must be ordered separately, will be mailed in separate
envelopes. The receiver of the two CD's runs an install program which
combines them so that they run together. A requires B to work, always.
2) A compression program has two different compression algorithms included.
It is statically linked as one executable file. A run-time option selects
which algorithm is used, only one can be selected at a time. There is no
way to separate the two linked subroutines. They could have been two
separate programs, though a little less convenient to the user.
In the first case they are logically one, but physically two separate
programs. In the second case it is physically one program, but logically
two.
-- glen
Glen Herrmannsfeldt
<vi...@parcelfarce.linux.theplanet.co.uk> wrote in message
news:b7orjh$bev$1...@parcelfarce.linux.theplanet.co.uk...
> In article <_7Rna.507669$L1.145710@sccrnsc02>,
> Glen Herrmannsfeldt <g...@ugcs.caltech.edu> wrote:
>
> >So, my result is that an ucompressed virtual file system is legal, but a
> >compressed one is not!
>
> That depends. If you can't take an fs image and replace the files in
> question, I would expect major objections even if files were substrings
> of the image.
There may be checksums on the data that would have to be changed if you
replaced one with the other. In a compressed file you generally can't
change part in the middle and expect to get anything out of later parts.
> If, OTOH, you provide tools that allow
> a) turn a collection of named files into valid fs image
> b) turn fs image into collection of named files
> and these tools are not patent-encumbered, etc., the common interpretation
> appears to be that "mere aggregation" still applies.
>
> If somebody tried to sue in situation above, I would expect a *lot*
> of seriously upset companies with money to burn and strong reasons to
> prevent establishing such precedents.
So as long as the program to separate the two is supplied, along with the
aggregate, it would count as "mere aggregation" in this definition? Note,
for example, that some executable file formats can be broken up into
separate object programs. That is, the linking process is reversable. See
my next follow-up where I discuss physical and logical coupling.
-- glen
Douglas H. Quebbeman
> useful.
It doesn't have to be someone's personal definition... the court
has demonstrated in the past that even if a well-established and
understood definition for a word is extant, the court can and will
do as it pleases.
Late 1800s, in spite of scientific testimony to the contrary,
the U.S. Supreme Court declared that the tomato is *not* a fruit,
but is rather instead a vegetable.
Nuts deciding between fruits and vegetables...
Peter da Silva
>the U.S. Supreme Court declared that the tomato is *not* a fruit,
>but is rather instead a vegetable.
What's wrong with that? It *is* a vegetable. The only fruits I know of
that aren't vegetable are mushrooms.
John R. Strohm
news:3ea03d25$1...@news.iglou.com...
"Tomato", "fruit", and "vegetable" are not, generally speaking, technical
legal terms. "Aggregation" and "mere aggregation" are. This means that,
for legal purposes, which includes of course the construction of legal
documents, such as software licenses, the legal meaning trumps the
commonplace meaning, all day, every day.
Recognize that the GPL is NOT a piece of random prose, but a
very-carefully-constructed legal document. It is probably much more
carefully-constructed than much of the software distributed under its terms.
Christopher C. Stacy
John> "Douglas H. Quebbeman" <do...@iglou.com> wrote in message
John> news:3ea03d25$1...@news.iglou.com...
>> "John R. Strohm" <str...@airmail.net> wrote in message
>> >
>> > In other words, your personal definition of "mere aggregation" is not
>> > useful.
>>
>> It doesn't have to be someone's personal definition... the court
>> has demonstrated in the past that even if a well-established and
>> understood definition for a word is extant, the court can and will
>> do as it pleases.
>>
>> Late 1800s, in spite of scientific testimony to the contrary,
>> the U.S. Supreme Court declared that the tomato is *not* a fruit,
>> but is rather instead a vegetable.
John> "Tomato", "fruit", and "vegetable" are not, generally speaking, technical
John> legal terms. "Aggregation" and "mere aggregation" are. This means that,
John> for legal purposes, which includes of course the construction of legal
John> documents, such as software licenses, the legal meaning trumps the
John> commonplace meaning, all day, every day.
John> Recognize that the GPL is NOT a piece of random prose, but a
John> very-carefully-constructed legal document. It is probably much more
John> carefully-constructed than much of the software distributed under its terms.
Black's defines "aggregate" to just mean what it means in English, and
"mere aggregation" doesn't appear as a term of art in that dictionary.
Googling around suggests that the term "mere aggregation" has only
been used as a licensing term by the {L}GPL, and therefore is the
subject of much controversy, discussion, and worry.
The phrase "mere aggregation" has been used descriptively in a USC
rulings, but not as term of art, and not as part of contractual
language. For example, they have used the phrase in a completely
different way, having to do with evaluating the quality of
patentability of a functional modification of a device.
Another example was a case involving a symbolic association
with a group of expressive elements. It has also, for example,
been used to describe a collection of people, in deciding what
constitutes a religion. It has also been used in various other
contexts, but in all these cases it was used to mean just what
it means in English; not a term of art.
Also consider that the GNU licenses have never been tested in court,
and their exact meanings are commonly hotly debated. Which leads one
to wonder if the GPL was as carefully constructed as is necessary.
It seems that it would be wise to shy away from it, if you are concerned
that the terms are not clearly understood, which they are not (and I believe
that opinion is also held by at least some IP lawyers who have looked at it.)
Just because there is some opinion as to what it means, presented by advocates
of the license, doesn't mean that a court would find that same meaning.
Glen Herrmannsfeldt
"Christopher C. Stacy" <cst...@dtpq.com> wrote in message
news:u7k9rv...@dtpq.com...
(snip relating to the definition of "mere aggregation")
I do believe that patent law is different enough for any definition there
not to apply.
> Also consider that the GNU licenses have never been tested in court,
> and their exact meanings are commonly hotly debated. Which leads one
> to wonder if the GPL was as carefully constructed as is necessary.
> It seems that it would be wise to shy away from it, if you are concerned
> that the terms are not clearly understood, which they are not (and I
believe
> that opinion is also held by at least some IP lawyers who have looked at
it.)
> Just because there is some opinion as to what it means, presented by
advocates
> of the license, doesn't mean that a court would find that same meaning.
But consider, again, the case of a patent. I patent holder does not want to
be taken to court, as that makes it possible to lose the patent. One might
say the same for GPL. Not that I necessarily believe it, but it could be
that it was carefully crafted so that it wouldn't be tested in court.
Still no comments on my physical vs. logical aggregation question.
-- glen
John R. Strohm
[ snip ]
> Still no comments on my physical vs. logical aggregation question.
Section 2 of the GPL reads as follows:
2. You may modify your copy or copies of the Program or any portion
of it, thus forming a work based on the Program, and copy and
distribute such modifications or work under the terms of Section 1
above, provided that you also meet all of these conditions:
a) You must cause the modified files to carry prominent notices
stating that you changed the files and the date of any change.
b) You must cause any work that you distribute or publish, that in
whole or in part contains or is derived from the Program or any
part thereof, to be licensed as a whole at no charge to all third
parties under the terms of this License.
Your "physical aggregation" scheme (your item 2) appears to be a unified
work, and as such would fall under section 2 (b) of the GPL.
Your "logical aggregation" scheme (your item 1) appears to be an obvious
attempt to circumvent the clear intent of section 2 (b) of the GPL. Courts
tend to take dim views of that kind of shenanigan.
John R. Strohm
> and their exact meanings are commonly hotly debated. Which leads one
> to wonder if the GPL was as carefully constructed as is necessary.
> It seems that it would be wise to shy away from it, if you are concerned
> that the terms are not clearly understood, which they are not (and I
believe
> that opinion is also held by at least some IP lawyers who have looked at
it.)
> Just because there is some opinion as to what it means, presented by
advocates
> of the license, doesn't mean that a court would find that same meaning.
It is precisely the nature of a well-crafted legal instrument that it will
never (have to) be tested in court. Obviously-infringing conduct is
OBVIOUSLY infringing. Obviously-noninfringing conduct is, similarly,
OBVIOUSLY non-infringing. What conduct falls into the "grey areas" is also
pretty obvious. A prudent individual is not going to bet his business on
the legal system agreeing that he is in an area that is sufficiently
lightly-toned, rather than farther into the shadows, and a prudent corporate
counsel is going to explain that principle to that prudent individual, at
length.
Glen Herrmannsfeldt
> It is precisely the nature of a well-crafted legal instrument that it will
> never (have to) be tested in court. Obviously-infringing conduct is
> OBVIOUSLY infringing. Obviously-noninfringing conduct is, similarly,
> OBVIOUSLY non-infringing. What conduct falls into the "grey areas" is
also
> pretty obvious. A prudent individual is not going to bet his business on
> the legal system agreeing that he is in an area that is sufficiently
> lightly-toned, rather than farther into the shadows, and a prudent
corporate
> counsel is going to explain that principle to that prudent individual, at
> length.
While I pretty much agree with this statement, at least for copyright
questions, I am less sure for patents. Try this one:
There is an application that some people I know are interested in working
on. Others have previously worked in this area, and have patented an
algorithm for doing it. Not only that, but they have patented ALL
algorithms for doing it. That is they way they wrote the patent, and that
is the way the USPTO granted it. There was an example in Scientific
American where someone has patented teaching using workbooks. I don't
expect that all teachers will now stop using workbooks in their classes, so
they are pretty much betting their business on the legal system agreeing
with them.
I hope I am not getting too far off the subject, but you were using the
patent definition of "mere aggregation" and I am trying to find out if I
agree with that.
thanks,
-- glen
Glen Herrmannsfeldt
OK, now you see the problem. In the patent example, putting two items
together is what I would call physical aggregation. Say there exists a
screwdriver and wrench and I put them together in one bag. I now want to
patent the aggregate, toolkit, and sell it. The patent office should refuse
that as "mere aggregation". Yet, you say that in terms of GPL that is a
unified work. I carefully wrote my description so that only one could be
used at a time.
On the other hand, it seems that the "logical aggregation" is what FSF is
encouraging. That people not sell unified works (pre-installed software)
but require the users to do the unification themself.
If neither of those is possible then I have to agree with others, that there
is no way to use GPL software.
-- glen
John R. Strohm
> OK, now you see the problem. In the patent example, putting two items
> together is what I would call physical aggregation. Say there exists a
> screwdriver and wrench and I put them together in one bag. I now want to
> patent the aggregate, toolkit, and sell it. The patent office should
refuse
> that as "mere aggregation". Yet, you say that in terms of GPL that is a
> unified work. I carefully wrote my description so that only one could be
> used at a time.
What you described was not a separate screwdriver and a separate wrench in a
bag.
What you described was a screwdriver and a wrench that had been welded
together, presumably at a 90 degree angle.
> On the other hand, it seems that the "logical aggregation" is what FSF is
> encouraging. That people not sell unified works (pre-installed software)
> but require the users to do the unification themself.
Glen, what you are OBVIOUSLY missing is that the Free Software Foundation
has an agenda, and the GPL was designed to further that agenda.
I am going to presume a motivation here, to make the salient point.
You are trying to circumvent section 2 (b) so that you can SELL your Part
"A", and retain all the rights,and keep the code a commercial secret, so you
can make megabucks by screwing users. Your Part "A" requires a significant
GPL Part "B" in order to be of any utility. This is the evil vampire that
the Free Software Foundation is trying to kill, and the GPL is one of their
stakes. If you had any intellectual honesty in this contrived example, what
you would do is spend the time and effort to develop your own Part B as
well, and then you could release it ALL under your own "All Your Drive Are
Belong To Us" license. You don't want to do that. You want to take the GPL
code private.
This is, essentially, what Gosling did to Stallman with Gosling/Unipress
Emacs. It was the fallout from this that caused Stallman to design the
"viral" features into the GPL.
It may help you to reread the last chapter of "Hackers", by Stephen Levy,
and also some of the position papers on the FSF website.
> If neither of those is possible then I have to agree with others, that
there
> is no way to use GPL software.
Not at all. Get with the program. Either release your Part "A" under the
GPL, or develop your own Part "B", that doesn't infringe on the GPL'ed Part
B, and everyone is happy.
John R. Strohm
>
> "John R. Strohm" <str...@airmail.net> wrote in message
> news:61D778A92A9224D6.2A559C73...@lp.airnews.net...
>
> > It is precisely the nature of a well-crafted legal instrument that it
will
> > never (have to) be tested in court. Obviously-infringing conduct is
> > OBVIOUSLY infringing. Obviously-noninfringing conduct is, similarly,
> > OBVIOUSLY non-infringing. What conduct falls into the "grey areas" is
> also
> > pretty obvious. A prudent individual is not going to bet his business
on
> > the legal system agreeing that he is in an area that is sufficiently
> > lightly-toned, rather than farther into the shadows, and a prudent
> corporate
> > counsel is going to explain that principle to that prudent individual,
at
> > length.
>
> While I pretty much agree with this statement, at least for copyright
> questions, I am less sure for patents. Try this one:
Discussion of patent law is non sequitur in the context of the GPL. The GPL
derives SOLELY from copyright law. Copyright law and patent law are two
entirely different animals.
> There is an application that some people I know are interested in working
> on. Others have previously worked in this area, and have patented an
> algorithm for doing it. Not only that, but they have patented ALL
> algorithms for doing it. That is they way they wrote the patent, and
that
> is the way the USPTO granted it. There was an example in Scientific
> American where someone has patented teaching using workbooks. I don't
> expect that all teachers will now stop using workbooks in their classes,
so
> they are pretty much betting their business on the legal system agreeing
> with them.
A patent is just a license to sue. Patents can be challenged, and
frequently are. Sometimes they are upheld, and sometimes they are thrown
out. The A&P Tea Company case I quoted was a patent challenge, and the
Supreme Court overturned the patent. In the case of an actual patent of
"teaching with workbooks", that patent will last about as long as it takes
for the defense counsel to explain the patent to the judge and then show the
judge a class workbook that clearly predates the filing date on the patent.
In the case of your associates, I would advise them to read the patent
carefully. If in fact the patent covers all possible algorithms for doing
the thing in question, then they probably should reconsider their business
plan, or contact the patent holder and negotiate a license. If they in fact
have a significantly better way to do it, the patent holder might well be
interested in working with them on it.
Alternatively, there ARE procedures for patenting a modification to an
earlier patented invention. The modification must meet the same
requirements for utility, innovation, and non-obviousness that the original
invention had to meet to get the patent.
> I hope I am not getting too far off the subject, but you were using the
> patent definition of "mere aggregation" and I am trying to find out if I
> agree with that.
You might also want to read what Don Lancaster has to say about patents.
Briefly, he is against them. You can make a lot more money in the long run
by writing an article for the popular press, that stashes the idea squarely
in the public domain, and pursuing work based on that idea, than you can by
spending thousands of dollars pursuing a patent and then realizing that the
world has passed you by.
John Saeger
> You haven't been paying attention. I would like to bundle some
> version of TOPS-20 Macsyma in a pre-built TOPS-20 filesystem.
Do you have access to a version of TOPS-20 Macsyma to bundle on a pre-built
filesystem, or is this discussion purely theoretical?
John
Mark Crispin
> Glen, what you are OBVIOUSLY missing is that the Free Software Foundation
> has an agenda, and the GPL was designed to further that agenda.
I don't think that anyone is missing that.
Some of us object to being compelled to be the agents of someone else's
agenda, EVEN IF they may agree (at least in part) with that agenda.
> This is, essentially, what Gosling did to Stallman with Gosling/Unipress
> Emacs.
I think that it is remarkably ill-advised to make statements such as this.
I don't know the details of the dispute between Gosling vs. Stallman, but
I suspect that Mr. Gosling's side of the story has not been heard or
considered.
> It may help you to reread the last chapter of "Hackers", by Stephen Levy
Levy's book is not an authoritative history of the 1960s and 1970s hacker
community. It has, shall we say, a bias. Nor is Levy an historian; he is
a tabloid writer.
Mark Crispin
> Also consider that the GNU licenses have never been tested in court, and
> their exact meanings are commonly hotly debated. Which leads one to
> wonder if the GPL was as carefully constructed as is necessary. It seems
> that it would be wise to shy away from it, if you are concerned that the
> terms are not clearly understood, which they are not (and I believe that
> opinion is also held by at least some IP lawyers who have looked at it.)
Amen to all of the above. I have legal advice from IP lawyers which says
in effect "stay away from GPL."
I've known Richard Stallman for nearly 30 years. Chris Stacy has known
him for about 25 years. So it isn't as if we don't know anything about
RMS or his goals.
> Just because there is some opinion as to what it means, presented by
> advocates of the license, doesn't mean that a court would find that same
> meaning.
If I were to trust the claims of the GPL advocates, I would put my
financial resources at risk. I can not do so unless the GPL advocates are
willing to indemnify me for taking their advice.
-- Mark --
http://panda.com/mrc
Democracy is two wolves and a sheep deciding what to eat for lunch.
Liberty is a well-armed sheep contesting the vote.
Mark Crispin
> It is precisely the nature of a well-crafted legal instrument that it will
> never (have to) be tested in court.
It is a Fallacy of Accident to infer that because the GPL has not been
tested in court, it is therefore well-crafted.
Anyone who might want to test GPL faces two problems.
First, the cost of testing GPL in court is greater than that of
reimplementing the necessary software to be sanitized of GPL.
Second, the entities that use GPL often have little or no financial
resources to speak of, and thus prevailing would be a Pyrrhic victory.
Peter da Silva
John R. Strohm <str...@airmail.net> wrote:
>Belong To Us" license. You don't want to do that. You want to take the GPL
>code private.
I'm sorry, I don't see the bit where he's refusing to publish the source to
any GPL code.
If I write a script that uses a feature of a webserver that's only available
on a webserver that runs under the GPL, even if my code never runs on the same
computer as the GPL-licensed webserver, by your logic my code should be forced
to be distributed under the GPL.
Paul Rubin
> If I were to trust the claims of the GPL advocates, I would put my
> financial resources at risk. I can not do so unless the GPL advocates are
> willing to indemnify me for taking their advice.
I don't think any of us here are insurers and therefore we're not in a
position to write you a liability policy even if we wanted to.
However, just about every company making an embedded Linux gizmo, as
well as OS integrators like Red Hat, are bundling GPL'd and non-free
software together in pretty much the same way that you want to do with
Tops-20 and Macsyma.
Since those companies are real businesses, I expect they have
liability insurance that covers any possible fallout from this kind of
thing. So while I don't know if what you're doing is commercial
enough that you're willing to pay insurance premiums for the
indemnification that you want, it seems likely to me that if you are,
then there are insurers out there who are willing to take your
business.
Mark Crispin
> Since those companies are real businesses, I expect they have
> liability insurance that covers any possible fallout from this kind of
> thing. So while I don't know if what you're doing is commercial
> enough that you're willing to pay insurance premiums for the
> indemnification that you want, it seems likely to me that if you are,
> then there are insurers out there who are willing to take your
> business.
I see.
Thank you for clarifying that the purpose of the GPL is to forbid anyone
who is not a big enough company from using GPL software.
I didn't realize that Richard Stallman was such a supporter of big
business.
Glen Herrmannsfeldt
(snip)
> Discussion of patent law is non sequitur in the context of the GPL. The
GPL
> derives SOLELY from copyright law. Copyright law and patent law are two
> entirely different animals.
You used a patent case as a legal definition of "mere aggregation". If we
agree that patent and copyright law are entirely different, then I think we
have to say that example doesn't apply.
> > There is an application that some people I know are interested in
working
> > on. Others have previously worked in this area, and have patented an
> > algorithm for doing it. Not only that, but they have patented ALL
> > algorithms for doing it. That is they way they wrote the patent, and
> that is the way the USPTO granted it.
(snip)
> A patent is just a license to sue. Patents can be challenged, and
> frequently are. Sometimes they are upheld, and sometimes they are thrown
> out. The A&P Tea Company case I quoted was a patent challenge, and the
> Supreme Court overturned the patent. In the case of an actual patent of
> "teaching with workbooks", that patent will last about as long as it takes
> for the defense counsel to explain the patent to the judge and then show
the
> judge a class workbook that clearly predates the filing date on the
patent.
>
> In the case of your associates, I would advise them to read the patent
> carefully. If in fact the patent covers all possible algorithms for doing
> the thing in question, then they probably should reconsider their business
> plan, or contact the patent holder and negotiate a license. If they in
fact
> have a significantly better way to do it, the patent holder might well be
> interested in working with them on it.
I think the "all possible algorithms" part will get thrown out the first
time it gets to court, but it has to get there first. Someone with lots of
money, and a reason, needs to challenge it.
> Alternatively, there ARE procedures for patenting a modification to an
> earlier patented invention. The modification must meet the same
> requirements for utility, innovation, and non-obviousness that the
original
> invention had to meet to get the patent.
>
> > I hope I am not getting too far off the subject, but you were using the
> > patent definition of "mere aggregation" and I am trying to find out if I
> > agree with that.
>
> You might also want to read what Don Lancaster has to say about patents.
> Briefly, he is against them. You can make a lot more money in the long
run
> by writing an article for the popular press, that stashes the idea
squarely
> in the public domain, and pursuing work based on that idea, than you can
by
> spending thousands of dollars pursuing a patent and then realizing that
the
> world has passed you by.
Probably true in many fields.
-- glen
Charles Shannon Hendrix
> I've got a webpage that should point you to everything you'll need.
> http://www.aracnet.com/~healyzh/pdp10emu.html
> The page has pointers to emulators, software, and documentation (actually
> the page has pointers to just about every piece of PDP-10 related info on
> the web).
>
> BTW, TWENEX isn't available, but TOPS-20 is. TOPS-20 V7 on KLH10 is quite
> nice!
Yeah, I need a block of completely idle time to relive some old
memories.
Charles Shannon Hendrix
Mark Crispin wrote:
> Since Maxima is under GPL, that may mean that a TOPS-20 Maxima would have
> to be released individually, rather than bundled with a TOPS-20
> distribution, because of the viral aspects of GPL.
I'm not sure the GPL is quite that viral. Nothing I've read suggests it
cannot be bundled into a closed system.
What about the version of Macsyma that Maxima was based on? Shouldn't
that one be free of the copyright problem, or is that why GNU did the
Maxima software?
> Is there a version of Maxima that has a non-viral open source license
> (such as free-fork)?
Not that I've found. I don't like aspects of the GPL either, but I
don't think its necessary for what you want to do. Maybe you should ask
GNU about it or something?
I looked up a few things about Maxima, and this same topic has come up
before for other systems, including commercial ones. They seemed to
believe that the GPL didn't prevent Maxima from being bundled, but I've
not had time to read for more than 20 minutes.
So much to read, so little time...
Charles Shannon Hendrix
<Pine.LNX.4.55.03...@shiva0.cac.washington.edu>, Mark
Crispin wrote:
> Clause 2(b) of the GPL is a problem. It requires that the GPL be applied
> to the whole of any work that contains any part of any GPL software.
Imagine the implications for ALL copyrighted works if GNU actually won
such a case in court, setting precedent.
I don't see this as legally or logically defensible myself.
There are some lawyers like Lawrence Rosen who seem quite convinced that
this cannot happen, and isn't what the GPL is designed for. Maybe fire
off a message to him? Assuming you trust him of course.
> The GPL has no clause permitting waiver of any of its terms by permission
> of the copyright holder. In other words, even if the copyright holder of
> Maxima says that he doesn't mind if a TOPS-20 Maxima is distributed as
> part of a TOPS-20 distribution, the GPL still prohibits doing so.
Yet again, not court tested. It would be interesting to see what
a judge says about a license which appears (to me anyway) to break
copyright law.
> Some people claim that the GPL was poorly-written, and that these viral
> characteristics are unintentional. I disagree. I believe that the GPL
> was very carefully written to be viral.
The GPL's worst part, IMHO, is the idea that it is considered by GNU
to be a living license. What I mean is that if you license under GPL
version N, they claim that when they write GPL version N+M, your code
falls under that license.
Of course, plenty of commercial licenses and contracts, in everything
from software to banking, do the same thing.
Charles Shannon Hendrix
<Pine.LNX.4.55.03...@shiva1.cac.washington.edu>, Mark
Crispin wrote:
> Don't be so sure about that. The way that the GPL is worded can be
> construed to give standing to any licensee who alleges that his rights
> under the GPL were infringed.
Something to think about, with regard to GNU people saying that the GPL
is simply not worded well: they've had years to write a new version
which fixes the problems.
> For that matter, I don't think that people residing in the US can legally
> distribute GPL version 2 software at all.
There are legal experts who say otherwise. I wish I knew. I'm sure a
lot of people do.
The Linux kernek is under GPL 2.0 I believe, permanently. Linus
Torvalds evidently stated that Linux would never be under the "living
GPL" license. Even the latest kernel sources are under the June 1991
version 2 GPL.
> You can claim that these aren't real problems until you're blue in the
> face. I don't think that you have sufficient assets to indemnify me in
> the event that you are wrong. I don't think that you can find an insurer
> who'd underwrite such indemnification either.
I think the very fact that there is so much controversy is itself a
problem.
However I still think that it would be OK to distribute Maxima with
TOPS-20.
Of course, it would be better to get Macsyma and update it like Maxima
has been updated, but that's a lot of work for someone.
Charles Shannon Hendrix
> Late 1800s, in spite of scientific testimony to the contrary,
> the U.S. Supreme Court declared that the tomato is *not* a fruit,
> but is rather instead a vegetable.
>
> Nuts deciding between fruits and vegetables...
Very good!
I think this is signature fodder for sure.
Charles Shannon Hendrix
Crispin wrote:
> If I were to trust the claims of the GPL advocates, I would put my
> financial resources at risk. I can not do so unless the GPL advocates are
> willing to indemnify me for taking their advice.
That seems to settle it then.
So back to the topic, which I'm interested in anyway, what's the plan to
get Macsyma into TOPS-20?
Also, how about contacting the author of Maxima to see if he will
release the code under a non-GPL license.
What if he released it under BSD, would that be OK?
Mark Crispin
> Also, how about contacting the author of Maxima to see if he will
> release the code under a non-GPL license.
> What if he released it under BSD, would that be OK?
Any non-viral open source license would be good.
Christopher C. Stacy
Charles> In article <Pine.WNT.4.55.03...@Shimo-Tomobiki.Panda.COM>,
Charles> Mark Crispin wrote:
>> Since Maxima is under GPL, that may mean that a TOPS-20 Maxima would have
>> to be released individually, rather than bundled with a TOPS-20
>> distribution, because of the viral aspects of GPL.
Charles> I'm not sure the GPL is quite that viral. Nothing I've read suggests it
Charles> cannot be bundled into a closed system.
Charles> What about the version of Macsyma that Maxima was based on? Shouldn't
Charles> that one be free of the copyright problem, or is that why GNU did the
Charles> Maxima software?
The Maxima software is a port, by the University of Texas, of a
program that was written at MIT, paid for by the U.S. Government
(Department of Defense and Department of Energy).
"GNU" had nothing to do with it.
Mark Crispin
> The Maxima software is a port, by the University of Texas, of a
> program that was written at MIT, paid for by the U.S. Government
> (Department of Defense and Department of Energy).
> "GNU" had nothing to do with it.
Other than their unfortunate decision to use the GPL instead of an
alternative non-viral open source license, that is.
Charles Shannon Hendrix
> Charles> What about the version of Macsyma that Maxima was based on?
> Shouldn't Charles> that one be free of the copyright problem, or is
> that why GNU did the Charles> Maxima software?
>
> The Maxima software is a port, by the University of Texas, of a
> program that was written at MIT, paid for by the U.S. Government
> (Department of Defense and Department of Energy).
> "GNU" had nothing to do with it.
I should have not said GNU, but the question still remains.
Mark Crispin
> > What about the version of Macsyma that Maxima was based on?
> > that why GNU did the Maxima software?
> I should have not said GNU, but the question still remains.
As a guess, we are talking about DOE Macsyma; and the University of Texas
had to obtain permission from DOE before placing Maxima under GPL.
I don't know how one would go about obtaining DOE Macsyma, much less
securing permission from DOE this many years later. My guess is that DOE
washed their hands of it after giving UTexas the right to distribute
Maxima under GPL.
Glen Herrmannsfeldt
> "Glen Herrmannsfeldt" <g...@ugcs.caltech.edu> wrote in message
> news:PT7oa.521510$3D1.286357@sccrnsc01...
> [ snip ]
> > Still no comments on my physical vs. logical aggregation question.
>
> Section 2 of the GPL reads as follows:
>
> 2. You may modify your copy or copies of the Program or any portion
> of it, thus forming a work based on the Program, and copy and
> distribute such modifications or work under the terms of Section 1
> above, provided that you also meet all of these conditions:
>
> a) You must cause the modified files to carry prominent notices
> stating that you changed the files and the date of any change.
>
> b) You must cause any work that you distribute or publish, that in
> whole or in part contains or is derived from the Program or any
> part thereof, to be licensed as a whole at no charge to all third
> parties under the terms of this License.
>
> Your "physical aggregation" scheme (your item 2) appears to be a unified
> work, and as such would fall under section 2 (b) of the GPL.
This is what I had always thought before. Though now you are going on what
GPL says, instead of what the legal system says.
> Your "logical aggregation" scheme (your item 1) appears to be an obvious
> attempt to circumvent the clear intent of section 2 (b) of the GPL.
Courts
> tend to take dim views of that kind of shenanigan.
So, for example, if I send a CD of Windows 2000, and a CD of the windows
version of GCC I would be considered in violation? That is pretty bad for
everyone.
-- glen
John R. Strohm
Come on, Glen, you can troll better than that.
Recalling your exact words from a saved local copy of your original message.
"1) Program A requires members of GPL'ed library B to work. I supply two
separate CD's, which must be ordered separately, will be mailed in separate
envelopes. The receiver of the two CD's runs an install program which
combines them so that they run together. A requires B to work, always."
Ignoring for the moment the question of whether you can legally send a CD of
Windows 2000, in *THIS* example scenario, "Program A" (Windows 2000) most
emphatically does not require (members of) "GPL'ed library B" (GCC) to
operate.
Your example this time around is simply NOT APPLICABLE.
Peter da Silva
John R. Strohm <str...@airmail.net> wrote:
>"1) Program A requires members of GPL'ed library B to work. I supply two
>separate CD's, which must be ordered separately, will be mailed in separate
>envelopes. The receiver of the two CD's runs an install program which
>combines them so that they run together. A requires B to work, always."
What if that library is on a GPLed webserver and the application
needs that library to perform the required service... and 'calls'
that library through HTTP?
>Ignoring for the moment the question of whether you can legally send a CD of
>Windows 2000, in *THIS* example scenario, "Program A" (Windows 2000) most
>emphatically does not require (members of) "GPL'ed library B" (GCC) to
>operate.
If Program A is Internet Explorer?
Glen Herrmannsfeldt
Oops, sorry, wrong example. I thought I had one that didn't say which
depended on the other.
OK, how about a real example. There is a company with a real product I
know of:
Program A has an option that uses GPL'ed library B. Because of the linkage
requirement they distribute B as a dynamic-link library, instead of
statically linking it. Because DLL's are resolved at load time, it is not
possible to use A without B, even if no features from B are actually being
used. Since I wasn't actually using any features from B, I wanted a dummy
DLL that didn't do anything (maybe print out a message), but they claimed
that GPL wouldn't let them do that.
I don't remember now if they actually shipped it on the same CD, or required
the user to download it from the original source. In any case, it was
required, linkage resolved by the user at load time, and carefully arranged
to get around GPL restrictions.
-- glen
Eric Smith
> As a guess, we are talking about DOE Macsyma; and the University of Texas
> had to obtain permission from DOE before placing Maxima under GPL.
Who holds the copyright of DOE Macsyma? If someone wants to get DOE Macsyma
released under a different license, the copyright holder is the party that
must approve of it.
The DOE, as a part of the United States Government, generally cannot
copyright any works they create. However, Macsyma was not originally
created by the DOE, so it may hold a copyright by another party, or that
party's copyright may have been assigned to the DOE.
Paul Rubin
> Who holds the copyright of DOE Macsyma? If someone wants to get DOE Macsyma
> released under a different license, the copyright holder is the party that
> must approve of it.
>
> The DOE, as a part of the United States Government, generally cannot
> copyright any works they create. However, Macsyma was not originally
> created by the DOE, so it may hold a copyright by another party, or that
> party's copyright may have been assigned to the DOE.
There was a lot of hassle (maybe even litigation) needed to get DOE
Macsyma free of Symbolics interference. Other people here probably
know the story better than I do. However, I doubt Macsyma as it came
from the DOE is directly useable. U Texas (I guess that means Bill
Schelter) had to do additional work to port it to Common Lisp. So
even if DOE Macsyma is now uncopyrighted (which I doubt), the porting
work may not be.
Carl Shapiro
> There was a lot of hassle (maybe even litigation) needed to get DOE
> Macsyma free of Symbolics interference. Other people here probably
> know the story better than I do. However, I doubt Macsyma as it came
> from the DOE is directly useable. U Texas (I guess that means Bill
> Schelter) had to do additional work to port it to Common Lisp. So
> even if DOE Macsyma is now uncopyrighted (which I doubt), the porting
> work may not be.
Why would Macsyma have to be ported to Common Lisp? Macsyma runs just
fine on MacLisp, which in turn runs just fine on the PDP-10.
Paul Rubin
> Why would Macsyma have to be ported to Common Lisp? Macsyma runs just
> fine on MacLisp, which in turn runs just fine on the PDP-10.
I thought we were talking about Maxima, which (unless I'm confused
which I might well be) is a version ported to Common Lisp to run on
later hardware. I'm not sure how it relates to Vaxima which ran under
Franz Lisp on the Vax.
John R. Strohm
"Peter da Silva" <pe...@taronga.com> wrote in message
news:b81ctt$1g7u$1...@citadel.in.taronga.com...
> In article
<2873682A5E395BF5.25DD73EC...@lp.airnews.net>,
> John R. Strohm <str...@airmail.net> wrote:
> >"1) Program A requires members of GPL'ed library B to work. I supply
two
> >separate CD's, which must be ordered separately, will be mailed in
separate
> >envelopes. The receiver of the two CD's runs an install program which
> >combines them so that they run together. A requires B to work, always."
>
> What if that library is on a GPLed webserver and the application
> needs that library to perform the required service... and 'calls'
> that library through HTTP?
I think I don't understand this scenario. Would you care to elaborate?
> >Ignoring for the moment the question of whether you can legally send a CD
of
> >Windows 2000, in *THIS* example scenario, "Program A" (Windows 2000) most
> >emphatically does not require (members of) "GPL'ed library B" (GCC) to
> >operate.
>
> If Program A is Internet Explorer?
I really don't understand this scenario.
In the first place, Microsoft Internet Explorer is most certainly NOT
GPL'ed. In the second place, Microsoft Internet Explorer most certainly
does not require GCC to run, nor does GCC require Microsoft Internet
Explorer to run. If you believe Microsoft, Windows 2000 does require some
code that is in Microsoft Internet Explorer in order to run, which is, as I
recall, part of what triggered one of the antitrust cases that Microsoft
LOST on the facts but won in the punishment phase.
John R. Strohm
Well, no one said you HAD to buy their product.
Disclaimer: I am not a lawyer, and this is not legal advice. I am assuming
that the DLL is in fact GPL'ed, and not LGPL'ed. (It makes a difference.)
In practical terms: they had to provide SOME kind of interface code to hook
up the DLL. Most likely, they #include'd a GPL'ed header file, and that MAY
be enough to bring them under the GPL. It depends on what they use from the
GPL'ed DLL. Someone would have to take a fairly hard look at their
executable, and see what it called from the DLL, and then look at the DLL to
see what they had to build or copy in order to set up the call. After that,
assuming that it was reasonably conclusive that they had overstepped the
bounds, the person that holds the copyright on the DLL would have to take
action, because he is the only one with standing.
Personally, I think their action is reprehensible. It may or may not be a
violation of the GPL.
Personally, I'd look into building a competing app, that did it better and
faster, and releasing it under GPL.
John R. Strohm
news:ouyhe8r...@panix3.panix.com...
It isn't that simple.
MIT developed Macsyma, with funding from various people. MIT sold *ALL*
*RIGHTS* to Macsyma, apparently including some rights that they didn't
actually have to sell, to Symbolics. The Department of Energy said "NOW
WAIT JUST A COTTON-PICKING MINUTE! We own some data rights and you don't
have any authority to sell our rights to Symbolics." I believe litigation
ensued. When the dust had settled, DOE had acquired the copyright to a
particular version of Macsyma, specifically a version that ran under NIL
(New Implementation of LISP) on a VAX, presumably under VMS, and Symbolics
held the copyrights to all other existing versions. DOE authorized the guy
at UT Austin to port their version to Common LISP under Unix, and later to
release Maxima, the ported version, under the GPL.
In short, Symbolics (or one of the successor companies) holds the copyright
on MacLisp Macsyma, and they aren't likely to let go of it any time soon.
Paul Rubin
> > OK, how about a real example. There is a company with a real product I
> > know of:
> >
> > Program A has an option that uses GPL'ed library B. Because of
> > the linkage requirement they distribute B as a dynamic-link
> > library, instead of statically linking it. Because DLL's are
> > resolved at load time, it is not possible to use A without B, even
> > if no features from B are actually being used. Since I wasn't
> > actually using any features from B, I wanted a dummy DLL that
> > didn't do anything (maybe print out a message), but they claimed
>
> Well, no one said you HAD to buy their product.
>
> Disclaimer: I am not a lawyer, and this is not legal advice. I am assuming
> that the DLL is in fact GPL'ed, and not LGPL'ed. (It makes a difference.)
> Personally, I think their action is reprehensible. It may or may not be a
> violation of the GPL.
According to the FSF, that is a GPL violation. It's been discussed
countless times and nobody knows for sure what would happen if such a
case actually went to court. The FSF is apparently willing to be a
plaintiff in such a case, but nobody has volunteered to be a defendant.
Glen Herrmannsfeldt
"John R. Strohm" <str...@airmail.net> wrote in message
news:114695C87B8620B5.5503FA98...@lp.airnews.net...
>
> "Peter da Silva" <pe...@taronga.com> wrote in message
> news:b81ctt$1g7u$1...@citadel.in.taronga.com...
> > In article
> <2873682A5E395BF5.25DD73EC...@lp.airnews.net>,
> > John R. Strohm <str...@airmail.net> wrote:
> > >"1) Program A requires members of GPL'ed library B to work. I supply
> two
> > >separate CD's, which must be ordered separately, will be mailed in
> separate
> > >envelopes. The receiver of the two CD's runs an install program which
> > >combines them so that they run together. A requires B to work,
always."
> >
> > What if that library is on a GPLed webserver and the application
> > needs that library to perform the required service... and 'calls'
> > that library through HTTP?
>
> I think I don't understand this scenario. Would you care to elaborate?
I think I understand it. GNU defines the linking of two programs in terms
of the data structures they share. So if IE is sharing data with a remote
web server, that could be considered linked in GNU terms. Though I might
say that it is not IE that is doing the linking, but more likely a script
file running under IE. The script is interpreted by IE, but the script
commands are in control. This would be comparable to a microcoded machine
executing a program.
I think there are a number of cases one could make similar arguments over
other cases. Can a proprietary OS run a GPL'ed program? At one point
linking was defined as being in the same address space, and many OS run in
the same address space as the program being run. Some even run all programs
in the same address space. Programs and OS share much data, which would
seem to fit the "linked" definition of GPL.
-- glen
jmfb...@aol.com
Running in the same address space could mean everything that
was run under timesharing since code and data can get swapped
out to make room for another completely unrelated job.
/BAH
Subtract a hundred and four for e-mail.
John R. Strohm
Glen, I am beginning to get the distinct impression that you have not
actually READ the GPL.
Essentially, the restrictions imposed by the GPL only come into play when
you create AND DISTRIBUTE a modified or derivative work. If you do not
DISTRIBUTE that work, if you just create it and keep it private, you can do
anything you want.
As for whether a proprietary OS can run a GPL'ed program: Look at (for
example) GNU Emacs or gcc. Look at the lists of machines and operating
systems for which ports are available, UNDER GPL, and you will find PLENTY
of proprietary sytems on those lists. I do not know offhand whether those
proprietary system distribution media include the various GPL'ed programs or
not, but I would *NOT* be surprised if they did.
John R. Strohm
news:7xr87v6...@ruckus.brouhaha.com...
By the way, are you the Paul Rubin who was at UT Austin in the 1970s? I
think we shared a desk on Don Good's project at one time.
Paul Rubin
> By the way, are you the Paul Rubin who was at UT Austin in the 1970s? I
> think we shared a desk on Don Good's project at one time.
Nope. Somebody (maybe you) asked me that same question by email a few
months ago. I've never been to Austin.
Douglas H. Quebbeman
> In article <3ea03d25$1...@news.iglou.com>,
> Douglas H. Quebbeman <do...@iglou.com> wrote:
> >Late 1800s, in spite of scientific testimony to the contrary,
> >but is rather instead a vegetable.
>
> What's wrong with that? It *is* a vegetable. The only fruits I know of
> that aren't vegetable are mushrooms.
There is a distinction between fruits and vegetables...
http://www.ag.uiuc.edu/~robsond/solutions/horticulture/docs/tomato.html
But it has attributes of both. Most fruits grow on "woody"
plants, as opposed to vegetable plants...
-dq
Peter da Silva
John R. Strohm <str...@airmail.net> wrote:
>> >Windows 2000, in *THIS* example scenario, "Program A" (Windows 2000) most
>> >emphatically does not require (members of) "GPL'ed library B" (GCC) to
>> >operate.
>> If Program A is Internet Explorer?
>I really don't understand this scenario.
>In the first place, Microsoft Internet Explorer is most certainly NOT
>GPL'ed.
No, that's why it's program "A".
>In the second place, Microsoft Internet Explorer most certainly
>does not require GCC to run, nor does GCC require Microsoft Internet
>Explorer to run.
I'm not buying "Internet Explorer", I'm buying "the functionality in program
B", so to get the functionality I'm paying for I need the GPLed program. But
the browser is also part of the total package I'm buying. Why does the fact
that IE calls the GPLed program through HTTP instead of calling a DLL make
a difference?
If it does make a difference, then I'll just write an HTTP interface to the
GPL-ed library, and add an HTTP-based client-server interface that the
application uses to access that functionality.
Peter da Silva
Douglas H. Quebbeman <do...@iglou.com> wrote:
>"Peter da Silva" <pe...@taronga.com> wrote in message
>news:b7prdt$so6$1...@citadel.in.taronga.com...
>> In article <3ea03d25$1...@news.iglou.com>,
>> Douglas H. Quebbeman <do...@iglou.com> wrote:
>> >Late 1800s, in spite of scientific testimony to the contrary,
>> >the U.S. Supreme Court declared that the tomato is *not* a fruit,
>> >but is rather instead a vegetable.
>> What's wrong with that? It *is* a vegetable. The only fruits I know of
>> that aren't vegetable are mushrooms.
>There is a distinction between fruits and vegetables...
Sure, a fruit is the part of a vegetable or fungus that carries the seeds
or spores.
>http://www.ag.uiuc.edu/~robsond/solutions/horticulture/docs/tomato.html
>But it has attributes of both. Most fruits grow on "woody"
>plants, as opposed to vegetable plants...
Woody plants are also vegetables, unless they're animals, bacteria, fungi,
or viruses. Or machines, I guess, but we're not that advanced yet.
Peter da Silva
John R. Strohm <str...@airmail.net> wrote:
>Essentially, the restrictions imposed by the GPL only come into play when
>you create AND DISTRIBUTE a modified or derivative work. If you do not
>DISTRIBUTE that work, if you just create it and keep it private, you can do
>anything you want.
And the question we're talking about is "what's a derived work"?
Let's say you have a GPLed web board, and I write a program that provides
an Aqua interface to that web board. Is my program a derived work? It depends
explicitly and inherently on a GPLed program, even though it doesn't "link"
to the program nor is it distributed with the program.
If the author of the web board tried to get me to release my Aqua program
under the GPL, wouldn't that be an attempt to abuse an interface copyright?
Would the same people have to support both sides of the lawsuit, wearing
their FSF hats on one side of the room, and switching to the LPF hat to cross
the courtroom to the other side?
What's the difference between an HTTP interface and a DLL interface?
Glen Herrmannsfeldt
<jmfb...@aol.com> wrote in message news:b838kr$cbm$1...@bob.news.rcn.net...
> In article <bi4pa.564009$L1.164790@sccrnsc02>,
> Running in the same address space could mean everything that
> was run under timesharing since code and data can get swapped
> out to make room for another completely unrelated job.
Even when not swapped. OS/360, OS/VS1, and the first versions of OS/VS2 ran
everything in a single address space. There are storage protection keys to
keep user programs from accessing other user programs, but only one address
space.
On Windows systems, part of the address space is reserved for the OS, and
the other part is used by the currently running program. Definitely
proprietary windows code and possibly GPL'ed user code is in the same
address space, and sharing complex data structures.
GNU makes a distinction between programs started by exec(), or some similar
call, and sharing only small amounts of data, like command line parameters.
This is definitely not the way programs run under any version of Windows.
-- glen
Glen Herrmannsfeldt
(snip)
> Glen, I am beginning to get the distinct impression that you have not
> actually READ the GPL.
I did actually read it some years ago, when I first started working with GPL
software. Last week I did a Google (tm) search on "mere aggregation" which
brought up many FSF web pages, and I read some of them.
> Essentially, the restrictions imposed by the GPL only come into play when
> you create AND DISTRIBUTE a modified or derivative work. If you do not
> DISTRIBUTE that work, if you just create it and keep it private, you can
do
> anything you want.
Yes, but software is much more useful when you can distribute it. My first
work with GPL software was the GNU file utilities and similar utilities when
I was using OS/2 version 1.0 and 1.2. I ported some for my own use. The
compiler I was using would also generate DOS executables, so I did that, and
gave copies to friends running DOS. I put the source on the same floppy, as
required, though everyone would ask why I did that.
> As for whether a proprietary OS can run a GPL'ed program: Look at (for
> example) GNU Emacs or gcc. Look at the lists of machines and operating
> systems for which ports are available, UNDER GPL, and you will find PLENTY
> of proprietary sytems on those lists. I do not know offhand whether those
> proprietary system distribution media include the various GPL'ed programs
or
> not, but I would *NOT* be surprised if they did.
Yes, and according to the ones I was reading, I believe those are in
violation of GPL. I am trying to find the boundary between what is legal
and what isn't, in real cases when possible, according to the GPL. I
appreciate your better understanding of the legal system, and the precise
meaning of some words.
One question I am asking is, does it make sense technically to make the
distinctions that FSF is making? Can one make an unambiguous distinction
between "mere aggregation" and a more intimate linking of two programs?
If those companies can distribute their software, then why is it that Mr.
Crispin can't distribute his?
-- glen
Douglas H. Quebbeman
it has attributes of both. Most fruits grow on "woody"
> >plants, as opposed to vegetable plants...
>
> Woody plants are also vegetables, unless they're animals, bacteria, fungi,
> or viruses. Or machines, I guess, but we're not that advanced yet.
I think a biologist *might* find your categories a bit
too imprecise. If the language level you're embracing
is common speech, sure, but this is a technical forum,
albeit one which is not devoted to vegetables (although
they appear to play quite a role from time to time...)
;)
John R. Strohm
> Yes, and according to the ones I was reading, I believe those are in
> violation of GPL. I am trying to find the boundary between what is legal
> and what isn't, in real cases when possible, according to the GPL. I
> appreciate your better understanding of the legal system, and the precise
> meaning of some words.
You are not going to find a sharp white/black boundary, like the one between
lunar day and lunar shadow. The best you will find is a gradual shading
from light to dark, with varying shades of grey. A well-crafted legal
document creates such a shaded continuum, to persuade a would-be scoundrel
not to try to sail too close to the edge of the world.
> One question I am asking is, does it make sense technically to make the
> distinctions that FSF is making? Can one make an unambiguous distinction
> between "mere aggregation" and a more intimate linking of two programs?
Yes, it does, and I believe one can.
The fundamental idea is that people should be able to use GPL'ed software,
and people should be able to redistribute GPL'ed software freely, and people
should be able to modify GPL'ed software, and redistribute the modifications
freely, but people should NOT be able to modify it and CHARGE for the
modified versions, nor should one be able to make a work that uses (links)
the GPL'ed code and charge for it.
It is worth noting that the Lesser GPL came about precisely because of
concerns raised by linking. Too many people were shying away from using a
GPL'ed standard library, and this was affecting the acceptance of the GNU
project goals. As a straw man example, who in their right minds would use a
development platform that requires them to release ALL of the code developed
on that platform to the public domain, or to some organization whose
interests might not coincide with theirs? (Note: I feel VERY safe in saying
that Microsoft would demand such a license for their products if they
thought they could get away with it.)
The language about "mere aggregation" allows the making of "shovelware"
CD-ROMs, for example, where one might have half a dozen, or even several
HUNDRED, unrelated programs, with different licenses, all on the same
physical media. This creates a convenience for the user, and makes it
easier to redistribute GPL'ed software.
There isn't any real doubt about the intent of the GPL. All of the
discussion is over the precise radiometry of the various regions of the grey
areas, and how to fake out the radiometer.
> If those companies can distribute their software, then why is it that Mr.
> Crispin can't distribute his?
Mr. Crispin has expressed fear of the legal system. That fear is,
unfortunately, fairly well-founded. It is my personal opinion that juries
actually do pretty well at dispensing justice, when they finally go into the
jury room and hash out their decisions, but it can cost a LOT of money to
get to that point. Good attorneys charge dozens to hundreds of dollars per
hour. My opinion is that Mr. Crispin's uncertainty about the GPL, combined
with his fear of the legal system and the traditional penury of University
professors, combine to persuade him to take the safe route.
Now, it may well be that Mr. Crispin is correct in his particular instance,
but not so much because of the GPL as because of the DEC Hobbyist license.
A casual reading of the Hobbyist license does not convince me that
(re)distribution of a prebuilt TOPS-20 is allowed at all, regardless of
whether GPL'ed software is merely aggregated or not.
John R. Strohm
> In article
<C0F6CC7CF6AAD9B5.473D490F...@lp.airnews.net>,
> John R. Strohm <str...@airmail.net> wrote:
> >Essentially, the restrictions imposed by the GPL only come into play when
> >you create AND DISTRIBUTE a modified or derivative work. If you do not
> >DISTRIBUTE that work, if you just create it and keep it private, you can
do
> >anything you want.
>
> And the question we're talking about is "what's a derived work"?
>
> Let's say you have a GPLed web board, and I write a program that provides
> an Aqua interface to that web board. Is my program a derived work? It
depends
> explicitly and inherently on a GPLed program, even though it doesn't
"link"
> to the program nor is it distributed with the program.
That is the key question, isn't it? My feeling would be that it probably
would be considered a derived work.
> If the author of the web board tried to get me to release my Aqua program
> under the GPL, wouldn't that be an attempt to abuse an interface
copyright?
No, it would not, not if it is a derived work.
In general, you are not allowed to make derived works at all, unless the
copyright holder of the original work grants you permission. In the case at
hand, he did grant that permission, subject to the terms of the GPL. By
making the derived work, you accepted his terms. If you now choose to
distribute your derived work, you are required to do so in accordance with
the terms under which you made the work.
You of course have a choice. You don't have to distribute your derived work
at all.
> Would the same people have to support both sides of the lawsuit, wearing
> their FSF hats on one side of the room, and switching to the LPF hat to
cross
> the courtroom to the other side?
>
> What's the difference between an HTTP interface and a DLL interface?
Probably nothing significant, as far as the determination of what
constitutes a "derivative work".
Paul Rubin
> One question I am asking is, does it make sense technically to make the
> distinctions that FSF is making? Can one make an unambiguous distinction
> between "mere aggregation" and a more intimate linking of two programs?
It's completely normal in the legal arena for it to be impossible to
make a sharp distinction between the legal and illegal. That's why
the whole court system exists--to decide on a case by case basis where
specific things fall, because most legal boundaries are in actuality
blurry. The law is not math and it does not promise or aim to deliver
mathematical precision.
> If those companies can distribute their software, then why is it that Mr.
> Crispin can't distribute his?
Mark is an extremely intelligent guy and I think he's letting himself
get carried away thinking up far-out theoretical ways to get into GPL
trouble, that (like almost anything) can't proven absolutely
impossible, but which probably don't have much to do with real life.
Most other people, including those companies, muddle through life
either not worrying about the highly improbable and going ahead taking
small risks, or else by paying someone else (i.e. an insurance
company) to assume the risk for them. Normally, nothing bad happens
as long as they act reasonably.
FWIW, I've never heard of someone getting sued over copyright (except
for blatant intentional infringement) without first getting a "hey,
that's ours, stop that or else" letter and plenty of opportunity to
avoid getting sued by simply saying "oops" and stopping whatever the
disputed activity is. IANAL, YMMV, etc.
Mark Crispin
> There isn't any real doubt about the intent of the GPL.
It depends upon how you define "doubt". There is certainly a great deal
of misinformation about the intent of the GPL, not the least propagated by
its advocates.
Among other things, there is quite a bit more to the GPL than merely:
> The fundamental idea is that people should be able to use GPL'ed
> software, and people should be able to redistribute GPL'ed software
> freely, and people should be able to modify GPL'ed software, and
> redistribute the modifications freely, but people should NOT be able to
> modify it and CHARGE for the modified versions, nor should one be able
> to make a work that uses (links) the GPL'ed code and charge for it.
GPL is a very restrictive license. GPL also attempts to force everything
that it touches to also be GPL.
It is ingenuous to claim, as some do, that GPL is about "freedom." GPL is
about control.
-- Mark --
http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.
Glen Herrmannsfeldt
"Paul Rubin" <http://phr...@NOSPAM.invalid> wrote in message
> "Glen Herrmannsfeldt" <g...@ugcs.caltech.edu> writes:
> > One question I am asking is, does it make sense technically to make the
> > distinctions that FSF is making? Can one make an unambiguous
distinction
> > between "mere aggregation" and a more intimate linking of two programs?
>
> It's completely normal in the legal arena for it to be impossible to
> make a sharp distinction between the legal and illegal. That's why
> the whole court system exists--to decide on a case by case basis where
> specific things fall, because most legal boundaries are in actuality
> blurry. The law is not math and it does not promise or aim to deliver
> mathematical precision.
(snip)
> FWIW, I've never heard of someone getting sued over copyright (except
> for blatant intentional infringement) without first getting a "hey,
> that's ours, stop that or else" letter and plenty of opportunity to
> avoid getting sued by simply saying "oops" and stopping whatever the
> disputed activity is. IANAL, YMMV, etc.
I don't know that I have heard details about many, but I think I agree with
this. If it is gray enough to get this much discussion, it should be enough
to get a warning before going to court.
-- glen