GNU readline usage
Paul Emsley
Is this sort of thing allowed? I thought readline was GPLed?
http://origo.imsb.au.dk/~mok/o-linux/
The program seems proprietary, since it needs a licence code
and has an expiry date.
Paul.
Paul D. Smith
pe> Is this sort of thing allowed? I thought readline was GPLed?
pe> http://origo.imsb.au.dk/~mok/o-linux/
pe> The program seems proprietary, since it needs a licence code and
pe> has an expiry date.
It says right in it that it will be distributed as binary only.
This clearly violates the Readline license; the author should be
informed that he will either have to change the licensing or remove
readline from the program.
--
-------------------------------------------------------------------------------
Paul D. Smith <psm...@baynetworks.com> Network Management Development
"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
-------------------------------------------------------------------------------
These are my opinions---Nortel Networks takes no responsibility for them.
Phil Hunt
psm...@baynetworks.com "Paul D. Smith" writes:
> %% Paul Emsley <pa...@chem.gla.ac.uk> writes:
>
> pe> Is this sort of thing allowed? I thought readline was GPLed?
>
> pe> http://origo.imsb.au.dk/~mok/o-linux/
>
> pe> The program seems proprietary, since it needs a licence code and
> pe> has an expiry date.
>
> It says right in it that it will be distributed as binary only.
>
> This clearly violates the Readline license; the author should be
> informed that he will either have to change the licensing or remove
> readline from the program.
Surely the author is also obliged to give full source code to anyone he
has distributed copies of the binary to?
--
Phil Hunt....philh@vision25.demon.co.uk
Paul D. Smith
>> This clearly violates the Readline license; the author should be
>> informed that he will either have to change the licensing or remove
>> readline from the program.
ph> Surely the author is also obliged to give full source code to
ph> anyone he has distributed copies of the binary to?
Well. One of those people would essentially have to take him to court
to force him to do that, if he didn't want to.
I consider it likely that the courts would allow him to back out of that
requirement as long as he complied with the license as soon as it was
brought to his attention. Likely he'd be forced to provide a full
refund at least for anyone who wanted it. It's extremely unlikely he'd
be forced to comply with the GPL.
This kind of thing has happened a few times in the past and the FSF has
never required that companies comply with the GPL on already-distributed
copies; they basically have said either comply now, or stop using the
GPL'd code immediately.
Of course, IANAL.
Ken Arromdee
> >> This clearly violates the Readline license; the author should be
> >> informed that he will either have to change the licensing or remove
> >> readline from the program.
> ph> Surely the author is also obliged to give full source code to
> ph> anyone he has distributed copies of the binary to?
>Well. One of those people would essentially have to take him to court
>to force him to do that, if he didn't want to.
>I consider it likely that the courts would allow him to back out of that
>requirement as long as he complied with the license as soon as it was
>brought to his attention. Likely he'd be forced to provide a full
>refund at least for anyone who wanted it. It's extremely unlikely he'd
>be forced to comply with the GPL.
I don't think he can be forced (and I've argued this before).
The GPL gets its effect from copyright--since only the GPL gives you the right
to distribute GPLed code, it can outline the conditions under which you
distribute, so you either agree to the GPL and follow those conditions, or
you don't agree to the GPL and you're a copyright violator.
Someone can always distribute anyway, _not_ agree to the GPL, and their crime
is copyright violation, not license violation. You can sue them for
copyright infringement, but you can't sue them to make them comply with the
license, since they haven't accepted the license.
>This kind of thing has happened a few times in the past and the FSF has
>never required that companies comply with the GPL on already-distributed
>copies; they basically have said either comply now, or stop using the
>GPL'd code immediately.
Probably because the FSF's lawyers know better (despite the wording in the
GPL itself).
--
Ken Arromdee |They said it was *daft* to build a space
arro...@inetnow.net |station in a swamp, but I showed them! It
karr...@nyx.nyx.net |sank into the swamp. So I built a second
http://www.inetnow.net/~arromdee|space station. That sank into the swamp too.
--------------------------------+My third space station sank into the swamp.
So I built a fourth one. That fell into a time warp and _then_ sank into the
swamp. But the fifth one... stayed up! --Monty Python/Babylon 5
NF Stevens
>%% ph...@vision25.demon.co.uk (Phil Hunt) writes:
>
> >> This clearly violates the Readline license; the author should be
> >> informed that he will either have to change the licensing or remove
> >> readline from the program.
>
> ph> Surely the author is also obliged to give full source code to
> ph> anyone he has distributed copies of the binary to?
>
>Well. One of those people would essentially have to take him to court
>to force him to do that, if he didn't want to.
>
>I consider it likely that the courts would allow him to back out of that
>requirement as long as he complied with the license as soon as it was
>brought to his attention. Likely he'd be forced to provide a full
>refund at least for anyone who wanted it. It's extremely unlikely he'd
>be forced to comply with the GPL.
>
>This kind of thing has happened a few times in the past and the FSF has
>never required that companies comply with the GPL on already-distributed
>copies; they basically have said either comply now, or stop using the
>GPL'd code immediately.
>
>Of course, IANAL.
I too ANAL but if the author in question is distributing
his program _dynamically_ linked against libreadline then
you would have to convince the court that dynamic linking
creates a derived work. I am not sure that you would be
successful in this.
Norman
Steven G. Johnson
> Someone can always distribute anyway, _not_ agree to the GPL, and their crime
> is copyright violation, not license violation. You can sue them for
> copyright infringement, but you can't sue them to make them comply with the
> license, since they haven't accepted the license.
In any case, they have to stop distributing their program, or agree to the
GPL in order to have the legal right to distribute. (Or, they can stop
using GPLed code.)
Cordially,
Steven G. Johnson
Paul D. Smith
amr> On 03 Jun 1999 13:30:16 -0400, Paul D. Smith
amr> <psm...@baynetworks.com> wrote:
>> Well. One of those people would essentially have to take him to court
>> to force him to do that, if he didn't want to.
amr> Has anyone simply written a polite letter to the author of this
amr> code, explained that readline is under the GPL, and asked him to
amr> change his license terms?
Don't know, but that's not the point. Someone was asking whether he had
to release the source to the people who already had the code, and that's
the (hypothetical) question we're discussing.
Tim Smith
>Someone can always distribute anyway, _not_ agree to the GPL, and their crime
>is copyright violation, not license violation. You can sue them for
>copyright infringement, but you can't sue them to make them comply with the
>license, since they haven't accepted the license.
There is a more interesting consequence of distributing without permission
that I have managed to overlook all the times I've studied copyright, and
most other people on the net have overlooked it as well.
Suppose party A releases a work. Party B makes a derivative work based on
A. If B has A's permission, B can copyright the derivative work, and all is
well. But what happens if B does *not* have A's permission to make the
derivative work? Answer: the derivative work cannot be copyrighted! A's
part of the derivative work remains under A's copyright, of course, but B's
part becomes public domain. Third parties who want to use the derivative work
only need A's permission, not B's.
--Tim Smith
Aaron M. Renn
>Well. One of those people would essentially have to take him to court
>to force him to do that, if he didn't want to.
Has anyone simply written a polite letter to the author of this code,
explained that readline is under the GPL, and asked him to change his
license terms?
--
Aaron M. Renn (ar...@urbanophile.com) http://www.urbanophile.com/arenn/
Robert Brady
>I too ANAL but if the author in question is distributing
>his program _dynamically_ linked against libreadline then
>you would have to convince the court that dynamic linking
>creates a derived work. I am not sure that you would be
>successful in this.
He's distributing statically linked binaries as well, from
what I could make out, so the point is moot.
--
Robert.
Morten Kjeldgaard
> Has anyone simply written a polite letter to the author of this code,
> explained that readline is under the GPL, and asked him to change his
> license terms?
Yes, someone just wrote me.
I am sorry this problem has arisen, I wasn't paying attention to the
legalities
involved.
I wanted to use the readline library, to give the linux version of the
program
an attractive feature, compared to the versions running on other
platforms.
I have immediately removed the statically linked binary from my site.
It was there because of Debian and (especially) Slackware users,
who often do not have the right version of the libc library. I guess
RedHat will be happy, then...
I am assuming it is alright to distribute the dynamically linked binary.
If not, I will of course immediately stop the distribution, and remove
the calls to readline library from the program.
My apologies for not being aware of this...
/Morten
---
Morten Kjeldgaard <m...@imsb.au.dk> | Phone : +45 89 42 50
26
Institute of Molecular and Structural Biology | Fax : +45 86 12 31
78
Aarhus University | Home : +45 86 18 81
80
Gustav Wieds Vej 10 C, DK-8000 Aarhus C, Denmark | icq : 27224900
Ed Avis
>Of course, you can also rerelease your software as GPLed software to
>amend the mistake. You can still charge for it whatever you want (the
>GPL permits this), but every recipient might choose to further
>distribute it around at his will. However, it is quite legal for you
>to refuse any kind of service or help to people that have not paid for
>the software, and you can try to offer additional services (like
>proprietary manuals and licenses) for pay.
Not proprietary licenses. Using a GPLed library means not only that
you must release a version of your program under the GPL, but also
that _all_ versions must be under that licence. You're not allowed to
grant exceptions, even if the terms were more liberal than the GPL.
--
Ed Avis
Robert Brady
>I am assuming it is alright to distribute the dynamically linked binary.
>If not, I will of course immediately stop the distribution, and remove
>the calls to readline library from the program.
It probably isn't allowable : the FSF would certainly claim it
wasn't, anyway.
The above is not legal advice, I am not a lawyer.
--
Robert.
David Kastrup
Sorry, I had my brain in "wrong-word" mode. I mean proprietary
manuals and services and other stuff.
--
David Kastrup Phone: +49-234-700-5570
Email: d...@neuroinformatik.ruhr-uni-bochum.de Fax: +49-234-709-4209
Institut für Neuroinformatik, Universitätsstr. 150, 44780 Bochum, Germany
Paul D. Smith
mk> I am assuming it is alright to distribute the dynamically linked
mk> binary.
No--well, again, we're not lawyers but if you were to ask the FSF
whether it was OK to use their code like that and take their response at
face value without resorting to the courts, their answer would be "no".
Readline, although it's a library, is actually licensed under the GPL,
not the LGPL. The GPL contains no special exceptions for the type of
linking, dynamic or otherwise. If the running app contains the code,
it's a derivative work (this is the FSF's position; it's never been
challenged in court that I know of).
If readline was LGPL'd, like glibc, for example, then dynamic linking
would be OK.
You might check around on the comp.unix.programmer group or somewhere:
it seems to me I've heard of other libraries along the lines of readline
(although significantly less powerful) which have BSD-like licenses.
Ottavio G. Rizzo
> %% Morten Kjeldgaard <m...@imsb.au.dk> writes:
>
> mk> I am assuming it is alright to distribute the dynamically linked
> mk> binary.
>
> No--well, again, we're not lawyers but if you were to ask the FSF
> whether it was OK to use their code like that and take their response at
> face value without resorting to the courts, their answer would be "no".
On the other hand, most people doubt that the FSF argument against
dynamical linking would hold in a court of law.
Of course, IANAL.
--
Ottavio Rizzo IRMAR, Campus de Beaulieu
ri...@maths.univ-rennes1.fr Université de Rennes 1
Tél +33 (0)2 99 28 67 92 35042 RENNES cedex
Fax +33 (0)2 99 28 67 90 FRANCE
Tim Smith
> mk> binary.
>
>No--well, again, we're not lawyers but if you were to ask the FSF
>whether it was OK to use their code like that and take their response at
>face value without resorting to the courts, their answer would be "no".
Unless there are other dynamic libraries with the same interface...then
FSF thinks its OK.
...
>You might check around on the comp.unix.programmer group or somewhere:
>it seems to me I've heard of other libraries along the lines of readline
>(although significantly less powerful) which have BSD-like licenses.
Do any of them have the same interface as GNU readline?
--Tim Smith
Barry Margolin
Ottavio G. Rizzo <ri...@maths.univ-rennes1.fr> wrote:
>psm...@baynetworks.com (Paul D. Smith) writes:
>
>> %% Morten Kjeldgaard <m...@imsb.au.dk> writes:
>>
>> mk> binary.
>>
>> No--well, again, we're not lawyers but if you were to ask the FSF
>> whether it was OK to use their code like that and take their response at
>> face value without resorting to the courts, their answer would be "no".
>
>dynamical linking would hold in a court of law.
>
>Of course, IANAL.
I don't think a concensus has ever been reached on that regard. Some of us
think it would, some think it wouldn't. IANAL, but my suspicion is that a
court would recognize that dynamic and static linking are effectively
equivalent, and wouldn't allow someone to get around the library's license
on such a technicality. However, this argument only holds when there's
only one available library. If there are multiple compatible libraries,
and the end user can decide which one to install on his system, it's no
longer the same as the calling program being derived from a specific
library.
--
Barry Margolin, bar...@bbnplanet.com
GTE Internetworking, Powered by BBN, Burlington, MA
*** DON'T SEND TECHNICAL QUESTIONS DIRECTLY TO ME, post them to newsgroups.
Please DON'T copy followups to me -- I'll assume it wasn't posted to the group.
Paul D. Smith
ogr> psm...@baynetworks.com (Paul D. Smith) writes:
>> %% Morten Kjeldgaard <m...@imsb.au.dk> writes:
mk> I am assuming it is alright to distribute the dynamically linked
mk> binary.
>> No--well, again, we're not lawyers but if you were to ask the FSF
>> whether it was OK to use their code like that and take their
>> response at face value without resorting to the courts, their
>> answer would be "no".
ogr> On the other hand, most people doubt that the FSF argument
ogr> against dynamical linking would hold in a court of law.
"Most" is _way_ overstating the case. Certainly _some_ people doubt
it.
It was exactly that position I had in mind when I wrote the above
paragraph, making it clear that if you care more about what the authors
of the code _want_ than what you can convince a court they techincally
_said_, then the answer is obvious.
I would hope "most" people would want to follow the wishes of the
author, whether or not they managed to translate them 100% correctly
into legalese.
Morten definitely strikes me as a member of this "most people" group.
PS. Without getting into a whole discussion, just to be clear: do you
doubt that use of _any_ dynamically linked library, regardless of
license, would constitute a derived work in the legal sense, or is
there something about the language of the GPL in particular that
gives you pause?
Paul D. Smith
ts> Paul D. Smith <psm...@baynetworks.com> wrote:
mk> I am assuming it is alright to distribute the dynamically linked
mk> binary.
>> No--well, again, we're not lawyers but if you were to ask the FSF
>> whether it was OK to use their code like that and take their
>> response at face value without resorting to the courts, their
>> answer would be "no".
ts> Unless there are other dynamic libraries with the same
ts> interface...then FSF thinks its OK.
Well, sure. Since the code is not given, only binaries, the dynamic
library would have to be _identical_ in order for the user to use
libreadline.so instead of the linked .so.
And of course, in this case the binary _ISN'T_ linked with readline, so
obviously the FSF has absolutely no interest in it at all. Certainly
it's not the FSF's position that you can't use anything but GNU code :).
If source _were_ provided, then again, it would come with a non-GPL'd,
non-readline library and again, why should the FSF care?
Remember that the GPL restricts _distribution_ only, not use, so if the
user installed readline and linked with it that's just fine.
>> You might check around on the comp.unix.programmer group or somewhere:
>> it seems to me I've heard of other libraries along the lines of readline
>> (although significantly less powerful) which have BSD-like licenses.
ts> Do any of them have the same interface as GNU readline?
Not that I recall. They were _very_ minimalist. I see where you're
going, to do the same thing they did with the GNU math library, and it's
definitely been mentioned here as a good goal before... but no one has
done it. It would have to be at least minimally functional in order to
pass muster; you can't just make some stubs that do nothing.
I'm not even sure where I saw these libraries, though; it's probably
been 4 years or so.
Barry Margolin
> doubt that use of _any_ dynamically linked library, regardless of
> license, would constitute a derived work in the legal sense, or is
> there something about the language of the GPL in particular that
> gives you pause?
I think that most people who think that dynamic linking doesn't violate the
GPL believe it's because it doesn't constitute a derived work because
there's no part of the library in the program that uses it. Thus, the GPL,
which derives its force from copyright protection, does not apply.
It's conceivable that some other license on a DLL could overcome this
limitation. The end user is making a derived work in memory when both the
program and library are loaded. The GPL doesn't apply in this case because
this derived work is not being redistributed, and that's the only activity
that the GPL places restrictions on, but some other license might have
additional restrictions. On the other hand, I believe that copying a into
memory for the purpose of using a legally-obtained program has been defined
to be fair use, and I presume this would apply to DLLs.
Barry Margolin
>user installed readline and linked with it that's just fine.
But if the program requires that the user install readline in order to run
it, it's effectively the same as if the program had been linked with it in
the first place. Is it really appropriate that the readline author's
control depends on whether the application author or the end user performs
that copy and link? It certainly doesn't seem to be in the spirit of the
copyright laws.
Paul D. Smith
bm> In article <p5lndz5...@baynetworks.com>,
bm> Paul D. Smith <psm...@baynetworks.com> wrote:
>> Remember that the GPL restricts _distribution_ only, not use, so if
>> the user installed readline and linked with it that's just fine.
bm> But if the program requires that the user install readline in
bm> order to run it, it's effectively the same as if the program had
bm> been linked with it in the first place.
Sure, but remember the context of this part of the thread is the app
comes with some "readline lookalike" which has the same API but is
perhaps less functional in some ways... but that still basically works.
In other words, the app is fully buildable and usable without readline,
but the user can drop in readline and get a much more featureful
version.
Alan McLean
> it seems to me I've heard of other libraries along the lines of readline
> (although significantly less powerful) which have BSD-like licenses.
BSD has libedit (editline(3) et al).
-amcl
Tim Smith
>But if the program requires that the user install readline in order to run
>it, it's effectively the same as if the program had been linked with it in
>the first place. Is it really appropriate that the readline author's
If that was true, one could not write programs for someone else's operating
system without permission, nor could one write plug ins for someone else's
application without permission, nor could one write scripts for someone
else's interpreter without permission.
--Tim Smith
NF Stevens
[snip]
>PS. Without getting into a whole discussion, just to be clear: do you
> doubt that use of _any_ dynamically linked library, regardless of
> license, would constitute a derived work in the legal sense, or is
> there something about the language of the GPL in particular that
> gives you pause?
Copyright law only covers the distribution of copyrighted works, not
their use. The fact that the memory image of a dynamically linked
work is a "derived work" is a red herring because no one is
distributing that memory image.
Norman
Barry Margolin
The difference is that the OS, application, etc. can be presumed to exist
on the user's system already -- they wouldn't be using your plug-in if they
didn't already have the application. But in the case of an application
that uses readline, there's no reason to expect the library to exist there.
If they don't already have it, they'll need to download it specifically to
make the program complete. Hence it's equivalent to including the library
in the application in the first place.
r...@greenend.org.uk
> I think that most people who think that dynamic linking doesn't
> violate the GPL believe it's because it doesn't constitute a derived
> work because there's no part of the library in the program that uses
> it. Thus, the GPL, which derives its force from copyright
> protection, does not apply.
What about header files?
Tim Smith
>The difference is that the OS, application, etc. can be presumed to exist
>on the user's system already -- they wouldn't be using your plug-in if they
>didn't already have the application. But in the case of an application
>that uses readline, there's no reason to expect the library to exist there.
Every Linux system I've checked has readline available as a shared library.
On the several that I've set up, I've never done anything to explicitly
install it, so it is either always installed, or it is installed as part
of some package that is very likely to be installed.
>If they don't already have it, they'll need to download it specifically to
>make the program complete. Hence it's equivalent to including the library
>in the application in the first place.
The end result is the same, but the route to that destination is different,
and it is the route that copyright law is concerned with.
--Tim Smith
Tim Smith
>Copyright law only covers the distribution of copyrighted works, not
>their use.
Nope, copyright law covers much more. In particular, it covers copying
without distribution, and it covers the creation of derivative works.
--Tim Smith
NF Stevens
You don't have to use the header files; in C functions don't have
to be declared before use, the actual names of structure fields
are irrelevant after compilation etc. etc. So you couldn't
necessarily prove that they were used. The link phase doesn't
need to use the actual GPL'd library. Any shared lib with the
relevant symbol will do.
In the case in question (readline) if you only used the readline
function itself "char *readline (const char *prompt)" then you
could provide a very simple implementation using fputs, fgets
and a static buffer in about half a dozen lines.
IIRC EEC law specifically allows reverse engineering in order to
reproduce an interface which would seem to imply that using the
interface is legitimate.
Norman
Barry Margolin
Tim Smith <t...@halcyon.com> wrote:
>Barry Margolin <bar...@bbnplanet.com> wrote:
>>The difference is that the OS, application, etc. can be presumed to exist
>>on the user's system already -- they wouldn't be using your plug-in if they
>>didn't already have the application. But in the case of an application
>>that uses readline, there's no reason to expect the library to exist there.
>
>Every Linux system I've checked has readline available as a shared library.
>On the several that I've set up, I've never done anything to explicitly
>install it, so it is either always installed, or it is installed as part
>of some package that is very likely to be installed.
If readline is indeed a standard part of Linux, then I think that a binary
targetted specifically to Linux would *not* be considered a derived work of
readline, any more than being a derived work of the Linux libc. But a
binary for Solaris *would* be, since readline is not normally found on
Solaris systems.
>>If they don't already have it, they'll need to download it specifically to
>>make the program complete. Hence it's equivalent to including the library
>>in the application in the first place.
>
>The end result is the same, but the route to that destination is different,
>and it is the route that copyright law is concerned with.
I have repeatedly said that I suspect a court would recognize the
equivalence and not allow the distributor to get around the copyright on
this technicality. Why should it make a difference if he distributes A+B
or distributes A with instructions for the user on how to get B and make
A+B from them?
Geoffrey KEATING
> I have repeatedly said that I suspect a court would recognize the
> equivalence and not allow the distributor to get around the copyright on
> this technicality. Why should it make a difference if he distributes A+B
> or distributes A with instructions for the user on how to get B and make
> A+B from them?
I suspect it's the difference between delivering a gun plus
instructions on how to shoot people, compared to delivering a dead
body with a bullet hole... The first is not necessarily illegal (if
the instructions are acted upon there may be a problem); the second
is, at the least, likely to be in violation of local health
regulations and will probably cause some rather pointed questions to
be asked.
--
Geoff Keating <Geoff....@anu.edu.au>
Ottavio G. Rizzo
> ogr> On the other hand, most people doubt that the FSF argument
> ogr> against dynamical linking would hold in a court of law.
>
> "Most" is _way_ overstating the case. Certainly _some_ people doubt
> it.
Sorry: I messed up with the English. I meant "many".
> It was exactly that position I had in mind when I wrote the above
> paragraph, making it clear that if you care more about what the authors
> of the code _want_ than what you can convince a court they techincally
> _said_, then the answer is obvious.
>
> I would hope "most" people would want to follow the wishes of the
> author, whether or not they managed to translate them 100% correctly
> into legalese.
>
> PS. Without getting into a whole discussion, just to be clear: do you
> doubt that use of _any_ dynamically linked library, regardless of
> license, would constitute a derived work in the legal sense, or is
> there something about the language of the GPL in particular that
> gives you pause?
I'm in no way competent to make such an assertion: from a moral point
of view, readline should not be used in a proprietary program. But law
has little to do with moral <grin>.
Ottavio G. Rizzo
> Barry Margolin <bar...@bbnplanet.com> writes:
>
> > I think that most people who think that dynamic linking doesn't
> > violate the GPL believe it's because it doesn't constitute a derived
> > work because there's no part of the library in the program that uses
> > it. Thus, the GPL, which derives its force from copyright
> > protection, does not apply.
>
> What about header files?
>
AFAIK, header files are just needed to interoperate with the library,
so it's always legal to use them. Moreover, UE laws explicitely allow
reverse engineering to the extent required for interoperability, so
it'd be completely legal to recreate them.
David Kastrup
> r...@greenend.org.uk writes:
>
> > Barry Margolin <bar...@bbnplanet.com> writes:
> >
> > > I think that most people who think that dynamic linking doesn't
> > > violate the GPL believe it's because it doesn't constitute a derived
> > > work because there's no part of the library in the program that uses
> > > it. Thus, the GPL, which derives its force from copyright
> > > protection, does not apply.
> >
> > What about header files?
> >
>
> AFAIK, header files are just needed to interoperate with the library,
> so it's always legal to use them. Moreover, UE laws explicitely allow
> reverse engineering to the extent required for interoperability, so
> it'd be completely legal to recreate them.
I severely doubt that UE law explicitly allows reverge engineering to
the extent required for weaseling around a license that explicitly
forbids linking into proprietary programs. Reverse Engineering is in
some contexts allowed for you to make *proper* use of software that
you acquired a license to. Linking it into proprietary software in a
breach of the basic premises of the license is hardly *proper* use.
Ottavio G. Rizzo
> > AFAIK, header files are just needed to interoperate with the library,
> > so it's always legal to use them. Moreover, UE laws explicitely allow
> > reverse engineering to the extent required for interoperability, so
> > it'd be completely legal to recreate them.
>
> I severely doubt that UE law explicitly allows reverge engineering to
> the extent required for weaseling around a license that explicitly
> forbids linking into proprietary programs. Reverse Engineering is in
> some contexts allowed for you to make *proper* use of software that
> you acquired a license to. Linking it into proprietary software in a
> breach of the basic premises of the license is hardly *proper* use.
I'm not claiming that linking is legal: I'm just saying that it'd be
legal to recreate the header files even in the case using the original
one were forbidden.
Barry Margolin
>> equivalence and not allow the distributor to get around the copyright on
>> this technicality. Why should it make a difference if he distributes A+B
>> or distributes A with instructions for the user on how to get B and make
>> A+B from them?
>
>I suspect it's the difference between delivering a gun plus
>instructions on how to shoot people, compared to delivering a dead
>body with a bullet hole...
Bad analogy. The purchaser of the gun could ignore the instructions, and
use it for target shooting instead.
But the program would be useless unless it's linked with readline. The
vendor is just getting the customer to perform a step that he would
normally have done himself, if it were not for the copyright issue.
David Kastrup
> In article <t8hfomy...@discus.anu.edu.au>,
> Geoffrey KEATING <geo...@discus.anu.edu.au> wrote:
> >Barry Margolin <bar...@bbnplanet.com> writes:
> >
> >> I have repeatedly said that I suspect a court would recognize the
> >> equivalence and not allow the distributor to get around the copyright on
> >> this technicality. Why should it make a difference if he distributes A+B
> >> or distributes A with instructions for the user on how to get B and make
> >> A+B from them?
> >
> >I suspect it's the difference between delivering a gun plus
> >instructions on how to shoot people, compared to delivering a dead
> >body with a bullet hole...
>
> Bad analogy. The purchaser of the gun could ignore the instructions, and
> use it for target shooting instead.
>
> But the program would be useless unless it's linked with readline. The
> vendor is just getting the customer to perform a step that he would
> normally have done himself, if it were not for the copyright issue.
So it's like selling an "inherited" house, only that the customer to
his surprise gets instructions how to shoot the rich uncle
unfortunately still in possession of both the house and good health.
Come on, somebody must be able to come up with even more off-the-wall
appalling analogies.
r...@greenend.org.uk
> I'm not claiming that linking is legal: I'm just saying that it'd be
> legal to recreate the header files even in the case using the
> original one were forbidden.
To do this to achieve the desired effect you'd have to recreate the
header files in such a way that they were definitely not a derived
work of the original library. That's IMHO a minimum requirement for
honourable behaviour; I wouldn't care to comment on what a court might
decide.
While that would be easy for function declarations, a complicated
struct or class definition could be a much harder proposition. If the
header file included complicated macros or inline functions, it gets
worse still.
Steve Peltz
Ed Avis <ep...@doc.ic.ac.uk> wrote:
>Not proprietary licenses. Using a GPLed library means not only that
>you must release a version of your program under the GPL, but also
>that _all_ versions must be under that licence. You're not allowed to
>grant exceptions, even if the terms were more liberal than the GPL.
It does NOT have to be under the GPL. If the terms are more liberal than
the GPL, that is perfectly acceptable.
In addition, you can, of course, release a version of your program
that doesn't use a GPLed library, under any license terms you want (by
"your program", I mean a program that contains only code you've written,
or have received permission to use under the license terms you want to
release it under).
I don't see any way that releasing a program that dynamically links to
a library, but that is not distributed with that library, can possibly
infringe on the copyright of the library if it is under the GPL. The user
must already have a copy of it, the GPL does not put any restrictions on
using the library, and there's no reason to think that the user is likely
to re-distribute the program along with the GPLed library. Where's the
infringement? The only theory that might allow infringement would be to
postulate an interface copyright.
The GPL does not have a clause along the lines of "you may not create
a derivative work unless you have the right to release it under the
GPL.", and I think such a clause would be a mistake to add.
Leslie Mikesell
Barry Margolin <bar...@bbnplanet.com> wrote:
>I have repeatedly said that I suspect a court would recognize the
>equivalence and not allow the distributor to get around the copyright on
>this technicality. Why should it make a difference if he distributes A+B
>or distributes A with instructions for the user on how to get B and make
>A+B from them?
I don't think there has ever been any question that an end user
can obtain A and B by himself and create A+B even if it takes some
modifications to B. Or even that he can hire someone to do it for
him. If you are going to talk about equivalence, how is this
changed by someone making the process easier? Note that the user
in question has the right to have and use A and B in their original
forms and the right to make any modifications for himself. Most
equivalent copyright issues would revolve around the right of the
user to receive any copy of the work in question and that isn't
the case here. The issue has something to do with whether or not
the modifications can be distributed at all even though they are
entirely a third party's work. Is a patch necessarily a
derived work under control of the original work's author?
Les Mikesell
l...@mcs.com
JD
> > doubt that use of _any_ dynamically linked library, regardless of
> > license, would constitute a derived work in the legal sense, or
is
> > there something about the language of the GPL in particular that
> > gives you pause?
>
> work is a "derived work" is a red herring because no one is
> distributing that memory image.
IANAL, but if I were a juror, that's exactly how I'd see it. The
argument others have made in this thread that there must be a
work-alike library that can be used instead of the GPL-infected
code is bogus as well. Consider this hypothetical scenerio:
1. I obtain a list of exported symbols provided by libreadline.so,
from a public (non-copyrighted) source.
2. I create my own libreadline.so containing these symbols, and
maybe a header file if data structures are involved.
3. I link my application with my shared library.
4. I distribute the code, minus shared library, with instructions
describing how to obtain and install GNU's version of
libreadline.so.
[You may ask how I debugged this program without a working libreadline;
maybe I didn't, it just works because I'm a careful programmer. Maybe
I used predicate logic or other well-known types of program correctness
proofs. Maybe my users beta test it for me. Suffice it to say that
I've never actually exercised that portion of the code myself.]
I defy anyone to tell me how my program can be claimed as a "derivative
work" of GNU readline. I have never seen or touched a copy of it, nor
has my program; therefore I have not even implicitly accepted its
redistribution license terms. The fact that it is necessary for correct
operation of the program is irrelevant; the end user has full rights
under the GPL to cause the library to be used by my program; he is not
redistributing it.
Sent via Deja.com http://www.deja.com/
Share what you know. Learn what you don't.
Decklin Foster
> Bad analogy. The purchaser of the gun could ignore the instructions, and
> use it for target shooting instead.
>
> But the program would be useless unless it's linked with readline. The
> vendor is just getting the customer to perform a step that he would
> normally have done himself, if it were not for the copyright issue.
But what if there exist the following (assume they are are all legal):
1. readline
2. another library that does the same thing, proprietary
3. another one, with X/MIT-style licensing
4. something under the QPL (just kidding! ;-)
Now it's more like a gun, isn't it? FWIW, i'm compiling rc right now,
and it allows you to do either ./configure --with-readline or
./configure --with-vrl. Probably more; but then i'm using readline
(heathen that I am.)
Ah, but you say, "forget the other choices, it still has the
readline-derivative code in it." (not sure if I accept this reasoning,
but i'll go with it.) So say I, snappy programmer that I am, write a
library somewhat Imlib that abstracts line-reading functions, handing
it off to 1, 2, 3 (or 4 <ahem>) underneath a unified API. (ugh, I
can't believe I just said "unified API"...) I put it out under the
LGPL, and this hypothetical proprietary program only writes code that
works with my interface.
I suppose I add a disclaimer to the LGPL that says, "permission to
link with this code notwithstanding, i take no responsibility for you
doing something illegal with it."
OK, you say, whatever, that just stretches out the same basic
problem... the propietary thing is a derivative of Linelib is a
derivative of readline.
Now we take this to it's logical conclusion... I write said library
first, and only include backends for libraries (just assume that many
exist, for rhetoric purposes) that have X/MIT or BSD-sans-ads
licenses, or are in the public domain. Then so-and-so writes this
propietary program that links against my library.
*Then*, someone else wants to use my library for a GPL project, but
finds the lack of a readline backend problematic, so he writes one. He
distributes such a patch widely.
So one day, Joe linux user hops onto #enlightenment and says, "hey
dudes, how do I get BitchX to to filename tab completion through
symlinked directories?" and <LiNUx-GuRu> says "go download this patch
and compile the latest Linelib from CVS. so you can use readline (you
have to fetch the RPM for that from ...)"
So the newbie user compiles this patched version and -- several weeks
later -- comes upon a corporate web site with this hypothetical
proprietary program. "Ooooo, kewl", he swoons, and downloads the RPM.
he then promptly runs it, and it finds the .so for Linelib which
proceeds to use readline to do something or other.
So the question is... who's responsible here?
;-)
[1] this is a bit of an rc in-joke...
[2] sorry, i couldn't contain myself... <stifles giggle>
--
"Men argue; nature acts." -- Voltaire
David Kastrup
> In article <3757B278...@doc.ic.ac.uk>,
> Ed Avis <ep...@doc.ic.ac.uk> wrote:
> >Not proprietary licenses. Using a GPLed library means not only that
> >you must release a version of your program under the GPL, but also
> >that _all_ versions must be under that licence. You're not allowed to
> >grant exceptions, even if the terms were more liberal than the GPL.
>
> It does NOT have to be under the GPL. If the terms are more liberal than
> the GPL, that is perfectly acceptable.
>
> In addition, you can, of course, release a version of your program
> that doesn't use a GPLed library, under any license terms you want (by
> "your program", I mean a program that contains only code you've written,
> or have received permission to use under the license terms you want to
> release it under).
>
> I don't see any way that releasing a program that dynamically links to
> a library, but that is not distributed with that library, can possibly
> infringe on the copyright of the library if it is under the GPL.
The problem is that "derived work" is something that is defined by the
law, not by the GPL. A program that is lacking serious functionality
without being compiled using the readline library *is* most probably a
derived work of the readline library according to the mad copyright
laws.
If that disturbs you, fight the laws. The GPL is not intended to be
weaker in its purpose than commercial licenses. In the FSF's ideal
world, no term of the GPL would be enforceable. But as long as the
law makes its mads definitions, they do not see to arbitrarily give up
the protection from proprietarization that the proprietary products
use as bulwark against the freeing of software.
> The user must already have a copy of it, the GPL does not put any
> restrictions on using the library, and there's no reason to think
> that the user is likely to re-distribute the program along with the
> GPLed library. Where's the infringement? The only theory that might
> allow infringement would be to postulate an interface copyright.
According to the law, you may not derive from software without being
granted permission to do so. The GPL grants this permission only
where the derived work is covered under the GPL again.
> The GPL does not have a clause along the lines of "you may not
> create a derivative work unless you have the right to release it
> under the GPL.", and I think such a clause would be a mistake to
> add.
Right at the start, the GPL states:
0. This License applies to any program or other work which contains
a notice placed by the copyright holder saying it may be distributed
under the terms of this General Public License. The "Program", below,
refers to any such program or work, and a "work based on the Program"
means either the Program or any derivative work under copyright law:
So it's the copyright law that is at fault.
Peter Seebach
Barry Margolin <bar...@bbnplanet.com> wrote:
>I've already addressed this case numerous times. The FSF has always
>maintained that it's only a derivative of the GPLed library if there's no
>alternative to using that library, and I agree with them. When the
>end-user has a choice of libraries to link in, the programmer cannot be
>considered to be deriving from any particular one of them. This was how
>the Gmp issue was solved several years ago.
I don't buy this, because the legality of writing something shouldn't depend
on external indeterminacy.
-s
--
Copyright 1999, All rights reserved. Peter Seebach / se...@plethora.net
C/Unix wizard, Pro-commerce radical, Spam fighter. Boycott Spamazon!
Will work for interesting hardware. http://www.plethora.net/~seebs/
Visit my new ISP <URL:http://www.plethora.net/> --- More Net, Less Spam!
Barry Margolin
Decklin Foster <dec...@home.com> wrote:
>Barry Margolin <bar...@bbnplanet.com> writes:
>
>> Bad analogy. The purchaser of the gun could ignore the instructions, and
>> use it for target shooting instead.
>>
>> But the program would be useless unless it's linked with readline. The
>> vendor is just getting the customer to perform a step that he would
>> normally have done himself, if it were not for the copyright issue.
>
>But what if there exist the following (assume they are are all legal):
>
>1. readline
>2. another library that does the same thing, proprietary
>3. another one, with X/MIT-style licensing
>4. something under the QPL (just kidding! ;-)
I've already addressed this case numerous times. The FSF has always
maintained that it's only a derivative of the GPLed library if there's no
alternative to using that library, and I agree with them. When the
end-user has a choice of libraries to link in, the programmer cannot be
considered to be deriving from any particular one of them. This was how
the Gmp issue was solved several years ago.
--
Barry Margolin
>In article <9zk63.375$KM3.122277@burlma1-snr2>,
>Barry Margolin <bar...@bbnplanet.com> wrote:
>
>>I have repeatedly said that I suspect a court would recognize the
>>equivalence and not allow the distributor to get around the copyright on
>>this technicality. Why should it make a difference if he distributes A+B
>>or distributes A with instructions for the user on how to get B and make
>>A+B from them?
>
>I don't think there has ever been any question that an end user
>can obtain A and B by himself and create A+B even if it takes some
>modifications to B. Or even that he can hire someone to do it for
>him. If you are going to talk about equivalence, how is this
>changed by someone making the process easier?
It has nothing to do with the end user having the right, or making the
process easier. I'm talking about the vendor *requiring* the end user to
perform this step. The only working program is A+B, but the vendor only
distributes A because he knows that A+B would be a derivative work of B and
he doesn't want to abide by B's copyright terms. So instead, he ships A
and requires the customer to add B to it himself.
Why should the owner of B be powerless to stop this when he could stop the
vendor from shipping A+B as a whole? This means that vendors could
routinely get around restrictions on distributing derivative works by
pushing the derivation step off to the end user.
Barry Margolin
Peter Seebach <se...@plethora.net> wrote:
>In article <sEc73.439$KM3.138805@burlma1-snr2>,
>Barry Margolin <bar...@bbnplanet.com> wrote:
>>I've already addressed this case numerous times. The FSF has always
>>maintained that it's only a derivative of the GPLed library if there's no
>>alternative to using that library, and I agree with them. When the
>>end-user has a choice of libraries to link in, the programmer cannot be
>>considered to be deriving from any particular one of them. This was how
>>the Gmp issue was solved several years ago.
>
>on external indeterminacy.
The person writing the program knows full well whether it requires a
one-of-a-kind library or something generic. When the installation
instructions include specific instructions to download and link with a
particular library, he knows that he's just trying to work around the fact
that he couldn't link it himself. There's nothing indeterminate about it.
On the other hand, if the instructions say "Install the readline
implementation you like and link with it -- here are locations of some
popular implementation" then it's clear that he's not deriving from any
specific one of them. He's writing a program that uses an API, not a
particular library.
About the best someone could do is profess ignorance. He could say he
thought there were multiple readline implementations, but he only happened
to know the download instructions for GNU readline. If it came to it, it
would be up to the court to decide whether he was telling the truth.
Furthermore, once he became aware, through this process, that there was
just one readline, he would have to abide by its license or stop deriving
from it.
Ken Arromdee
>It has nothing to do with the end user having the right, or making the
>process easier. I'm talking about the vendor *requiring* the end user to
>perform this step. The only working program is A+B, but the vendor only
>distributes A because he knows that A+B would be a derivative work of B and
>he doesn't want to abide by B's copyright terms. So instead, he ships A
>and requires the customer to add B to it himself.
>
>Why should the owner of B be powerless to stop this when he could stop the
>vendor from shipping A+B as a whole?
I have a book (Windows '95 Annoyances) which describes a binary patch to
modify Windows to alter the start menu. Distributing the entire derivative
work (Windows with the patch applied) is not legal. Distributing the patch
alone is, even though the patch is of no use unless applied to Windows.
>This means that vendors could
>routinely get around restrictions on distributing derivative works by
>pushing the derivation step off to the end user.
There is no such thing as a "restriction on distributing derivative works";
there's a restriction on making copies.
For most software it is illegal both to copy for someone else and to copy from
someone else. You can't give, and you can't receive. But if you don't follow
the GNU license, you are only restricted from giving, not from receiving.
The vendor's ability to get around the restriction depends on this asymmetry.
The asymmetry exists only for GPL and GPL-like licenses, so the vendor could
_not_ do this routinely just because he could do it for GPL software.
The GPL _could_ have had a clause requiring that people agree to the license
before they receive and use a copy. If someone got a copy of GNU software off
of a friend without agreeing to the clause, he would not be allowed to use it
at all, just as he would not be allowed to use a copy of Windows that he got
from a friend. But the GPL isn't written that way.
--
Ken Arromdee |They said it was *daft* to build a space
arro...@inetnow.net |station in a swamp, but I showed them! It
karr...@nyx.nyx.net |sank into the swamp. So I built a second
http://www.inetnow.net/~arromdee|space station. That sank into the swamp too.
--------------------------------+My third space station sank into the swamp.
So I built a fourth one. That fell into a time warp and _then_ sank into the
swamp. But the fifth one... stayed up! --Monty Python/Babylon 5
Steve Peltz
Barry Margolin <bar...@bbnplanet.com> wrote:
>The person writing the program knows full well whether it requires a
>one-of-a-kind library or something generic. When the installation
>instructions include specific instructions to download and link with a
>particular library, he knows that he's just trying to work around the fact
>that he couldn't link it himself. There's nothing indeterminate about it.
Here's my installation instructions: go out and buy a copy of
"Merriam-Webster's Collegiate Dictionary Tenth Edition Copyright 1993";
on page 679, insert after the 3rd entry on the page (between lintwhite
and lion) the following: "Linux: an open source implementation of a
Unix-like operating system".
Is that a copyright violation? If not, why not? I'm just trying to work
around the fact that I can't make the change myself and distribute the
result...
If it is a copyright violation, I suppose it must be because it is the
equivalent of my printing up copies of the dictionary, making the change,
and distributing them. Right?
A different example: when Photoshop was first released, before there
were any other programs that took Photoshop plug-ins, would it have been
a violation of Photoshop's copyright to write a plug-in? Would it have
been a violation to write a program that used a plug-in?
Steve Peltz
Barry Margolin <bar...@bbnplanet.com> wrote:
>Why should the owner of B be powerless to stop this when he could stop the
>vendor from shipping A+B as a whole?
Because the owner of B can only control the copying of B. If the vendor
ships A+B, then he's violating B's copyright. If the vendor only ships A,
then he is not. There might be some aspect of contributory infringement
IF the customer was prohibited from combining A and B, or IF the customer
was encouraged to re-distribute the combination (but the former isn't
the case if B is under the GPL, and the latter isn't likely if A is
distributed under a restrictive license).
Barry Margolin
I was referring to the spirit of the law. Does it make sense for a law to
prohibit someone from shipping A+B when it allows shipping A along with
instructions saying "to make this work, you have to add B to it"?
Peter Seebach
Barry Margolin <bar...@bbnplanet.com> wrote:
>The person writing the program knows full well whether it requires a
>one-of-a-kind library or something generic. When the installation
>instructions include specific instructions to download and link with a
>particular library, he knows that he's just trying to work around the fact
>that he couldn't link it himself. There's nothing indeterminate about it.
In a particular case, yes. But I don't buy the general argument. Is it
legal to distribute something that you believe links with a public domain
library if you turn out to be wrong?
Honestly, I think this is a Microsoft/Apple concept of "derived work".
Intended-to-link-with is at the level of "look and feel".
Isaac
>
>I was referring to the spirit of the law. Does it make sense for a law to
>prohibit someone from shipping A+B when it allows shipping A along with
>instructions saying "to make this work, you have to add B to it"?
>
It wouldn't make logical sense for the law to say this, but the
problem is created by the GPL rather than by copyright law. Users
are not prohibited from downloading GPL'd code for their own use.
The GPL explicitly rejects such restrictions, while reserving other
limits as the law allows for copyright holders. This is the cause of
the non sensical conclusions and not the law. There is no legal reason
why the GPL couldn't have been written to prevent this, but it wasn't.
Isaac
Isaac
>
>The problem is that "derived work" is something that is defined by the
>law, not by the GPL. A program that is lacking serious functionality
>without being compiled using the readline library *is* most probably a
>derived work of the readline library according to the mad copyright
>laws.
>
a programs functionality would be intact with even a marginally
effective substitute. I think the rest of your argument is mooted
since you didn't establish your initial premise.
Isaac
Tim Smith
>Why should the owner of B be powerless to stop this when he could stop the
>vendor from shipping A+B as a whole?
The owner of B isn't powerless to stop it. He/she just has to put in the
license of B something that says the end user can't combine B with A, and
make sure that this license term is enforceable.
It was B's owner's choice to use a license like GPL that gives the end user
the freedom to combine B with pretty much anything the end user wants to, as
long as the end user does not distribute.
--Tim Smith
Tim Smith
>I was referring to the spirit of the law. Does it make sense for a law to
>prohibit someone from shipping A+B when it allows shipping A along with
>instructions saying "to make this work, you have to add B to it"?
Yes, because in the second case, the owner of B has control--when user U
asks the owner of B for a copy of B, the owner can say "Are you going to
use it with A?", and if U says "yes", the owner of B can say "You can't
have a copy of B". In the first case, the owner of B is being bypassed.
It's hard to imagine what B's owner could complain about: "Your Honor, I
allege that the vendor of A urged people to obtain lawful copies of my
program and use them in accordance with my licensing terms!". :-)
--Tim Smith
Lynn Winebarger
Barry Margolin <bar...@bbnplanet.com> wrote:
>In article <7jjud7$iso$1...@harbinger.nn.com>,
>Steve Peltz <pe...@medusa.nn.com> wrote:
>>Because the owner of B can only control the copying of B. If the vendor
>>ships A+B, then he's violating B's copyright. If the vendor only ships A,
>>then he is not. There might be some aspect of contributory infringement
>>IF the customer was prohibited from combining A and B, or IF the customer
>>was encouraged to re-distribute the combination (but the former isn't
>>the case if B is under the GPL, and the latter isn't likely if A is
>>distributed under a restrictive license).
>
>prohibit someone from shipping A+B when it allows shipping A along with
>instructions saying "to make this work, you have to add B to it"?
>
of the original work, in the second it's about restricting the user's
ability to _use_ their copy as they see fit.
One of the questions that have to be weighed in a copyright
infringement case (read the statute) is if the alleged infringement
harms the market. Arguably, by B's author requiring the user to acquire
A from A's author, B is actually _improving_ the market for A. Whereas
he would clearly be harming the market for A if he just included A
wholesale in his distribution.
Lynn
Lynn Winebarger
Isaac <irc...@latveria.castledoom.org> wrote:
>It wouldn't make logical sense for the law to say this, but the
>problem is created by the GPL rather than by copyright law. Users
>are not prohibited from downloading GPL'd code for their own use.
>The GPL explicitly rejects such restrictions, while reserving other
>limits as the law allows for copyright holders. This is the cause of
>the non sensical conclusions and not the law. There is no legal reason
>why the GPL couldn't have been written to prevent this, but it wasn't.
How could it do so without becoming an "End User License Agreement"
parading as a contract with the user concerning using the software?
The GPL is a license on the copyright, not the copy (except the part
about "NO WARRANTY"). You're not bound by a contract if you simply use
the software - only if you doing things protected by copyright. i.e.
copying, modifying, redistributing to others.
Lynn
Stefaan A Eeckels
se...@plethora.net (Peter Seebach) writes:
> In article <9wd73.446$KM3.138805@burlma1-snr2>,
> Barry Margolin <bar...@bbnplanet.com> wrote:
>>The person writing the program knows full well whether it requires a
>>one-of-a-kind library or something generic. When the installation
>>instructions include specific instructions to download and link with a
>>particular library, he knows that he's just trying to work around the fact
>>that he couldn't link it himself. There's nothing indeterminate about it.
>
> In a particular case, yes. But I don't buy the general argument. Is it
> legal to distribute something that you believe links with a public domain
> library if you turn out to be wrong?
>
> Honestly, I think this is a Microsoft/Apple concept of "derived work".
> Intended-to-link-with is at the level of "look and feel".
By that reasoning, each and every Windows program is a "derived
work" from Windows (as it is intended to link with the Windows
DLLs). If a book contains a reference to another book, and the
latter is vital to the comprehension of the former, does that
make that a derivative work? Does it cease to be a derivative
when you (or someone else) publish a brief overview of the
matter covered in the referenced book?
I can't see how the appearance of "readline()" once in a program
can ever make that program a "derived work".
--
Stefaan
--
PGP key available from PGP key servers (http://www.pgp.net/pgpnet/)
___________________________________________________________________
Perfection is reached, not when there is no longer anything to add,
but when there is no longer anything to take away. -- Saint-Exupéry
Fergus Henderson
>I was referring to the spirit of the law. Does it make sense for a law to
>prohibit someone from shipping A+B when it allows shipping A along with
>instructions saying "to make this work, you have to add B to it"?
Sure. Take the case where B is a restricted substance of some kind,
which only properly licensed persons are allowed to possess or distribute,
because it could be (ab)used for some anti-social purpose.
And suppose A is a device which performs some _benign_ function using B.
Shipping A+B or B alone might well be illegal because of possibility of
people abusing B for antisocial purposes, but there's no reason to make
shipping A alone illegal.
For example, suppose A is a CAT-scanner and B is the radioactive isotope
that you need to inject into the patient before scanning them.
For another example, suppose that A is a program for deleting X-rated
images or sounds from DVD movies and suppose that B is an X-rated DVD movie.
--
Fergus Henderson <f...@cs.mu.oz.au> | "I have always known that the pursuit
WWW: <http://www.cs.mu.oz.au/~fjh> | of excellence is a lethal habit"
PGP: finger f...@128.250.37.3 | -- the last words of T. S. Garp.
Ken Arromdee
>>The GPL explicitly rejects such restrictions, while reserving other
>>limits as the law allows for copyright holders. This is the cause of
>>the non sensical conclusions and not the law. There is no legal reason
>>why the GPL couldn't have been written to prevent this, but it wasn't.
> How could it do so without becoming an "End User License Agreement"
>parading as a contract with the user concerning using the software?
> The GPL is a license on the copyright, not the copy (except the part
>about "NO WARRANTY"). You're not bound by a contract if you simply use
>the software - only if you doing things protected by copyright. i.e.
In order to use the software, you have to get it. Getting it, which involves
making a copy, _is_ protected by copyright. If a friend made you a copy of
Microsoft Windows, you would not have any right to use that copy at all.
Thus, a modified GPL which let you use the copy under certain conditions would
be adding rights to the default of no rights at all.
That works just like the current GPL. You don't have to agree to the contract,
but only the contract gives you any rights at all. It would be a copyright
violation to use the software without agreeing, just like it would _always_
be a copyright violation to use that copy of Windows that your friend made.
(Though there would be one quirk: if you got a copy directly from the FSF,
and they made the copy for you, this modified license wouldn't apply.)
Kenneth P. Turvey
[Snip]
>
>I've already addressed this case numerous times. The FSF has always
>maintained that it's only a derivative of the GPLed library if there's no
>alternative to using that library, and I agree with them. When the
This whole issue does seem to be fraught with problems. Is a version #3
of the GPL in the works? It seems that modifying the GPL to handle this
very important case well is enough to justify a revision.
--
Kenneth P. Turvey <ktu...@SprocketShop.com>
----------------- http://www.tranquility.net/~kturvey
Ninety percent of the politicians give the other ten percent a bad name.
-- Henry Kissinger
Leslie Mikesell
Barry Margolin <bar...@bbnplanet.com> wrote:
>>>I have repeatedly said that I suspect a court would recognize the
>>>equivalence and not allow the distributor to get around the copyright on
>>>this technicality. Why should it make a difference if he distributes A+B
>>>or distributes A with instructions for the user on how to get B and make
>>>A+B from them?
>>
>>I don't think there has ever been any question that an end user
>>can obtain A and B by himself and create A+B even if it takes some
>>modifications to B. Or even that he can hire someone to do it for
>>him. If you are going to talk about equivalence, how is this
>>changed by someone making the process easier?
>
>It has nothing to do with the end user having the right, or making the
>process easier.
Yes it does - you have made the case too simple by describing the
only other entity as A. That isn't the real problem.
>I'm talking about the vendor *requiring* the end user to
>perform this step. The only working program is A+B, but the vendor only
>distributes A because he knows that A+B would be a derivative work of B and
>he doesn't want to abide by B's copyright terms. So instead, he ships A
>and requires the customer to add B to it himself.
No, let's say A is an existing standalone library available under
some non-GPL terms. A user of A also has GPLed work B, creates
some modifications (C) to combine them. He knows lots of other people
already have A and can get B and would find his modification useful.
Why should he be prevented from distributing C?
>Why should the owner of B be powerless to stop this when he could stop the
>vendor from shipping A+B as a whole?
Why should the distribution of C be prevented when it contains none
of A or B and is entirely another person's work? Why should the
owner of A have any control whatever over this? Now the author
of C may or may not have some relationship with library A. That
becomes a separate issue.
>This means that vendors could
>routinely get around restrictions on distributing derivative works by
>pushing the derivation step off to the end user.
The end user is always allowed to create the derivitaves - or hire
someone to do it for him. The issue is restricting the distribution
of modifications. For the concept to make any sense you would have
to claim that the modifications where copyright-protected derivatives
of work B. How can that be when the modification is the difference
from the original, not a copy of it? Would it make a difference if
the only modifications neccessary could be the addition of a library
that intercepted existing function calls? Would it be necessary to
use a 'clean-room' approach to avoid the possibility of copyright
claims? (And does this mean that anyone who has seen GPL'd code
might be subject to copyright infringment claims if they ever write
anything similar for distribution under different terms?).
Les Mikesell
l...@mcs.com
NF Stevens
[snip]
>I was referring to the spirit of the law. Does it make sense for a law to
>prohibit someone from shipping A+B when it allows shipping A along with
>instructions saying "to make this work, you have to add B to it"?
It makes sense when the copy of B cannot be obtained for free.
This is the situation envisioned by those drafting the
copyright laws.
Norman
Lynn Winebarger
Ken Arromdee <karr...@nyx.nyx.net> wrote:
>In order to use the software, you have to get it. Getting it, which involves
>making a copy, _is_ protected by copyright. If a friend made you a copy of
>Microsoft Windows, you would not have any right to use that copy at all.
>Thus, a modified GPL which let you use the copy under certain conditions would
>be adding rights to the default of no rights at all.
Owning the copy would give me rights to it. While the distributor
might well be liable for not following the terms of the GPL in making a
copy, once I own the copy, it's my property. Any rights the copyright
holder strictly end at the act of making copies.
>That works just like the current GPL. You don't have to agree to the contract,
>but only the contract gives you any rights at all. It would be a copyright
No, ownership of the copy gives me rights to use that copy, read it,
sell it to someone else, etc. I don't have a right to make copies and
sell them (though I can make copies to the extent that it falls under
fair use, that would probably preclude selling them).
>violation to use the software without agreeing, just like it would _always_
>be a copyright violation to use that copy of Windows that your friend made.
>
I would like to see a lawyer address this issue. If I own a book
that was published by a "pirate", does that mean that I am a copyright
infringer or that the author holds a property interest in the copy of
the book?
By the way, _using_ a literary work is not a copyright violation,
unless it's for a public performance or the like.
Lynn
David Kastrup
> In article <92894432...@iris.nyx.net>,
> Ken Arromdee <karr...@nyx.nyx.net> wrote:
> >In order to use the software, you have to get it. Getting it, which involves
> >making a copy, _is_ protected by copyright. If a friend made you a copy of
> >Microsoft Windows, you would not have any right to use that copy at all.
> >Thus, a modified GPL which let you use the copy under certain conditions would
> >be adding rights to the default of no rights at all.
>
> Owning the copy would give me rights to it. While the distributor
> might well be liable for not following the terms of the GPL in making a
> copy, once I own the copy, it's my property.
I recommend that you reread the law. Acquiring ownership from
somebody not entitled to transfer it to you is not possible. You
cannot buy ownership of stolen goods.
If you get a copy from the copyright holder, you can do what owning a
copy allows you. If you get an illegal copy from some distributor,
this copy does not become your property.
> I would like to see a lawyer address this issue. If I own a
> book that was published by a "pirate", does that mean that I am a
> copyright infringer or that the author holds a property interest in
> the copy of the book?
You cannot own a book published by a pirate. Its contents belong to
the copyright holder that did not give you any usage rights. This
book can be confiscated without reimbursement. Any use of it for any
purpose (except heating, which does not infringe on the copyright) is
illegal.
> By the way, _using_ a literary work is not a copyright violation,
> unless it's for a public performance or the like.
It is if you do not own the copy. If you have stolen it at a library,
reading it for your amusement is not legal. If you have bought it at
a book pirate, reading it is not legal.
Lynn Winebarger
David Kastrup <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
>owin...@ezinfo.ucs.indiana.edu (Lynn Winebarger) writes:
>> Owning the copy would give me rights to it. While the distributor
>> might well be liable for not following the terms of the GPL in making a
>> copy, once I own the copy, it's my property.
>
>I recommend that you reread the law. Acquiring ownership from
>somebody not entitled to transfer it to you is not possible. You
>cannot buy ownership of stolen goods.
The person is certainly entitled to transfer the physical medium -
the copyright owner does not automatically gain a property interest in
all copies that are made - ownership of a copyright and a copy are two
completely separate things (see section 202 of Title 17).
One of the remedies for infringement _is_ confiscation and
destruction of all infringing copies that were made, but that is not the
same as saying that the copyright owner owned the copies and the
possessors of the infringing copies did not.
>If you get a copy from the copyright holder, you can do what owning a
>copy allows you. If you get an illegal copy from some distributor,
>this copy does not become your property.
>
Again, I would like to see a lawyer address these issues. It's clear
that the person doing the copying (or perhaps stealing, if the document
is unpublished) is going to suffer major legal problems, but it's not
clear the purchaser would - particularly if the purchase were in good
faith.
While it's true that if you acquire a stolen car from a car thief,
the property does not transfer to you, that's because someone else
already owns it. In the case of a copy of a book, the copyright holder
does not automatically hold a property interest in the book (again see
section 202). For example, if I make a copy of an academic paper for
use in a class I'm teaching, that's a fair use. The copyright owner
does not own the copies, the students who I give them to do.
>It is if you do not own the copy. If you have stolen it at a library,
>reading it for your amusement is not legal. If you have bought it at
>a book pirate, reading it is not legal.
>
I have major doubts about this - do you have references to statute
or case law to prove this assertion? I'm willing to accept this may be
the case where the work is unpublished, but it seems to me showing this
claim applies to published work is a whole 'nother kettle of fish.
Lynn
Ken Arromdee
>>In order to use the software, you have to get it. Getting it, which involves
>>making a copy, _is_ protected by copyright. If a friend made you a copy of
>>Microsoft Windows, you would not have any right to use that copy at all.
>>Thus, a modified GPL which let you use the copy under certain conditions would
>>be adding rights to the default of no rights at all.
>might well be liable for not following the terms of the GPL in making a
>copy, once I own the copy, it's my property.
You seem to be saying that if the distributor makes you a copy illegally, the
distributor can be sued, but you own the copy.
This isn't true. Again, consider what happens if it is Windows 95 instead of
GPL software. Would you claim that the distributor can be sued by Microsoft,
but you can use the copy of Windows as you see fit?
Of course not. If the copy is illegal, the person receiving the copy doesn't
end up owning it. They only own the copy if it was legally made.
So a modified GPL could state conditions under which the recipient is allowed
to have a copy. If the recipient doesn't agree to the conditions, then he
cannot legally get a copy at all, just like under the present GPL someone who
doesn't agree to the conditions can't distribute a copy.
ma...@usai.asiainfo.com
: I've already addressed this case numerous times. The FSF has always
: maintained that it's only a derivative of the GPLed library if there's no
: alternative to using that library, and I agree with them. When the
: considered to be deriving from any particular one of them. This was how
: the Gmp issue was solved several years ago.
Would any really feeble readline function work? This seems to be
compatible at the API level.
char *readline(char *prompt)
{
puts(prompt);
fflush(stdout);
return gets(malloc(1000));
}
--
Mark Harrison "Open the floppy disk door, Hal."
AsiaInfo Computer Networks http://usai.asiainfo.com:8080/
Beijing, China / Santa Clara, CA ma...@usai.asiainfo.com
Lynn Winebarger
Ken Arromdee <karr...@nyx.nyx.net> wrote:
>You seem to be saying that if the distributor makes you a copy illegally, the
>distributor can be sued, but you own the copy.
I'm saying a lawyer should address these issues. What is clear is
that the copyright holder does not own the copy, and that the court may
order the copy's confiscation and destruction. That is _not_ the same
as the possessor of the copy not owning it.
For example, marijuana is illegal in (at least most of) the United
States. That does not mean that I can not own some, only that it is
illegal to possess it in the US. And that the gov't can confiscate and
destroy it.
>This isn't true. Again, consider what happens if it is Windows 95 instead of
>GPL software. Would you claim that the distributor can be sued by Microsoft,
>but you can use the copy of Windows as you see fit?
>
This is probably not the case - for example, if I was a company
making Bargain PC's and distributing copies of Windows98 that I had
obtained from a second source, I'd guess the courts would take a dim
view of that. But that's not the same as the end user not owning the copy.
>Of course not. If the copy is illegal, the person receiving the copy doesn't
>end up owning it. They only own the copy if it was legally made.
See above.
>
>So a modified GPL could state conditions under which the recipient is allowed
>to have a copy. If the recipient doesn't agree to the conditions, then he
>cannot legally get a copy at all, just like under the present GPL someone who
>doesn't agree to the conditions can't distribute a copy.
I hope you just don't realize the implications of what you've
suggested. I hate to think of what would happen to our society if
literary works that were widely published had this kind of restriction
on who could read it.
Lynn
David Kastrup
> In article <92899943...@iris.nyx.net>,
> Ken Arromdee <karr...@nyx.nyx.net> wrote:
> >You seem to be saying that if the distributor makes you a copy illegally, the
> >distributor can be sued, but you own the copy.
>
> I'm saying a lawyer should address these issues. What is clear
> is that the copyright holder does not own the copy,
But the copyright holder owns the rights to his content on the copy,
if not on the physical medium itself. Since these contents have no
business being on the copy, he can demand them to be removed, which in
many cases can be accomplished only by destruction of the copy. He
can also ask for punishment of the persons reading the book unless
they have done it in good faith, in which he is only entitled to
destruction of the copy.
If you buy a pirated book, you have acquired the right to burn the
book in your chimney, or to wipe your behind with pages from it, or to
roll cigarettes with its pages and smoke them. You have not acquired
the right to read it.
> court may order the copy's confiscation and destruction. That is
> _not_ the same as the possessor of the copy not owning it.
You own the physical medium, but you own none of the rights to use the
copyrighted material unless regularly obtained with consent of the
copyright holder.
> For example, marijuana is illegal in (at least most of) the
> United States. That does not mean that I can not own some, only
> that it is illegal to possess it in the US. And that the gov't can
> confiscate and destroy it.
You are confusing two things here. I can buy some marijuana, in which
case I am the owner of it: nobody else can make any claims to me to
return the marijuana. I can, however, also steal marijuana in which
case I have not acquired the ownership of it.
There are different crimes involved if I buy, say, Cocaine from a
dealer, or steal it from a police department. In the latter, I may
have to return it if being caught, in the former, it will be
confiscated (by which act it passes into the possession of the state
in which it was not before).
Ken Arromdee
> This is probably not the case - for example, if I was a company
>making Bargain PC's and distributing copies of Windows98 that I had
>obtained from a second source, I'd guess the courts would take a dim
>view of that. But that's not the same as the end user not owning the copy.
So if a company distributes illegal copies of Windows 98, the end user of
a Windows copies still owns it?
Steve Peltz
David Kastrup <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
>If you buy a pirated book, you have acquired the right to burn the
>book in your chimney, or to wipe your behind with pages from it, or to
>roll cigarettes with its pages and smoke them. You have not acquired
>the right to read it.
The copyright owner owns the right to COPY the material. He does not
own the right to control who can READ the material, once it is published.
The GPL explicitly states that deficiencies by a distributor do not
inhibit the recipient in any way. There is no requirement that you agree
to the GPL in order to RECEIVE a copy of it, and it is pretty clear that
is the intent of the GPL. Adding in extra conditions such as "you must
agree to the conditions of the GPL in order to read, compile or modify
the program, and you must not give a copy of the program to anyone who
doesn't agree to these conditions" would be totally at odds with the
intent of the GPL.
Steve Peltz
Barry Margolin <bar...@bbnplanet.com> wrote:
>I was referring to the spirit of the law. Does it make sense for a law to
>prohibit someone from shipping A+B when it allows shipping A along with
>instructions saying "to make this work, you have to add B to it"?
Of COURSE it makes sense. If I write a book about how to use Windows
1901, it is clearly not very useful unless you actually have a copy of
that program.
There is a HUGE difference between my selling my book, and my selling
the book along with a copy of Windows 1901 (other than the fact that I
couldn't sell my book for 3 more years). Do you actually believe that
copyright law says I have to have permission from Microsoft in order to
even write my book, much less distribute it?
Selling a program (or even giving away a program) that runs on a Microsoft
OS is not the equivalent of selling the program along with a copy of
the OS. I'm not "working around" that I can't distribute them together
by having the user do the installation. Now, if Microsoft had something
in the license agreement saying "users may not install programs on this
operating system, unless they are officially approved by Microsoft",
AND such a license was held to be valid, then possibly selling such a
program could be held to be an infringement, since it would be encouraging
the user to violate the license agreement.
However, the GPL explicitly, and by intent, does NOT restrict the user
from doing anything besides distributing copies.
Lynn Winebarger
Ken Arromdee <karr...@nyx.nyx.net> wrote:
>owin...@ezinfo.ucs.indiana.edu (Lynn Winebarger) wrote:
>> This is probably not the case - for example, if I was a company
>>making Bargain PC's and distributing copies of Windows98 that I had
>>obtained from a second source, I'd guess the courts would take a dim
>>view of that. But that's not the same as the end user not owning the copy.
>
>So if a company distributes illegal copies of Windows 98, the end user of
>a Windows copies still owns it?
That I don't know - what I said was that no one has shown here that
the end user does _not_ own the copy, or that (in the case you suggest)
that MS owns the copy. Clearly MS controls the copyright, but that is not
the same as the copy, or even the expression itself - it merely means it
has the exclusive right to make copies (and a few other things, like
publicly perform) of the work. Nothing more. Of course, that copy may
be confiscated and destroyed by court order resulting from a lawsuit -
that still would not mean MS owned the copy, or that the end user did
not.
Like I said, a lawyer experienced in these issues could probably
address this issue with no trouble. I would still like to see some
evidence (case or statute citations) that it is illegal to make use of
an infringing copy, especially when the original work is published
(unpublished works usually get more protection against fair use
arguments). That's a very strong claim.
Lynn
Barry Margolin
Steve Peltz <pe...@medusa.nn.com> wrote:
>In article <7zf73.460$KM3.139232@burlma1-snr2>,
>Barry Margolin <bar...@bbnplanet.com> wrote:
>>I was referring to the spirit of the law. Does it make sense for a law to
>>prohibit someone from shipping A+B when it allows shipping A along with
>>instructions saying "to make this work, you have to add B to it"?
>
>Of COURSE it makes sense. If I write a book about how to use Windows
>1901, it is clearly not very useful unless you actually have a copy of
>that program.
Sure it is. I don't use Windows, but I bought an MCSE (or maybe it was
MCNE) book because I often have to provide technical support to customers
who do. While instructional books like this have additional utility when
used in conjunction with what they're describing, they can also stand alone
for intellectual purposes.
On the other hand, a program in binary form serves no practical purpose
unless it can be run. If it has to be linked with a particular library,
it's just a waste of disk space if the library isn't there.
>Selling a program (or even giving away a program) that runs on a Microsoft
>OS is not the equivalent of selling the program along with a copy of
>the OS. I'm not "working around" that I can't distribute them together
>by having the user do the installation. Now, if Microsoft had something
>in the license agreement saying "users may not install programs on this
>operating system, unless they are officially approved by Microsoft",
>AND such a license was held to be valid, then possibly selling such a
>program could be held to be an infringement, since it would be encouraging
>the user to violate the license agreement.
>
>However, the GPL explicitly, and by intent, does NOT restrict the user
>from doing anything besides distributing copies.
I think there are some practical, common sense issues involved here.
Microsoft ships an OS with the express purpose of allowing customers to
load applications onto it. These applications are not considered
derivative works of the OS. They're separate pieces of software, and
application developers can safely assume that the OS will be preexisting on
the system that the application is loaded onto.
Readline is different. Except perhaps for some Linux releases, readline
cannot be assumed to exist already. For the application to be complete,
the vendor should include it. If he doesn't do so, but instead says
"Customer must finish putting this program together by downloading readline
and linking them together", something seems wrong. If this is allowed,
what's the point of preventing the vendor from doing the link himself? All
it does is make things harder for the end user, which is the group that GPL
is intended to benefit. And what's the point of distinguishing the GPL
from the LGPL? It seems like the use of shared libraries can effectively
turn any GPL'ed library into an LGPL'ed library.
Fergus Henderson
>owin...@ezinfo.ucs.indiana.edu (Lynn Winebarger) wrote:
>>>In order to use the software, you have to get it. Getting it, which involves
>>>making a copy, _is_ protected by copyright. If a friend made you a copy of
>>>Microsoft Windows, you would not have any right to use that copy at all.
>>>Thus, a modified GPL which let you use the copy under certain conditions would
>>>be adding rights to the default of no rights at all.
>> Owning the copy would give me rights to it. While the distributor
>>might well be liable for not following the terms of the GPL in making a
>>copy, once I own the copy, it's my property.
>
>You seem to be saying that if the distributor makes you a copy illegally, the
>distributor can be sued, but you own the copy.
Yes. A court can order that the illegal copies be confiscated,
but until and unless such a court order is made, you own the copy.
I'm not a lawyer, but I know how to search for legislation on the
Internet. Here's a quote from the Australian copyright act.
| COPYRIGHT ACT 1968 - SECT 116
|
| Rights of owner of copyright in respect of infringing copies
|
| SECT. (1) The owner of the copyright in a work or other subject-matter
| may bring an action for conversion or detention in relation to:
|
| (a) an infringing copy
...
| (1A) In an action for conversion or detention, a court may grant to the
| owner of the copyright all or any of the remedies that are available in
| such an action as if:
|
| (a) the owner of the copyright had been the owner of the infringing
| copy since the time the copy was made
Note that the legislation says that a court may make an order AS IF
the copyright owner had been the owner of the infringing copy. It seems
to me that this implies that if an infringing copy is sold to some purchaser,
then up until such an order is made, the copyright owner is NOT the owner
of the infrinding copy -- the purchaser rather than the copyright
owner is the owner of the infringing copy.
>This isn't true. Again, consider what happens if it is Windows 95 instead of
>GPL software. Would you claim that the distributor can be sued by Microsoft,
>but you can use the copy of Windows as you see fit?
>
>Of course not. If the copy is illegal, the person receiving the copy doesn't
>end up owning it. They only own the copy if it was legally made.
I don't find your assertions convincing here. What are they based on?
I think it might well be the case that you can use the copy of Windows
as you see fit, since copyright law only governs copying (and creation
of derivative works, public performance, etc.), not use. The copyright
owner can "bring an action for conversion or detention", but unless and
until such a court order is made, I think you can indeed use the copy
as you see fit.
I agree with Lynn Winebarger, a lawyer's opinion would be helpful here.
Isaac
>Readline is different. Except perhaps for some Linux releases, readline
>cannot be assumed to exist already. For the application to be complete,
>the vendor should include it. If he doesn't do so, but instead says
>"Customer must finish putting this program together by downloading readline
>and linking them together", something seems wrong. If this is allowed,
This anomaly only exists because readline is already distributed in
every nook and cranny of the net with a license that says to users
that they can link it with anything they chose. No one needs to distribute
readline at all. Further what about programs specifically targetted for
Linux? Your argument allows the possibility that such programs could
be linked against readline, since we know that readline is already present
on those systems (AFAIK all and not just some).
>what's the point of preventing the vendor from doing the link himself? All
>it does is make things harder for the end user, which is the group that GPL
>is intended to benefit. And what's the point of distinguishing the GPL
>from the LGPL? It seems like the use of shared libraries can effectively
>turn any GPL'ed library into an LGPL'ed library.
This argument basically says, that you find the conclusion unacceptable,
and thus the premise must be wrong. This is essentially the FSF's
argument as well. I find such arguments unpersuasive. Since the
anomaly results strictly from the licensing in the GPL, you have to
consider the possibility that the GPL is flawed.
Isaac
Leslie Mikesell
David Kastrup <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
>If you buy a pirated book, you have acquired the right to burn the
>book in your chimney, or to wipe your behind with pages from it, or to
>roll cigarettes with its pages and smoke them. You have not acquired
>the right to read it.
When did copyright laws start attempting to control how you could
use something beyond duplicating it? And in this scenario, what
would you expect from someone who read the contents before learning
that the copy was not published legally? Does he have to forget
that he read it?
Les Mikesell
l...@mcs.com
NF Stevens
[snip]
>This argument basically says, that you find the conclusion unacceptable,
>and thus the premise must be wrong. This is essentially the FSF's
>argument as well. I find such arguments unpersuasive. Since the
>anomaly results strictly from the licensing in the GPL, you have to
>consider the possibility that the GPL is flawed.
Or that the copyright laws do not give the FSF the power
to enforce their wishes.
Norman
Lynn Winebarger
David Kastrup <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
>owin...@ezinfo.ucs.indiana.edu (Lynn Winebarger) writes:
>
>> In article <92903238...@iris.nyx.net>,
>> Ken Arromdee <karr...@nyx.nyx.net> wrote:
>> >owin...@ezinfo.ucs.indiana.edu (Lynn Winebarger) wrote:
>> >> This is probably not the case - for example, if I was a company
>> >>making Bargain PC's and distributing copies of Windows98 that I had
>> >>obtained from a second source, I'd guess the courts would take a dim
>> >>view of that. But that's not the same as the end user not owning the copy.
>> >
>> >So if a company distributes illegal copies of Windows 98, the end user of
>> >a Windows copies still owns it?
>>
>> That I don't know - what I said was that no one has shown here that
>> the end user does _not_ own the copy, or that (in the case you suggest)
>> that MS owns the copy.
>
>physical material of the copy, but Microsoft owns the intellectual
>property on it. Microsoft has the right to demand that you hand over
Here's where your wrong, at least in the US. You don't own
intellectual "property", you have monopoly rights on certain aspects of
its use - copying, public performance, etc. You do not own the
information or ideas expressed.
>
>It is a very bland and ordinary claim. Otherwise, pirated copies of
>books might be sold far and wide quite legally, and only the immediate
>producer of the copy might fear any problems, where as all bookstores
>and intermediaries would claim that they have all the right to make
>use of this infringing copy by selling it for a profit.
>
Any claim about not being able to read something that is widely
published is a strong claim. As for the bookstores, they are different
from the end users - reselling pirated copies (particularly on a wide
basis) is different from reading those copies as an end user.
Lynn
Fergus Henderson
>f...@cs.mu.oz.au (Fergus Henderson) writes:
>
>[pirated copy as example]
>
>> I think it might well be the case that you can use the copy of
>> Windows as you see fit, since copyright law only governs copying
>> (and creation of derivative works, public performance, etc.), not
>> use.
>
>Nope, I can't, because I have never indulged in any act to make the
>intellectual property accessible to me legally.
Which law school did you say you went to?
Which legislation or case law is your argument based on?
Or did you just make it up, based on your own "common sense"
understanding of what you *think* the law is or should be?
Last time I looked, the Australian copyright act never used the words
"intellectual property", but instead talked about "copyright",
and never prohibited reading books, but instead only prohibited
copying them (and certain related acts such as creating derived
works, making public performances, etc.).
If you can cite some legislation, case law, legal authority, or precedent
to back up your case, I might believe you. But if your argument is
based only on your analogy between physical property and intellectual
property, then I find it unconvincing.
Isaac
<d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
>
>But the material in the pirated book is not "published" by him. His
>copyright ensures that his material may be used according to fair use
>only when some consideration has passed to the author. In the case of
>a pirated book, this has not happened. You cannot acquire the right
>to read the stuff in a pirated book, in the same way as you cannot
>acquire the right to use or own something more material you buy from
>someone who has stolen it.
>
You overstate the rights that copyright law gives to the copyright
holder (at least in the US). Copyright law allows the copyright
holder the right to control copying, public performance, creating
derived works, distribution and other specifically enumerated things.
The right to restrict reading a book, or to letting a friend read the
book at your house are not amoung such rights, unless a transfer
or distribution is concerned.
People go into bookstores and read entire books without buying them
all the time. Some bookstores encourage this. The author gets
nothing from such readers. No consideration passed, period.
For the US, the place to read about copyright is Title 17 of the
US Code. Section 106 lists the exclusive rights of the copyright
holder. It's online, check it out.
Isaac
Geoffrey KEATING
> David Kastrup <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> writes:
>
> >f...@cs.mu.oz.au (Fergus Henderson) writes:
> >
> >[pirated copy as example]
> >
> >> I think it might well be the case that you can use the copy of
> >> Windows as you see fit, since copyright law only governs copying
> >> (and creation of derivative works, public performance, etc.), not
> >> use.
> >
> >Nope, I can't, because I have never indulged in any act to make the
> >intellectual property accessible to me legally.
...
> Last time I looked, the Australian copyright act never used the words
> "intellectual property", but instead talked about "copyright",
> and never prohibited reading books, but instead only prohibited
> copying them (and certain related acts such as creating derived
> works, making public performances, etc.).
>
> If you can cite some legislation, case law, legal authority, or precedent
> to back up your case, I might believe you. But if your argument is
> based only on your analogy between physical property and intellectual
> property, then I find it unconvincing.
Most non-free computer software, including Windows, is licensed, not
sold. So you don't buy a copy of Windows, what you buy is something
like a lease on the copy. Usually the license says something like
"you own the physical media that the software is recorded on, but we
own the software and permit you to use it on these conditions".
I'm not completely convinced that this in enforceable. Certainly, ads
that say things like "we sell Office 2000 for $2000" are quite likely
to be misleading advertising.
This is one of the nice things about the GPL; you own any copies
of GPLed software that you make.
--
Geoff Keating <Geoff....@anu.edu.au>
ma...@usai.asiainfo.com
Martin Moller Pedersen <tu...@daimi.au.dk> wrote:
>Mark Harrison wrote:
>
>-> Would any really feeble readline function work? This seems to be
>-> compatible at the API level.
>->
>-> char *readline(char *prompt)
>-> {
>-> puts(prompt);
>-> fflush(stdout);
>-> return gets(malloc(1000));
>-> }
>
>The readline function is a lot more than just this above.
>readline works together with the history.h lib.
OK, adding
void add_history(char* line) {}
makes this simple (but useful) skeleton application work:
main()
{
char *line;
while ((line = readline("hello: ")) != NULL) {
if (*line)
add_history(line);
puts(line);
free(line);
}
}
So, suppose I put my crummy readline() and add_history() in a shared
library and distribute that shared library with my non-GPL program,
and that the end user chooses to replace my shared library with
readline.so.
My source code is non-GPL, since I didn't incorporate any
other GPL code, right? So has this tiny bit of code
circumvented the readline GPL?
Ottavio G. Rizzo
> Most non-free computer software, including Windows, is licensed, not
> sold. So you don't buy a copy of Windows, what you buy is something
> like a lease on the copy. Usually the license says something like
> "you own the physical media that the software is recorded on, but we
> own the software and permit you to use it on these conditions".
>
> I'm not completely convinced that this in enforceable. Certainly, ads
> that say things like "we sell Office 2000 for $2000" are quite likely
> to be misleading advertising.
>
> This is one of the nice things about the GPL; you own any copies
> of GPLed software that you make.
No way: it's still just licenced to you. If you actually owned it,
you'd be able to re-release it under a different licencing scheme.
--
Ottavio Rizzo IRMAR, Campus de Beaulieu
ri...@maths.univ-rennes1.fr Université de Rennes 1
Tél +33 (0)2 99 28 67 92 35042 RENNES cedex
Fax +33 (0)2 99 28 67 90 FRANCE
David Kastrup
> Geoffrey KEATING <geo...@discus.anu.edu.au> writes:
>
> > Most non-free computer software, including Windows, is licensed, not
> > sold. So you don't buy a copy of Windows, what you buy is something
> > like a lease on the copy. Usually the license says something like
> > "you own the physical media that the software is recorded on, but we
> > own the software and permit you to use it on these conditions".
> >
> > I'm not completely convinced that this in enforceable. Certainly, ads
> > that say things like "we sell Office 2000 for $2000" are quite likely
> > to be misleading advertising.
> >
> > This is one of the nice things about the GPL; you own any copies
> > of GPLed software that you make.
>
> No way: it's still just licenced to you. If you actually owned it,
> you'd be able to re-release it under a different licencing scheme.
You own the copy and might do all with it copyright law permits. If
you accept the license, you can do more.
In contrast, standard licenses prohibit you to do a lot of things
copyright law permits. For that reason they usually include some
phrase "if you don't relish that indeed we want you to relinquish all
sorts of rights you might have been led into believing you have
acquired, give the stuff back for a refund".
Fergus Henderson
>Geoffrey KEATING <geo...@discus.anu.edu.au> writes:
>
>> This is one of the nice things about the GPL; you own any copies
>> of GPLed software that you make.
>
>No way: it's still just licenced to you. If you actually owned it,
>you'd be able to re-release it under a different licencing scheme.
No, you're confusing owning the copies with owning the copyright.
David Kastrup
> On 11 Jun 1999 12:14:49 +0200, David Kastrup
> <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
> >
> >But the material in the pirated book is not "published" by him. His
> >copyright ensures that his material may be used according to fair use
> >only when some consideration has passed to the author. In the case of
> >a pirated book, this has not happened. You cannot acquire the right
> >to read the stuff in a pirated book, in the same way as you cannot
> >acquire the right to use or own something more material you buy from
> >someone who has stolen it.
>
> You overstate the rights that copyright law gives to the copyright
> holder (at least in the US). Copyright law allows the copyright
> holder the right to control copying, public performance, creating
> derived works, distribution and other specifically enumerated
> things.
Correct. So it is the copyright holder who is in control of who has a
legal copy.
> The right to restrict reading a book, or to letting a friend read the
> book at your house are not amoung such rights, unless a transfer
> or distribution is concerned.
The right to have this book in the first place is. There is no
inherent right of a person buying a pirated copy of the book to be
allowed to make use of the pirated content.
> People go into bookstores and read entire books without buying them
> all the time. Some bookstores encourage this. The author gets
> nothing from such readers. No consideration passed, period.
But those people do not have the *right* to read it. You cannot go
into a bookstore and demand that you may read through entire books.
Lynn Winebarger
Geoffrey KEATING <geo...@discus.anu.edu.au> wrote:
>
>Most non-free computer software, including Windows, is licensed, not
>sold. So you don't buy a copy of Windows, what you buy is something
>like a lease on the copy. Usually the license says something like
>"you own the physical media that the software is recorded on, but we
>own the software and permit you to use it on these conditions".
>
same as actually being licensed. To determine whether your software is
_really_ licensed or just sold, break one of the terms of the "EULA",
inform the company, and be ready to go to court.
>I'm not completely convinced that this in enforceable. Certainly, ads
>that say things like "we sell Office 2000 for $2000" are quite likely
>to be misleading advertising.
>
Or, perhaps "EULA"'s are <almost fraudulent> claims.
Lynn
David Kastrup
> David Kastrup <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> writes:
>
> >f...@cs.mu.oz.au (Fergus Henderson) writes:
> >
> >[pirated copy as example]
> >
> >> I think it might well be the case that you can use the copy of
> >> Windows as you see fit, since copyright law only governs copying
> >> (and creation of derivative works, public performance, etc.), not
> >> use.
> >
> >Nope, I can't, because I have never indulged in any act to make the
> >intellectual property accessible to me legally.
>
> Which law school did you say you went to?
Even if I did, it would have been a German one.
> Last time I looked, the Australian copyright act never used the words
> "intellectual property", but instead talked about "copyright",
> and never prohibited reading books, but instead only prohibited
> copying them (and certain related acts such as creating derived
> works, making public performances, etc.).
Right. Acquiring a legal copy makes it possible for you to read the
stuff. Acquiring an illegal copy gives you no right to read the
stuff. You might as well have broken into the author's home and
copied his manuscripts. That would give you the same rights than
acquiring an illegal copy does.
Copyright law governs what constitutes a legally usable copy. You
might as well claim that it is perfectly legal to watch TV on a stolen
TV set, since property laws do not pose restrictions on watching TV.
Isaac
<d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
>irc...@latveria.castledoom.org (Isaac) writes:
>> People go into bookstores and read entire books without buying them
>> all the time. Some bookstores encourage this. The author gets
>> nothing from such readers. No consideration passed, period.
>
>But those people do not have the *right* to read it. You cannot go
>into a bookstore and demand that you may read through entire books.
Okay, but the point is not that they have the right, but whether the
copyright holder rather than the bookstore owner is the entity that
has the right to stop them.
The copyright holder cannot stop me from wallpapering my house
with pages from his/her book and then allowing everbody who visits
my house to read the book. The copyright holder can keep me from
making copies of the book, but not from using the original pages.
Isaac
Isaac
>
>Right. Acquiring a legal copy makes it possible for you to read the
>stuff. Acquiring an illegal copy gives you no right to read the
>stuff. You might as well have broken into the author's home and
>copied his manuscripts. That would give you the same rights than
>acquiring an illegal copy does.
>
A person who steals a book and reads it is a thief and not a
copyright violator. I don't see any protected right to restrict
reading (so long as a copy is not made) in US copyright law.
The right to perform, copy, or create derivative works is reserved
to the copyright holder, but reading is not. Could you cite a
provision of German or US law that would demonstrate your point?
For software the situation is different because usage implies
making a copy in memory. You cannot use software without
either violating copyright or invoking fair use, so your argument
would be true for software, but not for books.
Isaac
Isaac
<ma...@usai.asiainfo.com> wrote:
>So, suppose I put my crummy readline() and add_history() in a shared
>library and distribute that shared library with my non-GPL program,
>and that the end user chooses to replace my shared library with
>readline.so.
>
to parse user input, I don't see how your stuff wouldn't
satisfy the requirement that an alternate library with the
same API be available.
In fact the exercise is so trivial in the case of readline, that
it seems silly to even require someone to do it. Of course,
the licence on readline has in at least one instance encouraged
someone to GPL their code. Perhaps this is reason enough for the
FSF to stick to their position.
>My source code is non-GPL, since I didn't incorporate any
>other GPL code, right? So has this tiny bit of code
>circumvented the readline GPL?
Yes. The FSF would have to assert an interface copyright to stop
you, and they are loathe to do that. But we know that the copyright
holder doesn't want readline used this way, so we shouldn't deliberately
do this. Legal or not, it's still wrong.
Isaac
David Kastrup
> On 18 Jun 1999 11:00:31 +0200, David Kastrup
> <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
> >irc...@latveria.castledoom.org (Isaac) writes:
> >> People go into bookstores and read entire books without buying them
> >> all the time. Some bookstores encourage this. The author gets
> >> nothing from such readers. No consideration passed, period.
> >
> >But those people do not have the *right* to read it. You cannot go
> >into a bookstore and demand that you may read through entire books.
>
> Okay, but the point is not that they have the right, but whether the
> copyright holder rather than the bookstore owner is the entity that
> has the right to stop them.
Yes. If a bookstore advertises that it has reading rooms where you
can read through entire books in perfect leisure, then several
publishers would make trouble. A bookstore has not the permission to
turn itself into a library or a reading room without negotiating with
the copyright defenders.
> The copyright holder cannot stop me from wallpapering my house
> with pages from his/her book and then allowing everbody who visits
> my house to read the book.
I am not so sure. I quote from a book:
Except in the Unites States of America, this book is sold subject
to the condition that it shall not, by way of trade or otherwise,
be lent, re-sold, hired out, or otherwise circulated without the
publisher's prior consent in any form of binding or cover other
than that in which it is published and without a similar condition
including this condition being imposed on the subsequent purchaser.
Of course, this might not be legally binding.
David Kastrup
> On 18 Jun 1999 15:12:27 +0200, David Kastrup
> <d...@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
> >
> >Right. Acquiring a legal copy makes it possible for you to read the
> >stuff. Acquiring an illegal copy gives you no right to read the
> >stuff. You might as well have broken into the author's home and
> >copied his manuscripts. That would give you the same rights than
> >acquiring an illegal copy does.
> >
>
> A person who steals a book and reads it is a thief and not a
> copyright violator.
You fail to understand my argument. Read it again. I was not talking
about stealing a book.
> The right to perform, copy, or create derivative works is reserved
> to the copyright holder, but reading is not.
You cannot read a book if you don't have a copy.
Isaac
>
>Yes. If a bookstore advertises that it has reading rooms where you
>can read through entire books in perfect leisure, then several
>publishers would make trouble. A bookstore has not the permission to
>turn itself into a library or a reading room without negotiating with
>the copyright defenders.
>
such challenges by publishers, but perhaps they've gotten permission
from every publisher : ). The place is an extremely comfortable place
to sit and browse/read. It's always full of people doing just that
when I go in. There's even a coffee shop inside. No reading rooms
I guess. Just comfortable chairs and tables.
>> The copyright holder cannot stop me from wallpapering my house
>> with pages from his/her book and then allowing everbody who visits
>> my house to read the book.
>
>I am not so sure. I quote from a book:
>
> Except in the Unites States of America, this book is sold subject
> to the condition that it shall not, by way of trade or otherwise,
> be lent, re-sold, hired out, or otherwise circulated without the
> publisher's prior consent in any form of binding or cover other
> than that in which it is published and without a similar condition
> including this condition being imposed on the subsequent purchaser.
>
>Of course, this might not be legally binding.
Perhaps it is binding somewhere. I notice it excludes the US where
copyright law expressly allows resale without the copyright holders
permission. Perhaps elsewhere such resale is subject to publishers
consent.
I don't think pasting pages on a wall in my house constitutes resale,
lending, hiring, trade, or circulating though. So even where the
provision is binding it's not clear that wall papering can be
prohibited.
Isaac