air traffic control software

air traffic control software

[Terry J. Westley]

I'm looking for the air surveillance and graphics display
portion of air traffic control (ATC) software.

Is there any such thing as GPL'd or GOTS ATC code?

What about AAS?  Is any of that code available for DOD programs?

--
Terry J. Westley, Principal Engineer
Veridian Engineering, Calspan Operations
P.O. Box 400, Buffalo, NY 14225
twestley@buffalo.veridian.com    http://www.veridian.com

GPL and "free" software (was: air traffic control software)

[Terry J. Westley]

Kevin Rigotti <rigotti@atc.dera.gov.uk> wrote:
> ATC software is rather expensive to produce, so you are
> not likely to find too much of it for free,

Robert Dewar <robert_dewar@my-dejanews.com> responded:
> First, this is not what he asked, he asked whether there
> was GPL'ed code available. This has nothing to do with
> being free of cost. There is GPL code that costs money,
> and non-GPL'ed code that you can get without paying
> anything. These two factors (license used for distribution,
> and cost charged) are not linked in the manner you assume.

Robert is correct.  I also asked about GOTS (Government Off
The Shelf).  Lot's of GOTS software is available "free" to
those who qualify.  A lot of it wasn't cheap to produce.

We are currently using (and integrating some into our products)
Samba, R, emacs, Tcl/Tk, perl, and other GPL'd and open source
products.

(On a personal note, I've recently decided to release TASH, an
Ada binding to Tcl/Tk [http://tash.calspan.com] under the GPL.)

I wish a lot of success to ACT, Scriptics, Redhat, and others
who are pioneering business models based on open source products.
Even though our product is not and never will be public open
source, I am trying to convince our management that our customers
need and want source code.  I believe this will build their
confidence in our product and give them a level of ownership
that enhances the product.  If it doesn't, we need to improve our
code, not hide it.

--
Terry J. Westley, Principal Engineer
Veridian Engineering, Calspan Operations
twestley@buffalo.veridian.com   http://www.veridian.com/
-------------------------------------------------------
Author of TASH, an Ada binding to Tcl/Tk.
Visit the TASH web site at http://tash.calspan.com.
-------------------------------------------------------

Re: GPL and "free" software (was: air traffic control software)

[dennison]

In article <7fl5ac$9sh$1@news2.tor.accglobal.net>,
  "Terry J. Westley" <twestley@buffalo.veridian.com> wrote:

> (On a personal note, I've recently decided to release TASH, an
> Ada binding to Tcl/Tk [http://tash.calspan.com] under the GPL.)

Is that the full unadulterated GPL, or the Gnat modified GPL (GMG?) ?

--
T.E.D.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software (was: air traffic control software)

[bill]

In article <7fl5ac$9sh$1@news2.tor.accglobal.net>, "Terry says...
>
 
>I wish a lot of success to ACT, Scriptics, Redhat, and others
>who are pioneering business models based on open source products.
>Even though our product is not and never will be public open
>source, I am trying to convince our management that our customers
>need and want source code.  I believe this will build their
>confidence in our product and give them a level of ownership
>that enhances the product.  If it doesn't, we need to improve our
>code, not hide it.

I second that.

I also think that someone working on opensource code, will try harder
to make their code better, for the simple reason, is that alot of people
will be looking at it. 

(Look at how well formated and consistant GNAT code is, I think the fact that
it is in the public eye, is what pushed ACT to do that).

I find that if a programmer knows someone will be looking and 
examining their code, they will more likely work harder to make
the code better (programmer pride thing).  I knew programmers who 
spend years coding, and no one looks at what they wrote. these 
programmer will in most cases, not bother too much in improving 
their code or even how well written and well documented it is.

Bill
  

Re: air traffic control software

[Kevin Rigotti]

Terry J. Westley wrote in message <7fibd5$jc7$1@news2.tor.accglobal.net>...
>I'm looking for the air surveillance and graphics display
>portion of air traffic control (ATC) software.
>
>Is there any such thing as GPL'd or GOTS ATC code?


Thankyou, it's nice to start the day with a good laugh :-)

ATC software is rather expensive to produce, so you are not likely to
find too much of it for free, but you might be interested in some of the ATC
research sponsored by Eurocontrol

Try www.eurocontrol.fr/projects/eons/


Kevin
ATC Systems Group, DERA Malvern

Re: air traffic control software

[Robert Dewar]

In article <7fjucn$k4p$1@trog.dera.gov.uk>,
  "Kevin Rigotti" <rigotti@atc.dera.gov.uk> wrote:

> ATC software is rather expensive to produce, so you are
> not likely to find too much of it for free,

First, this is not what he asked, he asked whether there
was GPL'ed code available. This has nothing to do with
being free of cost. There is GPL code that costs money,
and non-GPL'ed code that you can get without paying
anything. These two factors (license used for distribution,
and cost charged) are not linked in the manner you assume.

Second, whether software is available under the GPL is
not necessarily a function of the cost of producing the
software. There is expensive software available under the
GPL, and there is software that must have cost almost
nothing to make sold as proprietary software under very
restrictive licenses.

For example, I am sure that the Netscape browser was quite
an expensive piece of software to build, yet it is released
under the NPL, which is quite similar to the GPL.

There are many different business models for the
distribution of software.

The old assumptions that you cannot afford to distribute
"expensive" software under the GPL or similar licenses are
being questioned by new announcements every day!

So you may laugh now, but we will see, perhaps the last
laugh will be in the other direction ...




-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software (was: air traffic control software)

[Robert Dewar]

In article <7fl9q5$ab7@drn.newsguy.com>,
  bill@nospam wrote:
> (Look at how well formated and consistant GNAT code is, I
> think the fact that it is in the public eye, is what
> pushed ACT to do that).

Well perhaps partly, but there is plenty of open source
code that is NOT well documented. The main key to well
documented code is *caring* about documentation!
>
> I find that if a programmer knows someone will be looking
> and  examining their code, they will more likely work
> harder to mak the code better (programmer pride thing).
> I knew programmers who spend years coding, and no one
> looks at what they wrote. these programmer will in most
> cases, not bother too much in improving  their code or
> even how well written and well documented it is.

Hmmm! Well I think I can fairly say that all the code I
ever wrote has been fanatically well documented (e.g. look
at the SPITBOL 360 sources from a long time back). Of
course most of this code was indeed proprietary, and no one
can ever see it which is a pity. I definitely like the
fact that other people can read the GNAT code.

What has been remarkable at ACT is that we have developed
a whole community of engineers all of whom really care
about the documentation and the quality of the sources, and
I am sure that the fact that this code is read by others
and is open played a significant part in being able to
create this community.

And thanks for the nice comments about the GNAT code, they
are appreciated :-)

Robert Dewar (and the GNAT team)

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software (was: air traffic control software)

[Robert Dewar]

In article <7fl9q5$ab7@drn.newsguy.com>,
  bill@nospam wrote:
> In article <7fl5ac$9sh$1@news2.tor.accglobal.net>, "Terry
> says...
> I also think that someone working on opensource code,
> will try harder
> to make their code better, for the simple reason, is that
> a lot of people will be looking at it.

One more point here. To me open source code MUST be well
documented, or the point of open source is lost. The GPL
can only require sources to be made available, not good
quality sources, but the spirit behind the GPL requires
that the sources be accessible.

One could meet the letter of the GPL by providing
obfuscated sources with no comments, but if all your
assignment statements look like

OOOO000OO0OOOO := O0000O000OO000O0 + O0000O0O00000;

with no comments, I think most people would agree that you
have not *really* makde the sources available.

Similarly if you supply a parser which has giant tables
of LRK junk, without supplying the programs that built
the tables, you might meet the letter of the GPL
requirement but definitely not the spirit.

I once worked on a compiler project where the contract had
a clause forbidding the use of any table driven techniques.
Why? Because on a previous project which required source
delivery, the contractor had done just that (delivered
LALR parsing tables without the tool that built them!)

Robert Dewar

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software (was: air traffic control software)

[Larry Kilgallen]

In article <7foo6s$qbm$1@nnrp1.dejanews.com>, Robert Dewar <robert_dewar@my-dejanews.com> writes:

> One could meet the letter of the GPL by providing
> obfuscated sources with no comments, but if all your
> assignment statements look like
> 
> OOOO000OO0OOOO := O0000O000OO000O0 + O0000O0O00000;
> 
> with no comments, I think most people would agree that you
> have not *really* makde the sources available.

Certainly not the sources from which the originator would
make changes.

Larry Kilgallen

Re: GPL and "free" software (was: air traffic control software)

[Tom Moran]

>Because on a previous project which required source
>delivery, the contractor had done just that (delivered
>LALR parsing tables without the tool that built them!)
  Presumably that kind of nonsense could be settled in court, with
expert testimony as to what constitutes "source".  (The courts being
supposed to require some credibility of the "expert witnesses"
nowadays.)

Re: GPL and "free" software (was: air traffic control software)

[Robert Dewar]

In article <1999Apr23.065716.1@eisner>,
  Kilgallen@eisner.decus.org.nospam wrote:
> Certainly not the sources from which the originator would
> make changes.

Well yes, but obviously this cannot easily be defined
legally, and the GPL certainly makes no attempt to do so!

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software (was: air traffic control software)

[Florian Weimer]

tmoran@bix.com (Tom Moran) writes:

>   Presumably that kind of nonsense could be settled in court, with
> expert testimony as to what constitutes "source".  (The courts being
> supposed to require some credibility of the "expert witnesses"
> nowadays.)

Perhaps.  But we do not know yet whether the GPL will hold up in
court anyway.

Re: GPL and "free" software (was: air traffic control software)

[Robert Dewar]

In article
<m34sm64cz2.fsf@deneb.cygnus.stuttgart.netsurf.de>,
  Florian Weimer <fw@cygnus.stuttgart.netsurf.de> wrote:
> tmoran@bix.com (Tom Moran) writes:
>
> Perhaps.  But we do not know yet whether the GPL will
> hold up in court anyway.

What do you mean by "hold up" here? Most certainly anyone
who copies software in clear violation of the permissions
of the GPL, e.g. by not making sources available, is
committing a simple copyright violation. The courts have
clearly affirmed the intention of congress that software
be subject to the full protection of the copyright law!

Now if there is judgment involved, e.g. I make the sources
available, but you have to wait XXX months for me to fill
your order for them, then some value of XXX is
unacceptable, and indeed the courts would have to make
a judgment on this issue.

But the presumption is pretty strong that anyone who copies
a copyrighted work must be able to demonstrate that they
have a license to do so, i.e. the burden of proof would be
on the copier in this case to show that the copying they
had done was not a copyright violation. This was for
example one of the arguments that Microsoft was using in
the Apple case, they argued they had a license for the
copying they did.

Of course the software has to be legitimately copyrighted
and there is a whole section of case law devoted to exactly
what that means!



-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Nick Roberts]

Wading rather trepidly into this issue, and with the understanding that I am
not any kind of legal expert (in UK or US law), I have wondered if there is
a potential problem with the GPL and similar licences, which I'll try to
describe.

Suppose party B obtains a work (of software), under the GPL, from party A
(paying nothing for it).  Then suppose party B modifies this work to form a
work which they then sell to party C (who is not aware of its GPL heritage)
for a substantial amount of money.

Now, suppose party C subsequently discovers the fact that the work is
derived from a GPL work; what can they do?  It seems to me that if party C
attempts to sue B, B can argue that C is not the injured party, because it
is not C's licence who's terms were breached.  On the other hand, it seems
to me that if party A attempts to sue B (on behalf of C), B can argue that A
(or any other party, for that matter) has not suffered any financial loss,
and so are not entitled to any compensation (except, perhaps, for court
costs and legal fees).

Please accept my apologies if this point: has been previously discussed to
death; is actually trivial, naive, or sophistic; is considered inappropriate
to comp.lang.ada or gnu.misc.discuss.

-------------------------------------
Nick Roberts
-------------------------------------

Re: GPL and "free" software

[spblunt]

Well B has done nothing wrong in selling C the software. Neither party can 
really sue B because B hasn't broken the licence (unless they refuse to 
give C the source code upon request).

B can charge as much as they like for the software, and they are under no
obligation to tell C that it can be obtained gratis elsewhere.

Nick Roberts <nickroberts@callnetuk.com> wrote:
> Wading rather trepidly into this issue, and with the understanding that I am
> not any kind of legal expert (in UK or US law), I have wondered if there is
> a potential problem with the GPL and similar licences, which I'll try to
> describe.

> Suppose party B obtains a work (of software), under the GPL, from party A
> (paying nothing for it).  Then suppose party B modifies this work to form a
> work which they then sell to party C (who is not aware of its GPL heritage)
> for a substantial amount of money.

> Now, suppose party C subsequently discovers the fact that the work is
> derived from a GPL work; what can they do?  It seems to me that if party C
> attempts to sue B, B can argue that C is not the injured party, because it
> is not C's licence who's terms were breached.  On the other hand, it seems
> to me that if party A attempts to sue B (on behalf of C), B can argue that A
> (or any other party, for that matter) has not suffered any financial loss,
> and so are not entitled to any compensation (except, perhaps, for court
> costs and legal fees).

Re: GPL and "free" software

[Paul Hughett]

Nick Roberts (nickroberts@callnetuk.com) wrote:
: Wading rather trepidly into this issue, and with the understanding that I am
: not any kind of legal expert (in UK or US law), I have wondered if there is
: a potential problem with the GPL and similar licences, which I'll try to
: describe.

: Suppose party B obtains a work (of software), under the GPL, from party A
: (paying nothing for it).  Then suppose party B modifies this work to form a
: work which they then sell to party C (who is not aware of its GPL heritage)
: for a substantial amount of money.

: Now, suppose party C subsequently discovers the fact that the work is
: derived from a GPL work; what can they do?  It seems to me that if party C
: attempts to sue B, B can argue that C is not the injured party, because it
: is not C's licence who's terms were breached.  On the other hand, it seems
: to me that if party A attempts to sue B (on behalf of C), B can argue that A
: (or any other party, for that matter) has not suffered any financial loss,
: and so are not entitled to any compensation (except, perhaps, for court
: costs and legal fees).

A lot will depend on the details of the transactions, but generally A
can sue B for copyright infringement and C can sue B for fraud.

Paul Hughett

Re: GPL and "free" software

[Robert Dewar]

In article <3723c38b@eeyore.callnetuk.com>,
  "Nick Roberts" <nickroberts@callnetuk.com> wrote:
> On the other hand, it seems
> to me that if party A attempts to sue B (on behalf of C),
> B can argue that A
> (or any other party, for that matter) has not suffered
> any financial loss,
> and so are not entitled to any compensation (except,
> perhaps, for court
> costs and legal fees).

B has violated a copyright. This is a statutory matter.
You should look at copyright law in more detail if you
want to think more about this (detailed discussion of
copyright law seems to far off topic for CLA for me to
wade in here :-)

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <gi71g7.gj.ln@joey>,
  <spblunt@ozemail.com.au> wrote:
> Well B has done nothing wrong in selling C the software.
> Neither party can
> really sue B because B hasn't broken the licence (unless
> they refuse to
> give C the source code upon request).
>
> B can charge as much as they like for the software, and
> they are under no
> obligation to tell C that it can be obtained gratis
> elsewhere.

The question of gratis has nothing to do with this at all,
it is a red herring. The GPL is about freedoms the
recipient has, it is nothing to do with the cost of the
software, so let's drop that confusion right away.

B is most DEFINITELY under the obligation to tell C that
this software is distributed under the GPL. As long as
this is done, there is no problem at all.

I understood the problem as posed to suppose that B did
NOT tell C that this was GPL'ed and that is of course
a clear copyright violation, since B has copied the
software under conditions not allowed by the license.

The GPL has nothing to say about costs.

For example, It is just fine to pick up the GNAT sources
put them on a convenient CD ROM, and charge money for the
CD. Several companies have done that, including most
recently Addison-Wesley, in conjunction with Michael
Feldman's new book.

As far as we are concerned, this is great! We encourage it
and work with the companies that do this. Anything that
helps GNAT distribution is a good thing for the Ada
community.

Now of course if someone tries to sell you a copy of
GNAT 3.11p for $100,000, they are trying to rip you off.
We hope you know enough to refuse, but it is nothing to
do with us in a legal sense -- we will most certainly
try to warn the buyer if we hear about it :-)



-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Barry Margolin]

In article <gi71g7.gj.ln@joey>,  <spblunt@ozemail.com.au> wrote:
>Well B has done nothing wrong in selling C the software. Neither party can 
>really sue B because B hasn't broken the licence (unless they refuse to 
>give C the source code upon request).

The GPL requires you to license the derivative work under the GPL or
something compatible with it.  If B didn't provide such a license to C,
they violated the GPL.

Regarding Nick's question about what rights C has, there's been an ongoing
debate about this in the "embedded systems and GPL" thread for at least a
week.

>B can charge as much as they like for the software, and they are under no
>obligation to tell C that it can be obtained gratis elsewhere.

This is true.  In fact, it might *not* be available elsewhere, unless some
other customer of B has chosen to make it available.

>
>Nick Roberts <nickroberts@callnetuk.com> wrote:
>> Wading rather trepidly into this issue, and with the understanding that I am
>> not any kind of legal expert (in UK or US law), I have wondered if there is
>> a potential problem with the GPL and similar licences, which I'll try to
>> describe.
>
>> Suppose party B obtains a work (of software), under the GPL, from party A
>> (paying nothing for it).  Then suppose party B modifies this work to form a
>> work which they then sell to party C (who is not aware of its GPL heritage)
>> for a substantial amount of money.
>
>> Now, suppose party C subsequently discovers the fact that the work is
>> derived from a GPL work; what can they do?  It seems to me that if party C
>> attempts to sue B, B can argue that C is not the injured party, because it
>> is not C's licence who's terms were breached.  On the other hand, it seems
>> to me that if party A attempts to sue B (on behalf of C), B can argue that A
>> (or any other party, for that matter) has not suffered any financial loss,
>> and so are not entitled to any compensation (except, perhaps, for court
>> costs and legal fees).


-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Robert Dewar]

In article <3723c38b@eeyore.callnetuk.com>,
  "Nick Roberts" <nickroberts@callnetuk.com> wrote:
> Suppose party B obtains a work (of software), under the
> GPL, from party A (paying nothing for it).  Then suppose
> party B modifies this work to form work which they then
> sell to party C (who is not aware of its GPL heritage)
                   ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^

Here is the copyright violation. When the software is sold
to party C, it is fine to charge anything you like, the
GPL has nothing at all to say about software distribution
costs (except for requiring reasonable incremental costs
for source distribution), but it DOES require that the
deriviative software be distributed under the GPL. If C
is unaware of this, it presumably means that this
requirement of the GPL has been violated, and that hence
B has no permission to make a copy for C.

> for a substantial amount of money.
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^
No issue here, anyone can sell GPL'ed software to anyone
else for any amount of money the buyer is willing to pay!


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <7g1sdh$nvo$2@netnews.upenn.edu>,
  hughett@chaplin.med.upenn.edu (Paul Hughett) wrote:
> and C can sue B for fraud.


You mean based on the failure to use the proper licensing
terms. Certainly there is no fraud involved in charging
for the software even if B got it free. This is not only
an acceptable occurrence, it is a *desirable* occurrence
in many situations.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Paul Hughett]

Robert Dewar (robert_dewar@my-dejanews.com) wrote:
: In article <7g1sdh$nvo$2@netnews.upenn.edu>,
:   hughett@chaplin.med.upenn.edu (Paul Hughett) wrote:
: > and C can sue B for fraud.


: You mean based on the failure to use the proper licensing
: terms. Certainly there is no fraud involved in charging
: for the software even if B got it free. This is not only
: an acceptable occurrence, it is a *desirable* occurrence
: in many situations.

I am assuming that B misrepresented the software as their own
and failed to inform C that C is free to use and/or redistribute
the software under the GPL; then I think C has a pretty good
case for fraud.

If B sells the software in a convenient form for C and/or with support
and hand-holding to get started, WITH correct information about its
provenance and licensing, then you are quite correct; this is both
legal and a good thing.  But that wasn't the situation I understood
from the question.

Paul Hughett

Re: GPL and "free" software

[Tim Smith]

Nick Roberts <nickroberts@callnetuk.com> wrote:
>Suppose party B obtains a work (of software), under the GPL, from party A
>(paying nothing for it).  Then suppose party B modifies this work to form a
>work which they then sell to party C (who is not aware of its GPL heritage)
>for a substantial amount of money.
>
>Now, suppose party C subsequently discovers the fact that the work is
>derived from a GPL work; what can they do?  It seems to me that if party C
>attempts to sue B, B can argue that C is not the injured party, because it
>is not C's licence who's terms were breached.  On the other hand, it seems
>to me that if party A attempts to sue B (on behalf of C), B can argue that A

Assuming party A is the copyright owner (if not, replace A with the owner),
then A does not sue on behalf of C.  A sues for himself for copyright
infringement.

--Tim Smith

Re: GPL and "free" software

[Tim Smith]

In article <gi71g7.gj.ln@joey>,  <spblunt@ozemail.com.au> wrote:
>Well B has done nothing wrong in selling C the software. Neither party can 
>really sue B because B hasn't broken the licence (unless they refuse to 
>give C the source code upon request).

B broke the license by not telling C that the software was GPL'ed.

--Tim Smith

Re: GPL and "free" software

[Robert Dewar]

In article <7g2gqk$ffk$1@52-a-usw.rb1.blv.nwnexus.net>,
  tzs@halcyon.com (Tim Smith) wrote:
> B broke the license by not telling C that the software
> was GPL'ed.


Well you can't exactly "break" a license. What B did was
to make an unauthorized copy for C, since the distribution
to C was not covered by the license. Thus this is a
copyright violation (there is really no such thing as a
license violation -- the GPL allows you to make certain
copies -- any copy outside these allowances is simply a
copyright violation).

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <7g2epv$k11$1@netnews.upenn.edu>,
  hughett@chaplin.med.upenn.edu (Paul Hughett) wrote:
> I am assuming that B misrepresented the software as their
> own and failed to inform C that C is free to use and/or
> redistribute the software under the GPL; then I think C
> has a pretty good case for fraud.

But in fact C is NOT free to further distribute this
software, since C does not hold a license. C has been
granted a license neither by B nor by A, thus C cannot
copy the software at all.

The question of whether C has any standing depends on
how the situation was presented to C. For example, if
there was a message attached saying "this is an
unauthorized copy, but we won't tell if you don't", then
C would have little recourse.

C as the recipient of an undisclosed illegal copy of
software would definitely have recourse against B if
the copy had been represented as legal.




>
> If B sells the software in a convenient form for C and/or
with support
> and hand-holding to get started, WITH correct information
about its
> provenance and licensing, then you are quite correct;
this is both
> legal and a good thing.  But that wasn't the situation I
understood
> from the question.
>
> Paul Hughett
>

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Phil Hunt]

In article <3723c38b@eeyore.callnetuk.com>
           nickroberts@callnetuk.com "Nick Roberts" writes:
> Wading rather trepidly into this issue, and with the understanding that I am
> not any kind of legal expert (in UK or US law), I have wondered if there is
> a potential problem with the GPL and similar licences, which I'll try to
> describe.
> 
> Suppose party B obtains a work (of software), under the GPL, from party A
> (paying nothing for it).  Then suppose party B modifies this work to form a
> work which they then sell to party C (who is not aware of its GPL heritage)
> for a substantial amount of money.
> 
> Now, suppose party C subsequently discovers the fact that the work is
> derived from a GPL work; what can they do?

They can demand that B give them the source.

They cannot sue anyone, unless B told C that they owned the copyright to
all the program they sold (in which case it is fraud). It is perfectly 
legal to distribute GPL'ed software for money.

-- 
Phil Hunt....philh@vision25.demon.co.uk

Re: GPL and "free" software

[Tim Smith]

Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
>   No, C _is_ free to do so.  The GPL is granted to everyone - it's a
>license on the copyright, not a particular copy.  That is C has, in
>fact, been granted a license by A, regardless of what B told them, or
>intends.  At least, it has been granted a license to the original work,
>and copyrights apply to derivative works as well (not necessarily the
>bare modifications, but the derivative work itself, after modification).

Close, but not quite correct.  To distribute a derivative work, you need
permission of both the original copyright owner and the owner of the
copyright on the modifications.  When someone makes a derivative work by
combining works licensed under incompatible terms, what results is a work
that is not distributable at all.

--Tim Smith

Re: GPL and "free" software

[Florian Weimer]

"Nick Roberts" <nickroberts@callnetuk.com> writes:

> Please accept my apologies if this point: has been previously discussed to
> death; is actually trivial, naive, or sophistic; is considered inappropriate
> to comp.lang.ada or gnu.misc.discuss.

I think it has.  AFAIK that's one of the reasons why the FSF requires
a copyright assignment for software which is part of the GNU Project.

Re: GPL and "free" software

[Kenneth P. Turvey]

On 26 Apr 1999 13:03:32 -0700, Tim Smith <tzs@halcyon.com> wrote:
>In article <gi71g7.gj.ln@joey>,  <spblunt@ozemail.com.au> wrote:
>>Well B has done nothing wrong in selling C the software. Neither party can 
>>really sue B because B hasn't broken the licence (unless they refuse to 
>>give C the source code upon request).
>
>B broke the license by not telling C that the software was GPL'ed.
>

It should be noted that `telling' C that the software is GPL'ed could be
simply including the standard COPYING file on the CD.

-- 
Kenneth P. Turvey <kturvey@SprocketShop.com> 

  Democracy is the theory that the common people know what they want and
  deserve to get it good and hard.
	-- H.L. Mencken

Re: GPL and "free" software

[Lynn Winebarger]

In article <7g4ncj$vme$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article
><m2ogkae2db.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de
>> Of course, traditional shrink-wrapped licenses have to
>> include a phrase where you may return the unused software
>> if you do not agree with the license.
>
>I do not know of any such requirement. I quickly checked
>and it is not in the Illinois statute as far as I can see,
>to which shrink-wrapped statute are you referring. It does
>not appear to be in the federal statute either.

   I think he's referring to the "ProCD v. Zeidenberg" decision, where a
shrink-wrap license was upheld partly because the software was
returnable if the purchaser didn't agree to the terms.  (The other part
was that they ruled Zeidenberg actually infringed on copyrighted
material, so the decision is somewhat muddled).

>This is quite wrong, the GPL is no different from any other
>software license, its terms may or may not be acceptable
    This statement is false on its face.  The GPL only deals with
creating derivative works, and redistributing the code.  It doesn't deal
with any particular copy of the software or usage thereof.  It's a
license on the copyrights of the software, not the copy of the software.
    And you don't get the copyrights when you get a copy, i.e. you've
not "purchased" anything.

>to the purchaser of the license. Indeed in the install
>shield for NT, we give people the chance to reject the
>GPL conditions, as seems appropriate. Note that you do not
>usually buy software, you buy the license, and the GPL is
>no different from any other license in this respect. If you
>"buy" GPL'ed software, you have not bought it in the legal
>sense, you have just bought the license.

   No, you do buy the software, though IP lawyers and the software
companies they work for would have you believe otherwise (that is,
unless it was a privately made contract, which just buying something in
a store, as a member of the general public, most certainly does _not_
constitute a privately made contract).
   And yes, when you buy a copy of GPL'ed software, that copy _is_
yours.

Lynn

Re: GPL and "free" software

[Lynn Winebarger]

In article <NllV2.96$jw4.11389@burlma1-snr2>,
Barry Margolin  <barmar@bbnplanet.com> wrote:
>In article <7g2v6u$a3p$1@flotsam.uits.indiana.edu>,
>Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
>>   No, C _is_ free to do so.  The GPL is granted to everyone - it's a
>>license on the copyright, not a particular copy.  That is C has, in
>>fact, been granted a license by A, regardless of what B told them, or
>
>I don't think so.  When C gets the software from B, he also gets the
>license that B specifies, and he's bound by that license.  If B was
>required to distribute the code under the GPL, but didn't, he violated his
>license, and wasn't authorized to distribute the software in the first
>place.  But he isn't automatically assumed to have distributed the source
>with a proper license.

   That's something a judge would decide, I think.  You should take a
look at Chapter 5 of Title 17 (the Copyright act).  The penalties
against an infringer are _very_ harsh.  Not only is the original
copyright holder owed any losses he incurred from the infringers use (in
the case of a GPL author, probably very little), he is also owed all
profits from the infringing activity (which could be a lot).  And that's
the tip of the iceberg.
http://lcweb.loc.gov/copyright/title17/
>
>The GPL has a clause that says that if B's license is terminated because he
>violates the GPL's terms, third parties who have received copies under the
>GPL do not lose their license.  But I don't think that applies to this
>case, since C didn't receive their copy under the GPL.  In other words, it
>says that if C was given the right to make copies, they don't lose them
>because B has lost them; but if C was never given the right to make copies
>(because B distributed under his own, more restrictive license) they don't
>suddenly get them when B's license is terminated.

    But B has no right to unilaterally license derivative works.  In
fact, I checked out the law: (title 17)
-----------------

    - 103. Subject matter of copyright: Compilations and derivative
works

    (a) The subject matter of copyright as specified by section 102
includes compilations and
    derivative works, but protection for a work employing preexisting
material in which copyright
    subsists does not extend to any part of the work in which such
material has been used
    unlawfully. 

    (b) The copyright in a compilation or derivative work extends only
to the material contributed
    by the author of such work, as distinguished from the preexisting
material employed in the
    work, and does not imply any exclusive right in the preexisting
material. The copyright in such
    work is independent of, and does not affect or enlarge the scope,
duration, ownership, or
    subsistence of, any copyright protection in the preexisting
material. 
---------------------------------------------

   The way I read part (a), any modifications to the GPL'ed software
that aren't independent of the GPL'ed software, are either distributed
under the GPL (as an authorized use of the right to derivative works) or
are used unlawfully and thus _do_not_have_copyright_protection.

   I could be wrong, of course, cause the wording is pretty strange.
I'm thinking a demonstrative example of how this rule would apply is if
you published a book where the first half was a revised (derivative)
version of someone else's work (unauthorized), and the second half was
totally original.  It seems to me that the author would not have
copyright protection on the first half, though he would on the second.

   So it reads to me as though, for unlawfully made derivative works,
the original copyright holder retains all the existent copyrights on the
derivative work (since the infringer is denied such rights by part (a),
and no one else can claim rights to derivative works).

Lynn

Re: GPL and "free" software

[Ronald Cole]

Robert Dewar <robert_dewar@my-dejanews.com> writes:
> One very curious case would be the case in which the
> copyright holder, A, gives B a program, ostensibly under
> the GPL, but then refuses to give B the sources.

That could easily be the case if A is a commercial binary-distributor
and binary-distribution recipient, B, let the written offer expire
before asking A for the source, per GPL2-3(b).

-- 
Forte International, P.O. Box 1412, Ridgecrest, CA  93556-1412
Ronald Cole <ronald@forte-intl.com>      Phone: (760) 499-9142
President, CEO                             Fax: (760) 499-9152
My PGP fingerprint: 15 6E C7 91 5F AF 17 C4  24 93 CB 6B EB 38 B5 E5

Re: GPL and "free" software

[Paul Hughett]

Barry Margolin (barmar@bbnplanet.com) wrote:
: In article <7g4jme$5sq$3@netnews.upenn.edu>,
: Paul Hughett <hughett@chaplin.med.upenn.edu> wrote:
: >Paragraph 7 of the GPL (version 2) grants C the right to
: >use/copy/modify/etc the software, even if B made and distributed the
: >copy in violation of the GPL.  C arguably has the same rights
: >with respect to B's modifications of the code, but I would want to do
: >some more study and thinking before I would be really sure of it.

: I don't see that in paragraph 7.  It describes the situation where other
: considerations (such as patent infringement) contradict the GPL, and says
: that the consequence of this is that you can't distribute the program at
: all.

One of us must have a garbled copy or got the wrong section by accident.
The paragraph that I am referring to reads as follows:

7.  Each time you redistribute the Program (or any work based on the
Program), the recipient automatically receives a license from the
original licensor to copy, distribute or modify the Program subject to
these terms and conditions.  You may not impose any further restrictions 
on the recipient's exercise of the rights granted herein.  You are not
responsible for enforcing compliance by third parties to this License.

Okay, not that we are both reading from the same text, do you still
disagree with my interpretation?

Also note that B does not have the right to distribute the software--even
his own modifications--under any license other than the GPL.  That's
why they call the GPL a virus.

Paul Hughett

Re: GPL and "free" software

[Joshua E. Rodd]

Paul Hughett wrote:
> Also note that B does not have the right to distribute the software--even
> his own modifications--under any license other than the GPL.  That's
> why they call the GPL a virus.

B could distribute patches, though, for their modifications and
distribute them under any licence they please. (A standard diff patch
wouldn't be acceptable as it includes lines of context which would make
the patch a derived work; an ed diff would be acceptable, however.)

D (i.e. any one) could get the patches from B, possibly under another
licence, and apply them to the GPL'd code. D (or B) could not redistribute
the patched sources or binaries, however, if the patch were incompatible
with the {L|}GPL.

Two instances which come to mind of people actually doing this are
some patches to Motif (that's all I remember) and the Minix-vmd patches
(before Minix was redistributable at all).

Cheers,
Joshua.

Re: GPL and "free" software

[Lynn Winebarger]

In article <7g2l0f$58g$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>But in fact C is NOT free to further distribute this
>software, since C does not hold a license. C has been
>granted a license neither by B nor by A, thus C cannot
>copy the software at all.

   No, C _is_ free to do so.  The GPL is granted to everyone - it's a
license on the copyright, not a particular copy.  That is C has, in
fact, been granted a license by A, regardless of what B told them, or
intends.  At least, it has been granted a license to the original work,
and copyrights apply to derivative works as well (not necessarily the
bare modifications, but the derivative work itself, after modification).

  Of course, IANAL.

Lynn

Re: GPL and "free" software

[spblunt]

Robert Dewar <robert_dewar@my-dejanews.com> wrote:
> In article <7g2epv$k11$1@netnews.upenn.edu>,
>   hughett@chaplin.med.upenn.edu (Paul Hughett) wrote:
>> I am assuming that B misrepresented the software as their
>> own and failed to inform C that C is free to use and/or
>> redistribute the software under the GPL; then I think C
>> has a pretty good case for fraud.

> But in fact C is NOT free to further distribute this
> software, since C does not hold a license. C has been
> granted a license neither by B nor by A, thus C cannot
> copy the software at all.

That was what I was sort of getting at, they could sue for the right
to modify and redistribute, but they probably couldn't sue for their 
money since B wasn't wrong in charging for the software.

Just reading the GPL again, it says

<quote>
  1. You may copy and distribute verbatim copies of the Program's
source code as you receive it, in any medium, provided that you
conspicuously and appropriately publish on each copy an appropriate
copyright notice and disclaimer of warranty; keep intact all the
notices that refer to this License and to the absence of any warranty;
and give any other recipients of the Program a copy of this License
along with the Program.

You may charge a fee for the physical act of transferring a copy, and
you may at your option offer warranty protection in exchange for a fee.
</quote>

Note "give any other recipients of the program a copy of this Licence along
with the program".  Does that mean you need not tell them about the GPL 
until the money has changed hands?

cya

Re: GPL and "free" software

[Robert Dewar]

In article <slrn7i9ohu.q9b.kturvey@pug1.sprocketshop.com>,
  kturvey@pug1.sprocketshop.com (Kenneth P. Turvey) wrote:

> It should be noted that `telling' C that the software is
> GPL'ed could be simply including the standard COPYING
> file on the CD.

Dubious. You really need a proper copyright statement, so
that you are sure you are receiving a license from the
copyright holder. Normally this would be done in a much
more explicit way. You also need a written offer at least
to provide the sources.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <mh23g7.ajc.ln@joey>,
  <spblunt@ozemail.com.au> wrote:
> Note "give any other recipients of the program a copy of
> this Licence along with the program".  Does that mean you
> need not tell them about the GPL  until the money has
> changed hands?

You can sell a box with a mystery gift in it and give
no information at all if you want, so yes, of course the
answer is yes. But so what? If you are concerned that B
is selling something to C that they have acquired from A
at a cheaper cost, then I would say:

a) this happens all the time, it is what commerce is about!

b) This could happen with either GPL'ed software or
non-GPL'ed software, since there is no connection between
the GPL and the price in $$$.

I think that part of the premise of this thread is that
the significant issue is what B paid to A, but that is
in fact completely irrelevant to this thread.

Yes, C might have paid more than they need to, but that
happens all the time (just look at ebay auctions!), and
there is nothing fraudulent about trying to sell things
for the highest price you can get for them.

If someone auctions a copy of GNAT 3.11p on ebay, and
manages to get someone to pay them $10,000 for it, then
they have located an ebay auction sucker, but there is
no fraud involved!


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <925098459snz@vision25.demon.co.uk>,
  philh@vision25.demon.co.uk wrote:
> They can demand that B give them the source.

Well they can demand it, but if B says no, it is A who
has to do something about it. The GPL does not compel
people to give sources, it just fails to license you to
redistribute sources without the sources, so if B refuses
to give C sources, they are violating A's copyright. C has
no standing in the matter.

It's all fairly straightforward, and of course in practice
these kind of marginal situations rarely arise, so in
practice the details can be left moot without really
affecting anyone significantly.

One very curious case would be the case in which the
copyright holder, A, gives B a program, ostensibly under
the GPL, but then refuses to give B the sources.

The result would be that B could not redistribute the
program since they do not have a valid license to do so,
since they cannot meet the source requirement for the
redistribution. There is no copyright issue here, but
presumably B can sue A for breach of contract (for not
meeting the requirements of the agreed on license vehicle).

I must say I have never heard of anything like the
occurrences discussed here happening.

The only GPL-shady thing I have heard even rumours of at
all is the business of distributing GPL'ed software for
proprietary chips together with a non-disclosure forbidding
disclosure of anything regarding the chip, including any
GPL'ed software that runs on it.


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Phil Hunt]

In article <7g2gqk$ffk$1@52-a-usw.rb1.blv.nwnexus.net>
           tzs@halcyon.com "Tim Smith" writes:
> In article <gi71g7.gj.ln@joey>,  <spblunt@ozemail.com.au> wrote:
> >Well B has done nothing wrong in selling C the software. Neither party can 
> >really sue B because B hasn't broken the licence (unless they refuse to 
> >give C the source code upon request).
> 
> B broke the license by not telling C that the software was GPL'ed.

AFAICT it is perfectly legal for B to to tell C this until after C
has bought it; after all, that is what Linux vendors do with
shrinkwrapped distrutions.

-- 
Phil Hunt....philh@vision25.demon.co.uk

Re: GPL and "free" software

[Lynn Winebarger]

In article <m2ogkae2db.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de>,
David Kastrup  <dak@mailhost.neuroinformatik.ruhr-uni-bochum.de> wrote:
>philh@vision25.demon.co.uk (Phil Hunt) writes:
>
>> AFAICT it is perfectly legal for B to to tell C this until after C
>> has bought it; after all, that is what Linux vendors do with
>> shrinkwrapped distrutions.

   Actually, the whole premise that a software license needs to be on
the outside is sort of bogus, and it doesn't really apply to the GPL
anyway.  Why?  Because, unlike most attempts to inflict a license on a
consumer, the GPL is not a license on a copy of the software, it's a
license on the copyrights themselves.  That is, you don't have to agree
to a contract on a copy of GPL'ed software, so it's really kind of
irrelevant as to whether you're given notice of the GPL or not (as far
as accepting a contract goes).
   Of course, I think that when you purchase a box with a copy of
software inside, you've purchased the copy, not a license, anyway.  It's
not because of the (fallacious) argument that you haven't read the terms
or had an opportunity to negotiate, it's because allowing companies to
sell "licenses" instead of actual products on the open market (as
opposed to an actual privately made contract) amounts to letting
companies get rights against the world, which is clearly the sole domain
of the federal government.  That's all in my non-lawyer opinion, of
course.  

>
>It would probably more complicated if the user acquired a 50-computer
>license for a large price over the 1-computer version, when the offer
>is not accompanied by appropriately more value, such as proportional
>support or media.
>
>In that case we are slowly getting into areas where, well, at least
>you would not buy a second time at that outlet.  Perhaps some sort of
>a court case could be made out of it, but I am not quite sure what.
>
   I was thinking the only real case where something like this would
happen would be where, say, a company makes modifications to GPL'ed
software that are hardware specific to the setup they're selling, then
don't pass along the source to the customers.  I think you'd get an
especially interesting case here, if the customer demanded the source.
One, I think distributing modified versions of GPL'ed software does
signal acceptance, and I think a judge would see it that way too.  The
only reason (that I can think of) that the seller might not have to
release the source code is if it is contractually bound to not do so
(maybe), and even then I tend to think that the sellers would be ordered
to pay damages to the customer for the costs of having to back out of
an installation (obviously, once an institution has widely deployed the
software/hardware combination, a mere refund of the purchase price could
be much less than the cost of switching).  
   
   Of course, what will be interesting is if/when a case actually
occurs.

Lynn

Re: GPL and "free" software

[Paul Hughett]

Robert Dewar (robert_dewar@my-dejanews.com) wrote:
: In article <7g2epv$k11$1@netnews.upenn.edu>,
:   hughett@chaplin.med.upenn.edu (Paul Hughett) wrote:
: > I am assuming that B misrepresented the software as their
: > own and failed to inform C that C is free to use and/or
: > redistribute the software under the GPL; then I think C
: > has a pretty good case for fraud.

: But in fact C is NOT free to further distribute this
: software, since C does not hold a license. C has been
: granted a license neither by B nor by A, thus C cannot
: copy the software at all.

Paragraph 7 of the GPL (version 2) grants C the right to
use/copy/modify/etc the software, even if B made and distributed the
copy in violation of the GPL.  C arguably has the same rights
with respect to B's modifications of the code, but I would want to do
some more study and thinking before I would be really sure of it.

Paul Hughett

Re: GPL and "free" software

[Robert Dewar]

In article
<m2ogkae2db.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de
> Of course, traditional shrink-wrapped licenses have to
> include a phrase where you may return the unused software
> if you do not agree with the license.

I do not know of any such requirement. I quickly checked
and it is not in the Illinois statute as far as I can see,
to which shrink-wrapped statute are you referring. It does
not appear to be in the federal statute either.
>
> This is not the case with GPLed software, as GPLed
> software grants you only additional freedoms over the
> case where you would buy software
> without an explicit license.

This is quite wrong, the GPL is no different from any other
software license, its terms may or may not be acceptable
to the purchaser of the license. Indeed in the install
shield for NT, we give people the chance to reject the
GPL conditions, as seems appropriate. Note that you do not
usually buy software, you buy the license, and the GPL is
no different from any other license in this respect. If you
"buy" GPL'ed software, you have not bought it in the legal
sense, you have just bought the license.


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Barry Margolin]

In article <7g39vc$o8p$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <slrn7i9ohu.q9b.kturvey@pug1.sprocketshop.com>,
>  kturvey@pug1.sprocketshop.com (Kenneth P. Turvey) wrote:
>
>> It should be noted that `telling' C that the software is
>> GPL'ed could be simply including the standard COPYING
>> file on the CD.
>
>Dubious. You really need a proper copyright statement, so
>that you are sure you are receiving a license from the
>copyright holder. Normally this would be done in a much
>more explicit way. You also need a written offer at least
>to provide the sources.

The GPL requires interactive programs to display a copyright notice when
they start up, but where would such a notice would go in other cases
(e.g. GCC, or a library)?  Normally it's in the source code, but if the
source code isn't supplied that obviously is moot.  The COPYING file is
usually the location of the copyright and license for the package as a
whole.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Barry Margolin]

In article <7g2v6u$a3p$1@flotsam.uits.indiana.edu>,
Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
>   No, C _is_ free to do so.  The GPL is granted to everyone - it's a
>license on the copyright, not a particular copy.  That is C has, in
>fact, been granted a license by A, regardless of what B told them, or

I don't think so.  When C gets the software from B, he also gets the
license that B specifies, and he's bound by that license.  If B was
required to distribute the code under the GPL, but didn't, he violated his
license, and wasn't authorized to distribute the software in the first
place.  But he isn't automatically assumed to have distributed the source
with a proper license.

The GPL has a clause that says that if B's license is terminated because he
violates the GPL's terms, third parties who have received copies under the
GPL do not lose their license.  But I don't think that applies to this
case, since C didn't receive their copy under the GPL.  In other words, it
says that if C was given the right to make copies, they don't lose them
because B has lost them; but if C was never given the right to make copies
(because B distributed under his own, more restrictive license) they don't
suddenly get them when B's license is terminated.

>intends.  At least, it has been granted a license to the original work,
>and copyrights apply to derivative works as well (not necessarily the
>bare modifications, but the derivative work itself, after modification).

>  Of course, IANAL.

NAI (Neither Am I).

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Barry Margolin]

In article <7g4jme$5sq$3@netnews.upenn.edu>,
Paul Hughett <hughett@chaplin.med.upenn.edu> wrote:
>Paragraph 7 of the GPL (version 2) grants C the right to
>use/copy/modify/etc the software, even if B made and distributed the
>copy in violation of the GPL.  C arguably has the same rights
>with respect to B's modifications of the code, but I would want to do
>some more study and thinking before I would be really sure of it.

I don't see that in paragraph 7.  It describes the situation where other
considerations (such as patent infringement) contradict the GPL, and says
that the consequence of this is that you can't distribute the program at
all.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[David Kastrup]

philh@vision25.demon.co.uk (Phil Hunt) writes:

> In article <7g2gqk$ffk$1@52-a-usw.rb1.blv.nwnexus.net>
>            tzs@halcyon.com "Tim Smith" writes:
> > In article <gi71g7.gj.ln@joey>,  <spblunt@ozemail.com.au> wrote:
> > >Well B has done nothing wrong in selling C the software. Neither party can 
> > >really sue B because B hasn't broken the licence (unless they refuse to 
> > >give C the source code upon request).
> > 
> > B broke the license by not telling C that the software was GPL'ed.
> 
> AFAICT it is perfectly legal for B to to tell C this until after C
> has bought it; after all, that is what Linux vendors do with
> shrinkwrapped distrutions.

Of course, traditional shrink-wrapped licenses have to include a
phrase where you may return the unused software if you do not agree
with the license.

This is not the case with GPLed software, as GPLed software grants you
only additional freedoms over the case where you would buy software
without an explicit license.

That the recipient of the software may at his choice further
distribute the software for an arbitrary price to whoever he wants is
not in itself to be suing for fraud.

It would probably more complicated if the user acquired a 50-computer
license for a large price over the 1-computer version, when the offer
is not accompanied by appropriately more value, such as proportional
support or media.

In that case we are slowly getting into areas where, well, at least
you would not buy a second time at that outlet.  Perhaps some sort of
a court case could be made out of it, but I am not quite sure what.

-- 
David Kastrup                                     Phone: +49-234-700-5570
Email: dak@neuroinformatik.ruhr-uni-bochum.de       Fax: +49-234-709-4209
Institut f�r Neuroinformatik, Universit�tsstr. 150, 44780 Bochum, Germany

Re: GPL and "free" software

[David Kastrup]

hughett@chaplin.med.upenn.edu (Paul Hughett) writes:

> Robert Dewar (robert_dewar@my-dejanews.com) wrote:
> : In article <7g2epv$k11$1@netnews.upenn.edu>,
> :   hughett@chaplin.med.upenn.edu (Paul Hughett) wrote:
> : > I am assuming that B misrepresented the software as their
> : > own and failed to inform C that C is free to use and/or
> : > redistribute the software under the GPL; then I think C
> : > has a pretty good case for fraud.
> 
> : But in fact C is NOT free to further distribute this
> : software, since C does not hold a license. C has been
> : granted a license neither by B nor by A, thus C cannot
> : copy the software at all.
> 
> Paragraph 7 of the GPL (version 2) grants C the right to
> use/copy/modify/etc the software, even if B made and distributed the
> copy in violation of the GPL.  C arguably has the same rights
> with respect to B's modifications of the code, but I would want to do
> some more study and thinking before I would be really sure of it.

But the GPL in this scenario covered only the transfer of software
from A to B.  C has no right whatsoever to demand that he might do
anything with the software in question if B did not license him to do
this.  The best C can do is tell A about the situation so that A can
sue B for breach of license.  If A is not interested in defending his
license in court, this does not give C any more rights to the stuff he
received from B.

-- 
David Kastrup                                     Phone: +49-234-700-5570
Email: dak@neuroinformatik.ruhr-uni-bochum.de       Fax: +49-234-709-4209
Institut f�r Neuroinformatik, Universit�tsstr. 150, 44780 Bochum, Germany

Re: GPL and "free" software

[Joshua E. Rodd]

Ken Arromdee wrote:
> In article <37264DE6.7AA43E60@noah.dhs.org>,
> Joshua E. Rodd <jerodd@noah.dhs.org> wrote:
> >> Also note that B does not have the right to distribute the software--even
> >> his own modifications--under any license other than the GPL.  That's
> >> why they call the GPL a virus.
> >B could distribute patches, though, for their modifications and
> >distribute them under any licence they please. (A standard diff patch
> >wouldn't be acceptable as it includes lines of context which would make
> >the patch a derived work; an ed diff would be acceptable, however.)
> Not with the FSF's interpretation which says that "user does the link" is
> impermissible.  "User does the patch" would be similar.

Can't be. The licence is not a EULA (which is unenforceable anyway,
at least until 2B goes into effect)--it doesn't limit what you can do
with anything (refer to section 5 of the GPL). So you're free to
mishmash GPL'd code and any other code as much as you like.

A user is free to link code as he or she pleases. This is interesting
with regard to the KDE/Qt situation: a user could freely link KDE with
Qt, but distributing it would be unlicenced IMHO.

[1] Courts will hopefully rule against the offensive provisions of
    the Digital Millenium Copyright Act.

Re: GPL and "free" software

[Barry Margolin]

In article <7g5cb2$bjn$1@netnews.upenn.edu>,
Paul Hughett <hughett@chaplin.med.upenn.edu> wrote:
>Barry Margolin (barmar@bbnplanet.com) wrote:
>: In article <7g4jme$5sq$3@netnews.upenn.edu>,
>: Paul Hughett <hughett@chaplin.med.upenn.edu> wrote:
>: >Paragraph 7 of the GPL (version 2) grants C the right to
>: >use/copy/modify/etc the software, even if B made and distributed the
>: >copy in violation of the GPL.  C arguably has the same rights
>: >with respect to B's modifications of the code, but I would want to do
>: >some more study and thinking before I would be really sure of it.
>
>: I don't see that in paragraph 7.  It describes the situation where other
>: considerations (such as patent infringement) contradict the GPL, and says
>: that the consequence of this is that you can't distribute the program at
>: all.
>
>One of us must have a garbled copy or got the wrong section by accident.
>The paragraph that I am referring to reads as follows:
>
>7.  Each time you redistribute the Program (or any work based on the
>Program), the recipient automatically receives a license from the
>original licensor to copy, distribute or modify the Program subject to
>these terms and conditions.  You may not impose any further restrictions 
>on the recipient's exercise of the rights granted herein.  You are not
>responsible for enforcing compliance by third parties to this License.

In my copy (the version that accompanies GNU Emacs 19.34, labeled GNU
GENERAL PUBLIC LICENSE, Version 2, June 1991) that's paragraph 6.  My
paragraphs are:

1. You may copy and distribute verbatim copies of the Program's
2. You may modify your copy or copies of the Program or any portion
3. You may copy and distribute the Program (or a work based on it,
4. You may not copy, modify, sublicense, or distribute the Program
5. You are not required to accept this License, since you have not
6. Each time you redistribute the Program (or any work based on the
7. If, as a consequence of a court judgment or allegation of patent
8. If the distribution and/or use of the Program is restricted in
9. The Free Software Foundation may publish revised and/or new versions
10. If you wish to incorporate parts of the Program into other free
11. BECAUSE THE PROGRAM IS LICENSED FREE OF CHARGE, THERE IS NO WARRANTY
12. IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING

>Okay, not that we are both reading from the same text, do you still
>disagree with my interpretation?

I think I agree.

>Also note that B does not have the right to distribute the software--even
>his own modifications--under any license other than the GPL.  That's
>why they call the GPL a virus.

You can use other licenses than the GPL, but they have to be compatible --
they can't impose additional restrictions, but they can give additional
freedoms.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Lynn Winebarger]

In article <JWIV2.146$jw4.15391@burlma1-snr2>,
Barry Margolin  <barmar@bbnplanet.com> wrote:
>
>You can use other licenses than the GPL, but they have to be compatible --
>they can't impose additional restrictions, but they can give additional
>freedoms.
>
   Just curious - what more freedoms could they give?

Lynn

Re: GPL and "free" software

[Robert Dewar]

In article <372753FF.11F381CF@noah.dhs.org>,
  "Joshua E. Rodd" <jerodd@noah.dhs.org> wrote:
> Not with the FSF's interpretation which says that "user
> does the link" is impermissible.  "User does the patch"
> would be similar.

I specifically discussed this issue with Richard Stallman
recently. His position is that a patch file, even one
without context, is so dependent on the original as to
be equivalent to a derived work, and he is confident that
a court would agree with this interpretation.

> Can't be

No one can be that definite (unless you are a judge :-).
I would certainly agree with Richard Stallman on this
point, but of course it would require a court decision
to get a definitive view.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Lynn Winebarger]

In article <7g7ro3$q91@www.inetnow.net>, Ken Arromdee <arromdee@tcp.com> wrote:
>In article <372753FF.11F381CF@noah.dhs.org>,
>Joshua E. Rodd <jerodd@noah.dhs.org> wrote:
>>> >B could distribute patches, though, for their modifications and
>>> >distribute them under any licence they please. (A standard diff patch
>>> >wouldn't be acceptable as it includes lines of context which would make
>>> >the patch a derived work; an ed diff would be acceptable, however.)
>>> Not with the FSF's interpretation which says that "user does the link" is
>>> impermissible.  "User does the patch" would be similar.
>>Can't be. The licence is not a EULA (which is unenforceable anyway,
>>at least until 2B goes into effect)--it doesn't limit what you can do
>>with anything (refer to section 5 of the GPL). So you're free to
>>mishmash GPL'd code and any other code as much as you like.
>
>According to the FSF, you might be able to do that, but nobody would be
>allowed to give you code with a stricter license, whose use is to be
>mishmashed together with GPL code.  If someone does that, that is considered
>a disguised way of distributing the combination, and since distributing the
>combination is not allowed, neither is distributing the component.
>
>No, I don't believe it either, but that's the way it's done.

   Well, I think it would be legal for a company to distribute patches
(under a stricter license even), as long as they did not also distribute
the original work.  Not that I would want to encourage that behaviour.

Lynn

Re: GPL and "free" software

[Samuel Mize]

In comp.lang.ada Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
> In article <JWIV2.146$jw4.15391@burlma1-snr2>,
> Barry Margolin  <barmar@bbnplanet.com> wrote:
>>
>>You can use other licenses than the GPL, but they have to be compatible --
>>they can't impose additional restrictions, but they can give additional
>>freedoms.
>>
>    Just curious - what more freedoms could they give?

One example: the freedom to distribute the binary version without
the source.  The distributor's freedom of action is restricted to
ensure the recipient's freedom to see the source.

Best,
Sam

-- 
Samuel Mize -- smize@imagin.net (home email) -- Team Ada
Fight Spam: see http://www.cauce.org/ \\\ Smert Spamonam

Re: GPL and "free" software

[Samuel Mize]

In comp.lang.ada Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
> In article <JWIV2.146$jw4.15391@burlma1-snr2>,
> Barry Margolin  <barmar@bbnplanet.com> wrote:
>>
>>You can use other licenses than the GPL, but they have to be compatible --
>>they can't impose additional restrictions, but they can give additional
>>freedoms.
>>
>    Just curious - what more freedoms could they give?

Please ignore my other reply, if you see it, it was a late-in-the-day
mental bobble.

For example, a license could require you make any changes
available via anonymous FTP.  It could require you to send the
copyright owner a copy of the changes so he can distribute them.

Best,
Sam Mize

-- 
Samuel Mize -- smize@imagin.net (home email) -- Team Ada
Fight Spam: see http://www.cauce.org/ \\\ Smert Spamonam

Re: GPL and "free" software

[Joshua E. Rodd]

Ken Arromdee wrote:
> According to the FSF, you might be able to do that, but nobody would be
> allowed to give you code with a stricter license, whose use is to be
> mishmashed together with GPL code.  If someone does that, that is considered
> a disguised way of distributing the combination, and since distributing the
> combination is not allowed, neither is distributing the component.

You could distribute ed context scripts however you wish.

What you *do* with the ed script is irrelevant.

Hmmm... maybe this could be used to sneak KDE into Debian <grin>.

Re: GPL and "free" software

[Joshua E. Rodd]

Robert Dewar wrote:
> I specifically discussed this issue with Richard Stallman
> recently. His position is that a patch file, even one
> without context, is so dependent on the original as to
> be equivalent to a derived work, and he is confident that
> a court would agree with this interpretation.

This is very true.

> No one can be that definite (unless you are a judge :-).
> I would certainly agree with Richard Stallman on this
> point, but of course it would require a court decision
> to get a definitive view.

Yes. If the patch were being used in a way the clearly
violated the spirit of the GPL, then the court probably
would rule that the licence were violated.

Come to think of it, any patch probably *is* a derived
work as it takes input from the GPL'd code.

<ponders a `clean-room' patch using specialised patch
 software>

Re: GPL and "free" software

[Joshua E. Rodd]

Lynn Winebarger wrote:
> In article <JWIV2.146$jw4.15391@burlma1-snr2>,
> Barry Margolin  <barmar@bbnplanet.com> wrote:
> >You can use other licenses than the GPL, but they have to be compatible --
> >they can't impose additional restrictions, but they can give additional
> >freedoms.
>    Just curious - what more freedoms could they give?

The MIT/X11 licence gives the ultimate freedom: the freedom to do
as you ought, and even as you not ought--you are granted the
freedom to take other's freedom away.

[wondering if this is relevant to comp.long.ada--please tell me
 if it's annoying folks]

Re: GPL and "free" software

[Barry Margolin]

In article <7g8ip8$skl@www.inetnow.net>, Ken Arromdee <arromdee@tcp.com> wrote:
>The way I heard it before, his position was that distributing a patch was
>contributory infringement, since it's only useful when other people get GPL
>software and combine it with the patch.  I didn't buy that version; you
>can't have contributory infringement without direct infringement and getting
>the GPL software is not direct infringement.
>
>The position that the patch itself is a derived work, rather than that the
>patch is contributory infringement which leads people to make derived works,
>seems a lot more defensible to me.  It also seems as scary as the first one,
>since if you patch Windows this now means that your patch is a derived work of
>Windows and Microsoft can sue you if you distribute it....

My take on this is that distributing a work and a program that can
transform that work is equivalent to distributing the derived work that
results from running the program, since the end result is the same -- the
end user has a copy of that derived work.  While it may not violate the
letter of the copyright laws, it certainly violates the spirit.  Why should
it matter whether the patch is applied before or after distributing?

The only thing I can think of that's like this that has been tested in
court is some device that was designed to hook into a video game player and
enhance/alter the games.  I think the game designer lost their suit against
the vendor of this device, but I don't know whether the issues are directly
applicable (I don't recall if they were suing based on copyright, patent,
or whatever).

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Barry Margolin]

In article <wccbtg7o3e2.fsf@world.std.com>,
Robert A Duff  <bobduff@world.std.com> wrote:
>Robert Dewar <robert_dewar@my-dejanews.com> writes:
>
>>... Actually people are always a bit surprised
>> that when you put something in the PD, it is perfectly
>> OK for someone to take it, make a minor change, copyright
>> the result, and sell it as a proprietary program.
>
>Having done so, does this restrict anyone's right to use the original
>work as they see fit?  Or is the person who copyrighted the result
>merely copyrighting the "derived work"?  Or merely the changes?

I believe they're copyrighting the derived work.  Of course, if they find a
copy of the original, they're free to do with t as they wish, so the
copyright notice only protects the changes strongly.  It protects the rest
of the work only insofar as the recipient doesn't look around for
alternatives.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Robert Dewar]

In article <37287742.F8EE7E23@aasaa.ofe.org>,
  David Starner <dstarner98@aasaa.ofe.org> wrote:
> Robert Dewar wrote:
> > They cannot exchange derived versions of gcc, since
> > these would have to be distributed under the GPL
> >*without*
> > any additional restrictions (like non-disclosure
> > agreements).
>
> In practice, does it really matter? Just agree not to
> release it, and
> don't worry about non-disclosure agreements. It's not
> quite as paranoid,
> but it should still work.

In practice this will work fine, but unfortunately lawyers
for certain large companies are unlikely to be happy at
this kind of informal protection of their trade secrets.
Remember that you protect trade secrets for two reasons.

1. To keep things secret

2. To be seen to be taking due diligence in keeping things
secret, so that legally the trade secrets retain their
status as secrets.

I am afraid that the informal passing of GPL'ed material
that you suggest, while it may meet criterion 1, may fail
to meet criterion 2.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Barry Margolin]

In article <3728AAFA.FAD4A827@doc.ic.ac.uk>,
Ed Avis  <epa98@doc.ic.ac.uk> wrote:
>Fraser Wilson wrote:
>
>>Ada libraries are commonly distributed with this additional freedom:
>>
>>-- As a special exception,  if other files  instantiate  generics from this --
>>-- unit, or you link  this unit with other files  to produce an executable, --
>>-- this  unit  does not  by itself cause  the resulting  executable  to  be --
>>-- covered  by the  GNU  General  Public  License.
>
>That seems silly.  Why not just distribute them under the LGPL, which
>is basically the GPL with the above exception built in?

The LGPL also has additional requirements.  You have to distribute the
unlinked object files and the source to the library, so that the recipient
can modify the library and relink.

I believe there are some standard GNU programs that include an exception
similar to the above clause.  For instance, programs produced using Bison
contain a copy of a sizable portion of Bison in the yyparse() function.
Bison was originally distributed under the normal GPL, which meant that
anything produced by it was a derived work, and had to be distributed under
the GPL (or something compatible).  An exception like Ada's was added to
Bison 1.24.

The info file regarding this exception mentions that GCC has always had
such an exception (an object file produced by a compiler includes large
chunks of code taken verbatim out of tables in the code generator).

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Robert Dewar]

In article <3728AAFA.FAD4A827@doc.ic.ac.uk>,
  Ed Avis <epa98@doc.ic.ac.uk> wrote:
> Fraser Wilson wrote:
> That seems silly.  Why not just distribute them under the
> LGPL, which is basically the GPL with the above exception
> built in?
>
> --
> Ed Avis

Ed I assume you are not an Ada folk, and therefore do not
see the critical point of the special language for
generics, something that is not addressed in the LGPL.

In addition, we choose not to use the LGPL quite
deliberately, we find it too restrictive to require
distribution of objects, with the intention of relinking,
and indeed this model of relinking does not really make
sense in the Ada compilation model anyway, where the binder
must check consistency and recompute an elaboration order
for the whole program if anything is changed.

So perhaps you would like to reconsider the "silly" here.
This language was very carefully crafted to meet specific
Ada requirements. The authors (I was one) were completely
aware of the LGPL, and carefully examined it. Note
incidentally that this language is derived from the
language used on the C runtime (which also does not use the
LGPL, but does not of course address the issue of Ada
generics).

Robert Dear

P.S. C++ templates raise similar issues ...


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[David Starner]

Ed Avis wrote:
> 
> Fraser Wilson wrote:
> 
> >Ada libraries are commonly distributed with this additional freedom:
> >
> >-- As a special exception,  if other files  instantiate  generics from this --
> >-- unit, or you link  this unit with other files  to produce an executable, --
> >-- this  unit  does not  by itself cause  the resulting  executable  to  be --
> >-- covered  by the  GNU  General  Public  License.
> 
> That seems silly.  Why not just distribute them under the LGPL, which
> is basically the GPL with the above exception built in?

Because with a template based library, it can be virtually impossible to
sepearte the library and program in the way that the LGPL requires. The
only way to permit relinking would be to include source files for the
program, something not acceptable for proprietary programs.

[Offtopic for cl.ada] Re: GPL and "free" software

[Joshua E. Rodd]

Barry Margolin wrote:
[distributing patches against GPL'd software in a non-freed manner]
> My take on this is that distributing a work and a program that can
> transform that work is equivalent to distributing the derived work that
> results from running the program, since the end result is the same -- the
> end user has a copy of that derived work.  While it may not violate the
> letter of the copyright laws, it certainly violates the spirit.  Why should
> it matter whether the patch is applied before or after distributing?

True. I'm not sure how a court would interpret this (mostly because I
can't predict the future), although I wish I did know <grin>.

> The only thing I can think of that's like this that has been tested in
> court is some device that was designed to hook into a video game player and
> enhance/alter the games.  I think the game designer lost their suit against
> the vendor of this device, but I don't know whether the issues are directly
> applicable (I don't recall if they were suing based on copyright, patent,
> or whatever).

IIRC the case is still going on, and it looks like the examiner of hardware
will not lose the case.

Note that under the new 2B legislation (DMCA), going around an anti-`piracy'
device would be `illegal'. EULAs would also be enforceable. Hopefully, a
judge will declare the law to be the unconstitutional provision it is.

Re: GPL and "free" software

[Robert Dewar]

In article <3728C535.4AFA09C2@doc.ic.ac.uk>,
  Ed Avis <epa98@doc.ic.ac.uk> wrote:
>
> I don't understand what you mean.  Surely the library can be stored in
> a separate shared object or DLL, and loaded just before the program
> starts running, which would be a strict separation of library code
> from program code.
>
> --
> Ed Avis

First of all, this obviously does not address the problem of C++
templates or Ada 95 generics. If this is not clear, it is perhaps
from a lack of familiarity with these features.

Second, for Ada, you can't just relink, you have to rebind, and
recompute the elaboration order, and obviously objects are not
sufficient for this.

All the world does not look like C

(and the LGPL was obviously designed with C in mind :-)


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Lynn Winebarger]

In article <7g9rh5$h5a$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>Suppose two companies want to work together on a super
>secret GCC implementation for a super secret chip (I will
>leave the imaginative reader to fill in appropriate
>details).
>
>They cannot exchange derived versions of gcc, since
>these would have to be distributed under the GPL *without*
>any additional restrictions (like non-disclosure
>agreements).
>
>But if patches were not covered, they could distribute
>patches.
>
>It was specifically in this context that I discussed the
>issue with Richard Stallman, and his view is that this
>attempt to subvert the intention of the GPL is not in
>accordance with the license.
>
   I think this depends.  If the two above companies are A and B, and a
third company - let's call it I, just to be different - develops a
supersecret chip, then company I might put A and B under identical NDA's
about the chip, that has a clause stating that information derived fromt
the documents may only be shared with other companies under an identical
NDA.  Then, in fact, A and B can both share their code and satisfy both
the NDA and the GPL - nothing in the GPL requires that you make your
modifications public, only that you make them available to people you
distribute the software to. (this assumes "I" isn't the one that gives
them the GPL'ed source code to start with)

   On the other hand, if it was A that was developing the chip and it
put B under NDA, it could not give B the source code changes without
either violating the GPL or partially nullifying the NDA (with regards
to any information in the modified source code).
 

   I am a full supporter of the GPL, but I also love these little logic
games.  

Lynn

Re: GPL and "free" software

[Joshua E. Rodd]

Ed Avis wrote:
> I don't understand what you mean.  Surely the library can be stored in
> a separate shared object or DLL, and loaded just before the program
> starts running, which would be a strict separation of library code
> from program code.

<thinks about the related Qt thread>

The answer to this is unclear.

Re: GPL and "free" software

[David Starner]

Ed Avis wrote:
> 
> David Starner wrote:
> 
> [why not distribute libraries under the GPL?]
> 
> >Because with a template based library, it can be virtually impossible to
> >sepearte the library and program in the way that the LGPL requires. The
> >only way to permit relinking would be to include source files for the
> >program, something not acceptable for proprietary programs.
> 
> I don't understand what you mean.  Surely the library can be stored in
> a separate shared object or DLL, and loaded just before the program
> starts running, which would be a strict separation of library code
> from program code.

In C terms, it would be similar to a "library" that was full of macros -
for example:

#define sqrt(a) ({int i; i = a/2; do {a = i/a;} \
	while (a*a - i > .1); i})

You could write most of the C library that way. Now, that would
embed a significant amount of code into the application in a way
that dynamic linking wouldn't help. This is effectively what
C++ templates and Ada generics do, but at a higher level.

If you used LGPL, you would have to ship source code - no other way to
allow someone to change the macros. With the GNAT-licensed code, you
don't have to.

Re: GPL and "free" software

[Joshua E. Rodd]

Lynn Winebarger wrote:
>    I think this depends.  If the two above companies are A and B, and a
> third company - let's call it I, just to be different - develops a
> supersecret chip, then company I might put A and B under identical NDA's
> about the chip, that has a clause stating that information derived fromt
> the documents may only be shared with other companies under an identical
> NDA.  Then, in fact, A and B can both share their code and satisfy both
> the NDA and the GPL - nothing in the GPL requires that you make your
> modifications public, only that you make them available to people you
> distribute the software to. (this assumes "I" isn't the one that gives
> them the GPL'ed source code to start with)

Ah, but let's say that the relationship between A and B soured. A decides
that B has done it a great wrong, and decides that the damage inflicted to
B outweighs the damage inflicted to themself by letting the code out (it
might even benefit A).

B would be hard-pressed to take A to court and hold them to their NDA
about the modifications given the licence under which the original sources
came.

>    On the other hand, if it was A that was developing the chip and it
> put B under NDA, it could not give B the source code changes without
> either violating the GPL or partially nullifying the NDA (with regards
> to any information in the modified source code).

Right--it would take cooperation (which isn't neccessarily a bad thing). 
But A could *not* restrict B's rights to make mods to GPL'd code freed.

>    I am a full supporter of the GPL, but I also love these little logic
> games.

It's fun, isn't it? <grin>

Re: GPL and "free" software

[Aidan Skinner]

On 29 Apr 1999 03:46:08 GMT, Lynn Winebarger
<owinebar@ezinfo.ucs.indiana.edu> wrote: 

>to a text?  That would qualify as "comment" under the fair use doctrine,
>and I think a patch file (particularly a source patch file) could
>reasonably be characterized this way.

A patch file isn't an annotation (those *are* comments ;>), but a
modification based on an origional work. I think it very clearly
violates the spirit of the GPL, though perhaps not the letter. OTOH
this is all meaningless unless and until it's brought up in court.

- Aidan

-- 
"Rotary barrel death stars: for when you absoloutely, positively, have
to obliterate every planet in the system"
http://www.skinner.demon.co.uk/aidan/
http://www.gla.ac.uk/Clubs/WebSoc/~9704075s/

Re: GPL and "free" software

[Lynn Winebarger]

In article <7g8fid$c7c$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <7g7tlc$j5g$1@jetsam.uits.indiana.edu>,
>  owinebar@ezinfo.ucs.indiana.edu (Lynn Winebarger) wrote:
>> Well, I think it would be legal for a company to
>> distribute patches (under a stricter license even), as
>> long as they did not also distribute the original work.
>> Not that I would want to encourage that behaviour.
>>
>> Lynn
>
>
>Lynn, you are expressing legal opinions here without
>giving a basis for the opinion. As I noted in a previous
   Well, I guess it boils down to this question: Can I publish a book
that describes how to physically modify a second book to create a third
book?  While the 3rd book is clearly a derived work, it is not so clear
that the 1st would be - it could be considered fair use.  Indeed, if you
look at the question of whether it hurts the market for the first, you
could only say "no", since performing the modifications requires having
the 2nd book.  There are other considerations, too, of course.  It would
depend on the particular patch file distributed, I think.
   But like I said, I think distributing such a patch would be grounds
for revoking the GPL on the original source, just not a copyright
infringement.  The effect would be that the purchaser would have to
obtain the 2 from different sources - it's not exactly a goldmine.
(The revoking would be because distributing them both _would_ constitute
a derived work - in my opinion, of course).

>post, Richard Stallman *strongly* disagrees with this
>assessment. He takes what seems to be the quite reasonable
>position that a patch file which only has meaning with
>regard to a work A is from a legal point of view a derived
>work, which would be protected. Note that the integrity
>of the GPL depends on this position, so this is not a
>trivial point. 
    Well, once again, it would depend on the particular patch file. 
For example, you can publish a topical index to a book, and that is not
a derived work, but a fair use.  (from "The Fair Use Privilege in
Copyright Law" by William F. Patry).  Or how about a book of annotations
to a text?  That would qualify as "comment" under the fair use doctrine,
and I think a patch file (particularly a source patch file) could
reasonably be characterized this way.
    I also think RMS's stance on the use of function calls from a GPL'ed
shared library is unreasonable (though it's not unreasonable when using
a static linker).  As far as I can see, writing a function call to a
shared library is equivalent to a reference to a scientific paper.
Static linking is equivalent to including it as an appendix.  The second
is clearly an infringement, and the first is not.  

>It has not been specifically adjudicated
>in court, but you should be aware that if you take action
>to follow your legal opinion, you might trigger a court
>to make the decision :-)

  Don't worry, that's not my intention.  I don't have any intention of
writing non-free software (I don't really intend to write non-GPL'ed
software, but I might relent on that in particular circumstances, as
long as the result is still free).

Lynn

Re: GPL and "free" software

[Robert Dewar]

In article <3727B3E3.28755434@noah.dhs.org>,
  "Joshua E. Rodd" <jerodd@noah.dhs.org> wrote:
> The MIT/X11 licence gives the ultimate freedom: the
> freedom to do as you ought, and even as you not
> ought--you are granted the
> freedom to take other's freedom away.

Of course putting something in the public domain has the
same effect. Actually people are always a bit surprised
that when you put something in the PD, it is perfectly
OK for someone to take it, make a minor change, copyright
the result, and sell it as a proprietary program.

They expect that something in the public domain should
belong to the public, i.e. that the public should have
a sort of virtual copyright on PD stuff. But that's not
the way things work, if something is in the PD there is
no copyright.

Indeed one way to characterize the intent of the GPL is to
achieve what many people *think* PD is about. With the GPL,
in a sense, the public does have a copyright interest,
since no one can make deriviative works that they alone
own, and the public is guaranteed full access to the
original work, and to any deriviative works that are
distributed.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <3727B37D.A4A5192E@noah.dhs.org>,
  "Joshua E. Rodd" <jerodd@noah.dhs.org> wrote:
> Come to think of it, any patch probably *is* a derived
> work as it takes input from the GPL'd code.


Note that this is quite an important issue with respect
to the distribution requirements of the GPL.

Suppose two companies want to work together on a super
secret GCC implementation for a super secret chip (I will
leave the imaginative reader to fill in appropriate
details).

They cannot exchange derived versions of gcc, since
these would have to be distributed under the GPL *without*
any additional restrictions (like non-disclosure
agreements).

But if patches were not covered, they could distribute
patches.

It was specifically in this context that I discussed the
issue with Richard Stallman, and his view is that this
attempt to subvert the intention of the GPL is not in
accordance with the license.


-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[David Starner]

Robert Dewar wrote:
> They cannot exchange derived versions of gcc, since
> these would have to be distributed under the GPL *without*
> any additional restrictions (like non-disclosure
> agreements).

In practice, does it really matter? Just agree not to release it, and
don't worry about non-disclosure agreements. It's not quite as paranoid,
but it should still work.

Re: GPL and "free" software

[Robert A Duff]

Robert Dewar <robert_dewar@my-dejanews.com> writes:

>... Actually people are always a bit surprised
> that when you put something in the PD, it is perfectly
> OK for someone to take it, make a minor change, copyright
> the result, and sell it as a proprietary program.

Having done so, does this restrict anyone's right to use the original
work as they see fit?  Or is the person who copyrighted the result
merely copyrighting the "derived work"?  Or merely the changes?

- Bob
-- 
Change robert to bob to get my real email address.  Sorry.

Re: GPL and "free" software

[Paul Hughett]

Lynn Winebarger (owinebar@ezinfo.ucs.indiana.edu) wrote:
: In article <JWIV2.146$jw4.15391@burlma1-snr2>,
: Barry Margolin  <barmar@bbnplanet.com> wrote:
: >
: >You can use other licenses than the GPL, but they have to be compatible --
: >they can't impose additional restrictions, but they can give additional
: >freedoms.
: >
:    Just curious - what more freedoms could they give?

The freedom to use the code in a proprietary product for which the source
is not revealed.  The BSD and MIT X licenses provide this freedom and are
preferred by some developers for that reason.  There is a periodic flamewar
in this group as to which license is better or more moral.

Paul Hughett

Re: GPL and "free" software

[Fraser Wilson]

paene lacrimavi postquam Lynn Winebarger scripsit:

>In article <JWIV2.146$jw4.15391@burlma1-snr2>,
>Barry Margolin  <barmar@bbnplanet.com> wrote:

>>You can use other licenses than the GPL, but they have to be compatible --
>>they can't impose additional restrictions, but they can give additional
>>freedoms.

>   Just curious - what more freedoms could they give?

Ada libraries are commonly distributed with this additional freedom:

-- As a special exception,  if other files  instantiate  generics from this --
-- unit, or you link  this unit with other files  to produce an executable, --
-- this  unit  does not  by itself cause  the resulting  executable  to  be --
-- covered  by the  GNU  General  Public  License.  This exception does not --
-- however invalidate  any other reasons why  the executable file  might be --
-- covered by the  GNU Public License.                                      --

Personally, I'm still wavering between free software purist and pragmatist,
so I have no idea whether I like it or not.

Fraser.

Re: GPL and "free" software

[Robert Dewar]

In article <7g80tn$pbt@news1.newsguy.com>,
  Samuel Mize <smize@imagin.net> wrote:
> For example, a license could require you make any changes
> available via anonymous FTP.  It could require you to
> send the copyright owner a copy of the changes so he can
> distribute them.

These are additional restrictions, and are clearly
incompatible with the GPL, which does not allow additional
restrictions to be added.

Indeed the second, the requirement to send the copyright
owner a copy, is quite an unacceptable requirement, which
would make such a license fail to qualify as free software.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <7g7tlc$j5g$1@jetsam.uits.indiana.edu>,
  owinebar@ezinfo.ucs.indiana.edu (Lynn Winebarger) wrote:
> Well, I think it would be legal for a company to
> distribute patches (under a stricter license even), as
> long as they did not also distribute the original work.
> Not that I would want to encourage that behaviour.
>
> Lynn


Lynn, you are expressing legal opinions here without
giving a basis for the opinion. As I noted in a previous
post, Richard Stallman *strongly* disagrees with this
assessment. He takes what seems to be the quite reasonable
position that a patch file which only has meaning with
regard to a work A is from a legal point of view a derived
work, which would be protected. Note that the integrity
of the GPL depends on this position, so this is not a
trivial point. It has not been specifically adjudicated
in court, but you should be aware that if you take action
to follow your legal opinion, you might trigger a court
to make the decision :-)

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Ed Avis]

Fraser Wilson wrote:

>Ada libraries are commonly distributed with this additional freedom:
>
>-- As a special exception,  if other files  instantiate  generics from this --
>-- unit, or you link  this unit with other files  to produce an executable, --
>-- this  unit  does not  by itself cause  the resulting  executable  to  be --
>-- covered  by the  GNU  General  Public  License.

That seems silly.  Why not just distribute them under the LGPL, which
is basically the GPL with the above exception built in?

-- 
Ed Avis

Re: GPL and "free" software

[Ed Avis]

David Starner wrote:

[why not distribute libraries under the GPL?]

>Because with a template based library, it can be virtually impossible to
>sepearte the library and program in the way that the LGPL requires. The
>only way to permit relinking would be to include source files for the
>program, something not acceptable for proprietary programs.

I don't understand what you mean.  Surely the library can be stored in
a separate shared object or DLL, and loaded just before the program
starts running, which would be a strict separation of library code
from program code.

-- 
Ed Avis

Re: GPL and "free" software

[Paul Hughett]

Barry Margolin (barmar@bbnplanet.com) wrote:

: My take on this is that distributing a work and a program that can
: transform that work is equivalent to distributing the derived work that
: results from running the program, since the end result is the same -- the
: end user has a copy of that derived work.  While it may not violate the
: letter of the copyright laws, it certainly violates the spirit.  Why should
: it matter whether the patch is applied before or after distributing?

: The only thing I can think of that's like this that has been tested in
: court is some device that was designed to hook into a video game player and
: enhance/alter the games.  I think the game designer lost their suit against
: the vendor of this device, but I don't know whether the issues are directly
: applicable (I don't recall if they were suing based on copyright, patent,
: or whatever).

The case is Lewis Galoob Toys, Inc. v. Nintendo of America, Inc.  The
Ninth Circuit Court of Appeals upheld the claim that Galoob's
enhancement device did not infringe on Nintendo's game.

Paul Hughett

Re: GPL and "free" software

[Robert Dewar]

In article <7gce89$94q$2@netnews.upenn.edu>,
  hughett@chaplin.med.upenn.edu (Paul Hughett) wrote:
> The case is Lewis Galoob Toys, Inc. v. Nintendo of America, Inc.  The
> Ninth Circuit Court of Appeals upheld the claim that Galoob's
> enhancement device did not infringe on Nintendo's game.

On the game? or on the patent? I had (perhaps mis)remembered this as
a patent case. I do not recall a claim that the Galoob device was a
deriviative work which violated the copyright (which is what would
be relevant to the thread of discussion here).

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Barry Margolin]

In article <3729D0F6.3DF2CCED@doc.ic.ac.uk>,
Ed Avis  <epa98@doc.ic.ac.uk> wrote:
>Surely not!  I'm sure that templates are implemented in a slightly
>more sensible way than by cutting and pasting object code - or is this
>not what you meant?.  Can't you write a template class in C++, and put
>it in a dynamically linked library?  I know that MFC, for example,
>includes template classes and yet is in a DLL.

Template definitions are in header files, not libraries.  They often work
at the source level, and are more like an elaborate macro facility (they
were actually added to the language to replace some common macro idioms).

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Lynn Winebarger]

In article <3728F132.116D44D1@noah.dhs.org>,
Joshua E. Rodd <jerodd@noah.dhs.org> wrote:
>Lynn Winebarger wrote:
>>    I think this depends.  If the two above companies are A and B, and a
>> third company - let's call it I, just to be different - develops a
>> supersecret chip, then company I might put A and B under identical NDA's
>> about the chip, that has a clause stating that information derived fromt
>> the documents may only be shared with other companies under an identical
>> NDA.  Then, in fact, A and B can both share their code and satisfy both
>> the NDA and the GPL - nothing in the GPL requires that you make your
>> modifications public, only that you make them available to people you
>> distribute the software to. (this assumes "I" isn't the one that gives
>> them the GPL'ed source code to start with)
>
>Ah, but let's say that the relationship between A and B soured. A decides
>that B has done it a great wrong, and decides that the damage inflicted to
>B outweighs the damage inflicted to themself by letting the code out (it
>might even benefit A).
>
>B would be hard-pressed to take A to court and hold them to their NDA
>about the modifications given the licence under which the original sources
>came.
>
    Oh, yeah.  I'm assuming the only worry was about breaking the NDA,
not about the modification being freely distributed.  I'm obviously
referring to situation like the one where Intel provides the docs on
their new chips to Linux developers, who share code amongst each other
and will be released as free eventually anyway.  I wasn't referring to
one were "I" was some sort of shadow company made for the devious
purpose of restraining free code.  That would be different.
   Of course, A is still bound by its contract with "I."  Unless there's
some legal way to show it was only a devious scheme and really
equivalent to B's being the source of the NDA, rather than their being
independent entities.  I haven't taken the time to do any reading on
that subject though.

Lynn

Re: GPL and "free" software

[spblunt]

Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
> In article <JWIV2.146$jw4.15391@burlma1-snr2>,
> Barry Margolin  <barmar@bbnplanet.com> wrote:
>>
>>You can use other licenses than the GPL, but they have to be compatible --
>>they can't impose additional restrictions, but they can give additional
>>freedoms.
>>
>    Just curious - what more freedoms could they give?

for example the xfree licence allows you to link with proprietary software,
but this licence is still GPL compatable.

cya

Re: GPL and "free" software

[David Kastrup]

<spblunt@ozemail.com.au> writes:

> Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
> > In article <JWIV2.146$jw4.15391@burlma1-snr2>,
> > Barry Margolin  <barmar@bbnplanet.com> wrote:
> >>
> >>You can use other licenses than the GPL, but they have to be compatible --
> >>they can't impose additional restrictions, but they can give additional
> >>freedoms.
> >>
> >    Just curious - what more freedoms could they give?
> 
> for example the xfree licence allows you to link with proprietary software,
> but this licence is still GPL compatable.

When distributed as part of a GPLed work, they cannot grant linking to
proprietary software.  So they cannot give additional freedoms to a
GPLed work linked with them.


-- 
David Kastrup                                     Phone: +49-234-700-5570
Email: dak@neuroinformatik.ruhr-uni-bochum.de       Fax: +49-234-709-4209
Institut f�r Neuroinformatik, Universit�tsstr. 150, 44780 Bochum, Germany

Re: GPL and "free" software

[Ed Avis]

David Starner wrote:

>In C terms, it would be similar to a "library" that was full of macros -
>for example:
>
>#define sqrt(a) ({int i; i = a/2; do {a = i/a;} \
>        while (a*a - i > .1); i})
>
>You could write most of the C library that way. Now, that would
>embed a significant amount of code into the application in a way
>that dynamic linking wouldn't help. This is effectively what
>C++ templates and Ada generics do, but at a higher level.

Surely not!  I'm sure that templates are implemented in a slightly
more sensible way than by cutting and pasting object code - or is this
not what you meant?.  Can't you write a template class in C++, and put
it in a dynamically linked library?  I know that MFC, for example,
includes template classes and yet is in a DLL.

-- 
Ed Avis

Re: GPL and "free" software

[Ed Avis]

Robert Dewar wrote:

>Ed I assume you are not an Ada folk, and therefore do not
>see the critical point of the special language for
>generics, something that is not addressed in the LGPL.

No, I've never used Ada.  I have used C++ templates though.  Could you
explain what the problem is, or is it too complex for the uninitiated?

>So perhaps you would like to reconsider the "silly" here.

Yes, I have already.  I said it 'seems silly', which it did.  It
doesn't seem silly now I know there are good reasons.

-- 
Ed Avis

Re: GPL and "free" software

[Ed Avis]

Barry Margolin wrote:

>Template definitions are in header files, not libraries.  They often work
>at the source level, and are more like an elaborate macro facility (they
>were actually added to the language to replace some common macro idioms).

I didn't know that any work compiled using header files counts as a
derivative work of those header files.  People write programs using
(say) Microsoft's header files for MFC, and yet they don't have to say
'portions Copyright Microsoft' for the finished work.

-- 
Ed Avis

Re: GPL and "free" software

[Lynn Winebarger]

In article <372ADFBD.4F4C241B@doc.ic.ac.uk>,
Ed Avis  <epa98@doc.ic.ac.uk> wrote:
>Lynn Winebarger wrote:
>
>>Just think how MS (or other proprietary OS'es) could legally limit the
>>authorship of programs written to be run under their OS.
>
>As part of their consent decree settlment with the DoJ a few years
>ago, MS are not allowed to limit programs compiled with MS tools to
>run only on Windows.  Well, not using licence agreements, anyway.
>
   I wasn't referring to their compiler, but rather to the notion that
making dynamic calls to a library (and the kernel of an OS can be seen
in this way - heck, the Amiga's OS was explicitly written this way)
makes the caller a derivative work.  
   Whether or not MS could do it because of particular consent decrees
is beside the question.  The problem is that with this interpretation of
derivative work, a OS company would be legally in the right to restrict
the authorship of any program that made calls to the OS.  This might
trigger antitrust action, but in MS'es case they owned a huge chunk of
the OS market - other OS manufacturers wouldn't have this consideration
against them.  In any case, I don't think it's beneficial to try and
extend copyright to cover this.

Lynn

Re: GPL and "free" software

[Phil Hunt]

In article <3729D0F6.3DF2CCED@doc.ic.ac.uk> epa98@doc.ic.ac.uk "Ed Avis" writes:
> David Starner wrote:
> >In C terms, it would be similar to a "library" that was full of macros -
> >for example:
> >
> >#define sqrt(a) ({int i; i = a/2; do {a = i/a;} \
> >        while (a*a - i > .1); i})
> >
> >You could write most of the C library that way. Now, that would
> >embed a significant amount of code into the application in a way
> >that dynamic linking wouldn't help. This is effectively what
> >C++ templates and Ada generics do, but at a higher level.

Is the ({ ... }) construct standard C and/or C++ these days? IIRC
gcc supports it.

> Surely not!  I'm sure that templates are implemented in a slightly
> more sensible way than by cutting and pasting object code - or is this
> not what you meant?. 

Effectively this is what templates do. The difference is that templates
allow more type checking than macro definitions.

(Bjarne Stroustrup is on record as saying that he wishes he'd 
initially implemented templates as an extension to #define macros).

> Can't you write a template class in C++, and put
> it in a dynamically linked library?

I would be surprised if this is the case. However instantiations of 
templates can be put in a DLL or .so.

> I know that MFC, for example,
> includes template classes and yet is in a DLL.


-- 
Phil Hunt....philh@vision25.demon.co.uk

Re: GPL and "free" software

[Ed Avis]

Ken Arromdee wrote:

>>As part of their consent decree settlment with the DoJ a few years
>>ago, MS are not allowed to limit programs compiled with MS tools to
>>run only on Windows.  Well, not using licence agreements, anyway.
>
>They're not?  It seems to be a subject that pops up every so often on the
>comp.emulators.ms-windows.wine newsgroup.  Look for the thread "Why you could
>be a software pirate" from the end of 1997.
>
>The IE 4 license says:
>* Installation and Use. Microsoft grants you the right to install and use
>copies of the SOFTWARE PRODUCT on your computers running validly licensed
>copies of the operating system for which the SOFTWARE PRODUCT was designed
>[e.g., Windows(r) 95; Windows NT(r), Windows 3.x, Macintosh, etc.].

I think IE4 is an exception because it is gratis.

>Compilers say the same thing, although I don't have any Microsoft compilers in
>front of me to check.

Compilers?  Ouch!

Maybe it's because the consent decree only covered 'application
programs'.  I'm just guessing though.

-- 
Ed Avis
Advertise here! epa98@doc.ic.ac.uk

Re: GPL and "free" software

[Phil Hunt]

In article <372C1824.1417502F@doc.ic.ac.uk> epa98@doc.ic.ac.uk "Ed Avis" writes:
> Phil Hunt wrote:
> >>Can't you write a template class in C++, and put
> >>it in a dynamically linked library?
> >
> >I would be surprised if this is the case. However instantiations of
> >templates can be put in a DLL or .so.
> 
> What do you mean by 'instantiations of templates'?
> 
> Surely if I write a class like this (pseudocode, I don't know Ada and
> my C++ is rusty):
> 
> class List
>   let T be any class
>   method add(T)
>   method get_first_item : returns T
>   method count_items : returns Int
> 
> then there will only one copy of the executable code in the object
> file.  If the program instantiates a List of Int, then the code for
> List is called and passed pointers (or references) to Ints.  If the
> program wants a List of Char, then the same object code can be used,
> being passed pointers to Chars.
> 
> Or have I got it horribly wrong?

I think you've got it horribly wrong (there might be C++ compilers
that work like that, but I imagine they are very rare, produce slow code
and are horrible to implement).

What C++ does is constructs code for the template with T changed to
int, compiles it, and produces code for the template with T changed
to char, and compiles that as well.

-- 
Phil Hunt....philh@vision25.demon.co.uk

Re: GPL and "free" software

[Ole-Hjalmar Kristensen]

Ed Avis <epa98@doc.ic.ac.uk> writes:

> Phil Hunt wrote:
> 
> >>Can't you write a template class in C++, and put
> >>it in a dynamically linked library?
> >
> >I would be surprised if this is the case. However instantiations of
> >templates can be put in a DLL or .so.
> 
> What do you mean by 'instantiations of templates'?
> 
> Surely if I write a class like this (pseudocode, I don't know Ada and
> my C++ is rusty):
> 
> class List
>   let T be any class
>   method add(T)
>   method get_first_item : returns T
>   method count_items : returns Int
> 
> then there will only one copy of the executable code in the object
> file.  If the program instantiates a List of Int, then the code for
> List is called and passed pointers (or references) to Ints.  If the
> program wants a List of Char, then the same object code can be used,
> being passed pointers to Chars.
> 
> Or have I got it horribly wrong?
> 
> -- 
> Ed Avis
> Advertise here! epa98@doc.ic.ac.uk

In principle, you're right that it could be done this way. However, the
majority of compilers seem to create a separate instance for each T,
just as a macro processor would have done.

Olw-Hj. Kristensen

-- 
E pluribus Unix

Re: GPL and "free" software

[Mitch Blevins]

Barry Margolin <barmar@bbnplanet.com> wrote:
> The GPL requires interactive programs to display a copyright notice when
> they start up, ...

Is this really true?
I can swing a dead cat with a blindfold on (on me, not the cat) and hit
at least 50 interactive GPL'd programs that do NOT do this.

Curious,
-Mitch
mblevin@debian.org

Re: GPL and "free" software

[Andy Isaacson]

In article <925572182snz@vision25.demon.co.uk>, Phil Hunt wrote:
> > >#define sqrt(a) ({int i; i = a/2; do {a = i/a;} \
> > >        while (a*a - i > .1); i})
[snip]
> Is the ({ ... }) construct standard C and/or C++ these days? IIRC
> gcc supports it.

No, it's not standard.  gcc does support it, however.

-andy

Re: GPL and "free" software

[Ed Avis]

Phil Hunt wrote:

>>Can't you write a template class in C++, and put
>>it in a dynamically linked library?
>
>I would be surprised if this is the case. However instantiations of
>templates can be put in a DLL or .so.

What do you mean by 'instantiations of templates'?

Surely if I write a class like this (pseudocode, I don't know Ada and
my C++ is rusty):

class List
  let T be any class
  method add(T)
  method get_first_item : returns T
  method count_items : returns Int

then there will only one copy of the executable code in the object
file.  If the program instantiates a List of Int, then the code for
List is called and passed pointers (or references) to Ints.  If the
program wants a List of Char, then the same object code can be used,
being passed pointers to Chars.

Or have I got it horribly wrong?

-- 
Ed Avis
Advertise here! epa98@doc.ic.ac.uk

Re: GPL and "free" software

[Sam Holden]

On 2 May 1999 12:40:52 GMT, Mitch Blevins <mblevin@debian.org> wrote:
>Barry Margolin <barmar@bbnplanet.com> wrote:
>> The GPL requires interactive programs to display a copyright notice when
>> they start up, ...
>
>Is this really true?
>I can swing a dead cat with a blindfold on (on me, not the cat) and hit
>at least 50 interactive GPL'd programs that do NOT do this.

No, it is suggested as part of the process of applying the terms of the
GPL to your software.

It is not part of the terms and conditions of the GPL.

-- 
Sam

The very fact that it's possible to write messy programs in Perl is also
what makes it possible to write programs that are cleaner in Perl than
they could ever be in a language that attempts to enforce cleanliness.
	--Larry Wall

Re: GPL and "free" software

[Ed Avis]

Phil Hunt wrote:

>>If the program instantiates a List of Int, then the code for
>>List is called and passed pointers (or references) to Ints.  If the
>>program wants a List of Char, then the same object code can be used,
>>being passed pointers to Chars.

>I think you've got it horribly wrong (there might be C++ compilers
>that work like that, but I imagine they are very rare, produce slow code
>and are horrible to implement).
>
>What C++ does is constructs code for the template with T changed to
>int, compiles it, and produces code for the template with T changed
>to char, and compiles that as well.

I see.  That seems a much less efficient way to do things, but what do
I know about compiler design?  I'm sure that not every language
supporting generic types does it this way.  For a start, some are
interpreted.

You've convinced me that a program compiled, and dynamically linked,
against a library containing template classes really does contain bits
of that library embedded in the program code.  I think you could
probably work around this, by creating a thin wrapper:

class List_wrapper
  method add(Object o) (or void *)
    call List.add(o)
  end method
(etc.)

and then substitute List_wrapper for List everywhere in your program. 
The object code for List_wrapper would contain bits of code from List,
but the rest of your program would not, and you could link the two
together at runtime.  You lose the benefit of template type-checking
this way, but AFAIK, it's only used at compile time, not runtime (at
least in C++), so you could develop without the wrapper, then stick it
in for your final release.

-- 
Ed Avis
Advertise here! epa98@doc.ic.ac.uk

Re: GPL and "free" software

[Robert Dewar]

In article <slrn7ip1tk.207.adisaacs@pirx.resnet.mtu.edu>,
  adisaacs@mtu.edu (Andy Isaacson) wrote:
> No, it's not standard.  gcc does support it, however.

It is actually more helpful to use the term GNU C here rather than
gcc, when you are talking about the language. GNU C is a superset of
C containing many useful extensions, including most notably nested
functions.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[dennison]

In article <372CBCA7.52206CB5@doc.ic.ac.uk>,
  Ed Avis <epa98@doc.ic.ac.uk> wrote:
> Phil Hunt wrote:
>
> >What C++ does is constructs code for the template with T changed to
> >int, compiles it, and produces code for the template with T changed
> >to char, and compiles that as well.
>
> You've convinced me that a program compiled, and dynamically linked,
> against a library containing template classes really does contain bits
> of that library embedded in the program code.  I think you could
> probably work around this, by creating a thin wrapper:

Ada has much the same problem with generics as well as the language's run-time
system. Even the LGPL does not prevent GPL "infection" of client code in this
situation.

The way the Gnat compiler got around it was to put an addendum to the GPL in
the headers of the source files in question, saying basicly that the meer act
of compiling and/or linking against that source file is not sufficient to
make the GPL apply to the client.

--
T.E.D.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

[O/T 4 cla] Re: GPL and "free" software

[Joshua E. Rodd]

Leslie Mikesell wrote:
> Now, if you never redistribute any GPL'd material you don't
> need the copyright exception that the GPL gives you, so
> why does it matter if you violate this licence by exchanging
> patches?

The patch is a derived work; you are not licenced to make
copies of the copyrighted code unless you agree to the licence.

Re: GPL and "free" software

[Barry Margolin]

In article <slrn7ioisp.lmv.sholden@pgrad.cs.usyd.edu.au>,
Sam Holden <sholden@cs.usyd.edu.au> wrote:
>On 2 May 1999 12:40:52 GMT, Mitch Blevins <mblevin@debian.org> wrote:
>>Barry Margolin <barmar@bbnplanet.com> wrote:
>>> The GPL requires interactive programs to display a copyright notice when
>>> they start up, ...
>>
>>Is this really true?
>>I can swing a dead cat with a blindfold on (on me, not the cat) and hit
>>at least 50 interactive GPL'd programs that do NOT do this.
>
>No, it is suggested as part of the process of applying the terms of the
>GPL to your software.
>
>It is not part of the terms and conditions of the GPL.

Clause 2c contains the requirement.  However, it only applies to derived
works where the original program had such behavior.  In other words, you
can't take a GPLed program that displays a copyleft/warranty notice when it
starts up, modify it so it doesn't do so, and then redistribute this
version.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Barry Margolin]

In article <7gdln1$vna$1@jetsam.uits.indiana.edu>,
Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
>   I doubt the macros in standard header files (at least the ones I've seen)
>would get much in the way of copyright protection, because they are
>usually very functional in nature.  I agree that templates are a
>different matter.  I think a sufficiently creative peice of macro code
>would get copyright protection, but for string copying macros and the
>like, I have my doubts.

Well, if AT&T could claim copyright on an *empty* shell script (/bin/true),
I expect they'd think that a header file also deserves copyright
protection.  Whether either of these is legally defensible, I can't say
(the copyright notice in /bin/true was presumably the result of someone
mechanically inserting copyright notices in all the shell scripts installed
with the system, rather than someone consciously deciding that this empty
script needs protection).

-- 
Barry Margolin, barmar@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.

Re: [O/T 4 cla] Re: GPL and "free" software

[Joshua E. Rodd]

Ken Arromdee wrote:
> In article <372DCFB9.57F91727@iiinet.dhs.org>,
> Joshua E. Rodd <jerodd@iiinet.dhs.org> wrote:
> >> Now, if you never redistribute any GPL'd material you don't
> >> need the copyright exception that the GPL gives you, so
> >> why does it matter if you violate this licence by exchanging
> >> patches?
> >The patch is a derived work; you are not licenced to make
> >copies of the copyrighted code unless you agree to the licence.
> Do you believe the same thing for patches to Microsoft Windows?

Yes.

Re: GPL and "free" software

[Joshua E. Rodd]

Ed Avis wrote:
> Barry Margolin wrote:
> 
> >When you #include a file,
> >the compiler treats it as if the contents of that file were actually
> >inserted in that place, and the resulting compiled file contains the
> >compiled versions of those statements, which are clearly derivative works
> >of those bits of source code.
> 
> But most of the time, header files don't actually include code.

FD_SET certainly includes code.

> They
> include function prototypes, which aren't strictly necessary (K&C
> managed for years without them), declarations of functions (which are
> just there to reassure the compiler that 'printf()' isn't a typo), and
> constant values such as #define SIGINT 2, which being just a few
> simple numbers, probably aren't copyrightable by themselves.

Huh? One could as well argue that an object file is `just a bunch
of unsigned numbers'. It's still copyrighted.

> Prototypes and declarations don't end up in the final object code -
> they're just 'scaffolding' for the compiler.

They very well could end up in the final code--think about Java.

Re: GPL and "free" software

[Joshua E. Rodd]

David Kastrup wrote:
> But the licenses on many of their own software explicitly prohibit you
> to try running their software on, say, something like Wine.

Ah, but one could still take Intercal Explorer and run it on any
platform you wish--emulator in Bochs, Wine (not that it's technically
feasible), or Win32s--because you've never agreed to a contract to
not run it in Wine.

If you signed a contract with Microsoft that you would not run
Intercal Explorer under Wine, that would be different. But not
signatures are exchanged when retreiving Intercal Explorer 5
beta 1.002.

> It would probably not be enforcible, and of course they are taking
> technical measures also all the time to make sure the programs break
> on anything but Windows, but it *is* there openly in the license.
> Perhaps the DoJ has not noticed?

Microsoft *probably* wouldn't bring a civil suit against someone
running IE in Wine, either.

Re: GPL and "free" software

[Joshua E. Rodd]

Ken Arromdee wrote:
> The IE 4 license says:
> * Installation and Use. Microsoft grants you the right to install and use
> copies of the SOFTWARE PRODUCT on your computers running validly licensed
> copies of the operating system for which the SOFTWARE PRODUCT was designed
> [e.g., Windows(r) 95; Windows NT(r), Windows 3.x, Macintosh, etc.].

That's a bit of stretch--how do you define ``designed''? One could
easily blur the line between Bochs[1] and Wine and Microsoft Windows
(none of this would be unlicenced provided you didn't redistribute
the resulting cornhash of binaries).

[1] For the purposes of this discussion, assume Bochs is licenced under
the GPL or MIT-style licence.

Re: GPL and "free" software

[Larry Kilgallen]

In article <372E3C51.47FA3994@iiinet.dhs.org>, "Joshua E. Rodd" <jerodd@iiinet.dhs.org> writes:

> Microsoft *probably* wouldn't bring a civil suit against someone
> running IE in Wine, either.

Or if they do, there is always a chance they will lose due to
having doctored the videotapes :-)

Larry Kilgallen

Re: GPL and "free" software

[Alan Braggins]

Ed Avis <epa98@doc.ic.ac.uk> writes:
> >When you #include a file,
> >the compiler treats it as if the contents of that file were actually
> >inserted in that place, and the resulting compiled file contains the
> >compiled versions of those statements, which are clearly derivative works
> >of those bits of source code.
> 
> But most of the time, header files don't actually include code.

They do if they include template definitions.

Re: GPL and "free" software

[Ed Avis]

Joshua E. Rodd wrote:

>>But most of the time, header files don't actually include code.
>
>FD_SET certainly includes code.

Okay.  I did say, 'most of the time'.  For the sake of argument, let's
just discuss 'pure' header files that don't contain actual code, (or
maybe the code is so trivial and short as to be not copyrightable).
 
>>They
>>include function prototypes, which aren't strictly necessary (K&C
>>managed for years without them), declarations of functions (which are
>>just there to reassure the compiler that 'printf()' isn't a typo), and
>>constant values such as #define SIGINT 2, which being just a few
>>simple numbers, probably aren't copyrightable by themselves.
>
>Huh? One could as well argue that an object file is `just a bunch
>of unsigned numbers'. It's still copyrighted.

An object file is several thousand rather long integers.  I meant that
something like:

#define BUFFER_SIZE 1000
#define MAX_THINGIES 6

probably isn't copyrightable, especially once you strip away the names
and are just left with the pair of integers '1000, 6'.  You could make
an enormous header file with #define FIRST_INSTRUCTION 0x28a45e92,
#define SECOND_INSTRUCTION 0xf3485dea, and so on, which would be big
enough to copyright.  Of course, IANAL.

>>Prototypes and declarations don't end up in the final object code -
>>they're just 'scaffolding' for the compiler.
>
>They very well could end up in the final code--think about Java.

Java doesn't have separate prototypes or header files.  AFAIK the
information on a method's name, what arguments it takes, and so on,
are stored together with the bytecode for that method - but then, they
came from the same source file anyway.  It's not a case of my object
code getting 'polluted' with copyrighted material from somebody else's
header files.

-- 
Ed Avis
Advertise here! epa98@doc.ic.ac.uk

Re: GPL and "free" software

[Ed Avis]

Joshua E. Rodd wrote:

>>But the licenses on many of their own software explicitly prohibit you
>>to try running their software on, say, something like Wine.

>If you signed a contract with Microsoft that you would not run
>Intercal Explorer under Wine, that would be different. But not
>signatures are exchanged when retreiving Intercal Explorer 5
>beta 1.002.

>Microsoft *probably* wouldn't bring a civil suit against someone
>running IE in Wine, either.

I think you are missing the point here.  Firstly, laws such as the
DMCA will make 'click-through' licences legally binding.  Secondly,
even though the agreement may not be legally binding, there is still
dishonour involved in breaking it.  (Not to say that I don't break
licence agreements myself, of course.)  Thirdly, most companies and
institutions take care to comply with licence agreements on software;
whether they do this out of charity towards Microsoft, or because
their lawyers have warned them that the EULAs do have some legal
force, I don't know.  But the point is, just because a few people on
gnu.misc.discuss may claim that EULAs are not legally binding, that
doesn't mean that the rest of the world will suddenly realize their
mistakes and start breaking licences with impunity.  For many people,
Microsoft's restrictions on where you can run IE _do_ need to be
complied with, whether or not they are actually legally valid.

-- 
Ed Avis
Advertise here! epa98@doc.ic.ac.uk

Re: GPL and "free" software

[Barry Margolin]

In article <372F2C65.4B7663C5@doc.ic.ac.uk>,
Ed Avis  <epa98@doc.ic.ac.uk> wrote:
>An object file is several thousand rather long integers.  I meant that
>something like:
>
>#define BUFFER_SIZE 1000
>#define MAX_THINGIES 6
>
>probably isn't copyrightable, especially once you strip away the names
>and are just left with the pair of integers '1000, 6'.  You could make
>an enormous header file with #define FIRST_INSTRUCTION 0x28a45e92,
>#define SECOND_INSTRUCTION 0xf3485dea, and so on, which would be big
>enough to copyright.  Of course, IANAL.

It certainly is copyrightable.  The choice of names is a significant part
of programming.  A header file that says:

#define ASDFASDF 1000
#define LKJLKJLK 6

is quite different from the above file.

IANAL, so I don't know how much punishment could be exacted for infringing
the copyright of such a small amount of code, but I think it's *still*
copyright infringement, just as taking a 50-cent candy bar is still theft.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Ed Avis]

Barry Margolin wrote:

>>I meant that something like:
>>
>>#define BUFFER_SIZE 1000
>>#define MAX_THINGIES 6
>>
>>probably isn't copyrightable, especially once you strip away the names
>>and are just left with the pair of integers '1000, 6'.

>It certainly is copyrightable.  The choice of names is a significant part
>of programming.  A header file that says:
>
>#define ASDFASDF 1000
>#define LKJLKJLK 6
>
>is quite different from the above file.

My point was, the only part of the header file which gets included in
the final executable (or, indeed, in the .S file) is the two numbers
1000 and 6.  So my header file and yours would give identical object
code.  I don't see how the number six can be copyrighted, except
perhaps in relation to the Children's Television Workshop.

>IANAL, so I don't know how much punishment could be exacted for infringing
>the copyright of such a small amount of code, but I think it's *still*
>copyright infringement, just as taking a 50-cent candy bar is still theft.

I thought that very small snippets of code were not copyrighted.

-- 
Ed Avis
Advertise here! epa98@doc.ic.ac.uk

Re: GPL and "free" software

[Peter Seebach]

In article <372F36AB.55833EA6@doc.ic.ac.uk>,
Ed Avis  <epa98@doc.ic.ac.uk> wrote:
>My point was, the only part of the header file which gets included in
>the final executable (or, indeed, in the .S file) is the two numbers
>1000 and 6.

Hmm.  Interesting point...  Not sure it's always true; some compilers
embed lots of information for debuggers.

>I thought that very small snippets of code were not copyrighted.

If this were true, you could probably "fair-use" a large piece of code
to death.

-s
-- 
Copyright 1999, All rights reserved.  Peter Seebach / seebs@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!

Re: GPL and "free" software

[Barry Margolin]

In article <372F36AB.55833EA6@doc.ic.ac.uk>,
Ed Avis  <epa98@doc.ic.ac.uk> wrote:
>Barry Margolin wrote:
>
>>>I meant that something like:
>>>
>>>#define BUFFER_SIZE 1000
>>>#define MAX_THINGIES 6
>>>
>>>probably isn't copyrightable, especially once you strip away the names
>>>and are just left with the pair of integers '1000, 6'.
>
>>It certainly is copyrightable.  The choice of names is a significant part
>>of programming.  A header file that says:
>>
>>#define ASDFASDF 1000
>>#define LKJLKJLK 6
>>
>>is quite different from the above file.
>
>My point was, the only part of the header file which gets included in
>the final executable (or, indeed, in the .S file) is the two numbers
>1000 and 6.  So my header file and yours would give identical object
>code.  I don't see how the number six can be copyrighted, except
>perhaps in relation to the Children's Television Workshop.

Suppose you have a program that contains:

#include <somefile.h>
...
printf ("%d, %d\n", BUFFER_SIZE, MAX_THINGIES);

If you compile this file, you'll get a different object file depending on
whether somefile.h contains:

#define BUFFER_SIZE 1000
#define MAX_THINGIES 6

or

#define BUFFER_SIZE 6
#define MAX_THINGIES 1000

Thus, the object file is derived from the header file.

>I thought that very small snippets of code were not copyrighted.

They are, but in some cases copying small snippets out of a larger work
might be fair use, and hence not infringing.  There are a number of
criteria that must be satisfied for a copy to be fair use, and size in
proportion to the entire work is just one of them.

Perhaps a court might decide that the way that header files are typically
used constitutes fair use.  AFAIK, it's never been tested.

Years ago I worked for a company that produced an operating system and
compilers.  It was quite common for programmers at the time to produce
compiler listings that included all the header files inserted.  IIRC, we
took the copyright notices *out* of our header files so that these listing
files wouldn't appear to be copyrighted by the vendor.  Our PL/I compiler
didn't support macros, so the only stuff in the header files were function
prototypes and structure declarations, and we weren't concerned about them
being copied.

-- 
Barry Margolin, barmar@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.

Re: [O/T 4 cla] Re: GPL and "free" software

[Leslie Mikesell]

In article <7gnb7u$hlu@www.inetnow.net>, Ken Arromdee <arromdee@tcp.com> wrote:
>In article <372E2E00.99917C63@iiinet.dhs.org>,
>Joshua E. Rodd <jerodd@iiinet.dhs.org> wrote:
>>> >> Now, if you never redistribute any GPL'd material you don't
>>> >> need the copyright exception that the GPL gives you, so
>>> >> why does it matter if you violate this licence by exchanging
>>> >> patches?
>>> >The patch is a derived work; you are not licenced to make
>>> >copies of the copyrighted code unless you agree to the licence.
>>> Do you believe the same thing for patches to Microsoft Windows?
>>Yes.
>
>You mean you think people should be sued for distributing unauthorized patches
>to Windows, even if the patch doesn't use any Windows code?

There is a rather long tradition of user groups sharing their patches
to commercial works.  It is usually encouraged rather than contested
as long as the participants all have their own copies of the base
code.  

 Les Mikesell
   les@mcs.com

Re: GPL and "free" software

[Nick Roberts]

Very briefly, C++ templates and Ada generics provide a way to encode an
algorithm only once, if it is the kind of algorithm which can be applied
unchanged to several variations of things that are basically 'static' in
languages such as C++ and Ada (types, functions/subprograms, etc.).  Each
such variation is called an 'instance', and setting up an instance ready for
use is called 'instantiation'.  C++ templates and Ada generic units work in
fundamentally the same way; the biggest difference being that, in C++,
templates are instantiated automatically.

The usual (default) implementation is for each instantiation of a template
or generic unit to cause the creation of a new piece of object code, varied
according to the template/generic parameters.  (This generally gives faster
execution speed, but at the cost of the overall program's object code being
bigger.)

But, some implementations (optionally) produce a single piece of object
code, which calls 'hooks' inside it to provide the variation according to
the different template/generic parameters; each instantiation then simply
generates a set of appropriate hooks.  A pointer to one of these sets of
hooks must be passed in on every call into the template/generic.  (This
generally gives slower speed, but reduces the size of the overall program's
object code.)

There are other schemes which provide a compromise between these two
extremes.

I think the relevant point is that, in any eventuality, if a program uses a
template or generic unit, then some object code corresponding to that
template or generic unit will end up in the program's object code.  I feel
sure it is the case that the source code for many templates and generic
units will constitute an 'original and substantive work' susceptible to a
claim of copyright.

-------------------------------------
Nick Roberts
-------------------------------------

Re: [O/T 4 cla] Re: GPL and "free" software

[James Youngman]

"Joshua E. Rodd" <jerodd@iiinet.dhs.org> writes:

> Leslie Mikesell wrote:
> > Now, if you never redistribute any GPL'd material you don't
> > need the copyright exception that the GPL gives you, so
> > why does it matter if you violate this licence by exchanging
> > patches?
> 
> The patch is a derived work; you are not licenced to make
> copies of the copyrighted code unless you agree to the licence.

What about "diff -e"?

-- 
ACTUALLY reachable as @free-lunch.demon.(whitehouse)co.uk:james+usenet

Re: [O/T 4 cla] Re: GPL and "free" software

[James Youngman]

arromdee@www.inetnow.net (Ken Arromdee) writes:

> In article <372DCFB9.57F91727@iiinet.dhs.org>,
> Joshua E. Rodd <jerodd@iiinet.dhs.org> wrote:
> >> Now, if you never redistribute any GPL'd material you don't
> >> need the copyright exception that the GPL gives you, so
> >> why does it matter if you violate this licence by exchanging
> >> patches?
> >The patch is a derived work; you are not licenced to make
> >copies of the copyrighted code unless you agree to the licence.
> 
> Do you believe the same thing for patches to Microsoft Windows?

The license which Mr Rodd is referring to is the license which allows
you legally to distribute derived works.  Unless you agree to the GPL,
you are not allowed to distribute derived works. 

In the case of Windows, the license is an end-user license, not a
license to distribute derived works.  While the GPL seeks to provide
rights, the Windows EULA seeks to remove them.   The two situations
are not comparable.

While the Windows EULA says "by using this software....", the GPL says
"By copying this software...", which is quite different.  You are
allowed to *use* but not *copy* GPLed software without taking on even
a single one of the obligations mentioned in the GPL.

-- 
ACTUALLY reachable as @free-lunch.demon.(whitehouse)co.uk:james+usenet

Re: [O/T 4 cla] Re: GPL and "free" software

[Leslie Mikesell]

In article <7gprnj$pau$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <7goi9s$qvq$1@Mercury.mcs.net>,
>  les@MCS.COM (Leslie Mikesell) wrote:
>
>> There is a rather long tradition of user groups sharing their patches
>> to commercial works.  It is usually encouraged rather than contested
>> as long as the participants all have their own copies of the base
>> code.
>
>This is potentially very misleading, I am aware of a major lawsuit
>where the EXACT point of contention is that such sharing of patches
>was a violation of the license agreement. You cannot generalize
>in this manner.

There are lawsuits over nearly everything.  Is there any court
decision that states that something you write is covered by
someone else's copyright when it contains none of the other
work?  Note that this has nothing to do with license agreements
since you can obtain GPL'd works without agreeing to the license,
unless you are redistributing the GPL'd code this is strictly
a copyright issue.  It might not be in a different situation
where you had agreed to a license that prohibited reverse
engineering or modifications.  How about an approach where you hire
the person who did the modification to make the same modification
for you as your agent for some token sum?  

  Les Mikesell
   les@mcs.com

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <7goi9s$qvq$1@Mercury.mcs.net>,
  les@MCS.COM (Leslie Mikesell) wrote:

> There is a rather long tradition of user groups sharing their patches
> to commercial works.  It is usually encouraged rather than contested
> as long as the participants all have their own copies of the base
> code.

This is potentially very misleading, I am aware of a major lawsuit
where the EXACT point of contention is that such sharing of patches
was a violation of the license agreement. You cannot generalize
in this manner.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Richard E. Hawkins Esq.]

In article <7g4ncj$vme$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article
><m2ogkae2db.fsf@mailhost.neuroinformatik.ruhr-uni-bochum.de
>> Of course, traditional shrink-wrapped licenses have to
>> include a phrase where you may return the unused software
>> if you do not agree with the license.

>I do not know of any such requirement. I quickly checked
>and it is not in the Illinois statute as far as I can see,
>to which shrink-wrapped statute are you referring. It does
>not appear to be in the federal statute either.

You will not find the overwhelming majority of contract law in 
any statutes; it is heavily common law.  As a practical matter,
if there is no such condition clear before the shrinkwrap is
cracked, the purported license terms would not attach.

rick, esq.
-- 
These opinions will not be those of ISU until it pays my retainer.

Re: GPL and "free" software

[Richard E. Hawkins Esq.]

In article <7gaiof$sjj$1@flotsam.uits.indiana.edu>,
Lynn Winebarger <owinebar@ezinfo.ucs.indiana.edu> wrote:
>In article <7g9rh5$h5a$1@nnrp1.dejanews.com>,
>Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>>Suppose two companies want to work together on a super
>>secret GCC implementation for a super secret chip (I will
>>leave the imaginative reader to fill in appropriate
>>details).

>>They cannot exchange derived versions of gcc, since
>>these would have to be distributed under the GPL *without*
>>any additional restrictions (like non-disclosure
>>agreements).
...
>>It was specifically in this context that I discussed the
>>issue with Richard Stallman, and his view is that this
>>attempt to subvert the intention of the GPL is not in
>>accordance with the license.

>   I think this depends.  If the two above companies are A and B, and a
>third company - let's call it I, just to be different - develops a
>supersecret chip, then company I might put A and B under identical NDA's
>about the chip, that has a clause stating that information derived fromt
>the documents may only be shared with other companies under an identical
>NDA.  Then, in fact, A and B can both share their code and satisfy both
>the NDA and the GPL - nothing in the GPL requires that you make your
>modifications public, only that you make them available to people you
>distribute the software to. (this assumes "I" isn't the one that gives
>them the GPL'ed source code to start with)

However, under this chip development scenario, the licensee of the GPL
is the conglomerate/joint enterprise/whatever of A, B, and I, 
complete with their NDA's, which can internally distribute it's 
modified version.  There is no subversion here; it's simply a question
of "who is the licensee."
-- 
These opinions will not be those of ISU until it pays my retainer.

Re: GPL and "free" software

[Robert Dewar]

In article <7gpqms$7qj$1@eyry.econ>,
  hawk@eyry.econ.iastate.edu (Richard E. Hawkins Esq.) wrote:

> However, under this chip development scenario, the licensee of the GPL
> is the conglomerate/joint enterprise/whatever of A, B, and I,
> complete with their NDA's, which can internally distribute it's
> modified version.  There is no subversion here; it's simply a question
> of "who is the licensee."
> --
> These opinions will not be those of ISU until it pays my retainer.


That does not work, unless there is a legal entity involved, not at all
clear that this could be set up. A license needs a legal entity, either
an individual, or a corporate entity. A collection of companies is not
an entity in this regard. It would be possible to establish an entity
for this purpose, but it would not be able to distribute to the other
entities involved.

It is quite intentional that the GPL makes this kind of cooperation without
open disclosure difficult!

Of course ultimately a court has to decide what is and what is not valid,
but certainly in the case of GCC, the copyright holder, Richard Stallman,
would probably object to any such dubious setup, as best I understand his
position.

RObert Dewar

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Richard E. Hawkins Esq.]

In article <7gpvsh$tdm$1@nnrp1.dejanews.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <7gpqms$7qj$1@eyry.econ>,
>  hawk@eyry.econ.iastate.edu (Richard E. Hawkins Esq.) wrote:

>> However, under this chip development scenario, the licensee of the GPL
>> is the conglomerate/joint enterprise/whatever of A, B, and I,
>> complete with their NDA's, which can internally distribute it's
>> modified version.  There is no subversion here; it's simply a question
>> of "who is the licensee."

>That does not work, unless there is a legal entity involved, not at all
>clear that this could be set up. A license needs a legal entity, either
>an individual, or a corporate entity. A collection of companies is not
>an entity in this regard. It would be possible to establish an entity
>for this purpose, but it would not be able to distribute to the other
>entities involved.

To start with, such an operation would be a partnership, unless a
contract or other action gave it another status (venture, corporation,
etc.).  It certainly is a legal entity.  

Additionally, it is not the collection of companies, but the development
effort; at first impression (which is as far as I'm going :), it doesn't
appear that this entity would be able to distribute to the to the
companies that formed it (though thoughtful planning might make it so,
but i don't have time to work with it [and noon'es paid my 
retainer :) ]).

>It is quite intentional that the GPL makes this kind of cooperation without
>open disclosure difficult!

It only makes it difficult without a minimally competent lawyer. 
However, it would prevent external distribution of binaries.


>Of course ultimately a court has to decide what is and what is not valid,
>but certainly in the case of GCC, the copyright holder, Richard Stallman,
>would probably object to any such dubious setup, as best I understand his
>position.

I won't grant you that it's dubious; these are rather simple legal
principles at work. 

However, all happens is that there is an internal version of GCC,
which could be useful during development, but could never be released
in binary only form to the customers who buy the chip.

Also, RMS's opinion on what is valid weighs in at about five billion
and one--after every one else on the planet. In interpreting
a document, the opinion of the author is the *very* last thing 
to consider.  For that matter, general contract principles call 
for interpreting any ambiguity *against* the author.

rick, esq.

-- 
These opinions will not be those of ISU until it pays my retainer.

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <x6zp3kk444.fsf@no-such-thing-as-a.free-lunch.demon.co.uk>,
  James Youngman <james@no-such-thing-as-a.free-lunch.demon.co.uk> wrote:

> What about "diff -e"?

What possible difference does that make. Remember that software
copyright is NOT based on literal text identity. You can have
two pieces with literal text identity that are not infringing,
and two pieces with no literal text identity which are infringing.

Either a patch is a derived work or it is not, the details of how
the patch is transmitted cannot matter. Given that the patch is
completely useless without the original work, and that in fact
the patch is operationally equivalent to the modified work if you
have the original, then it is hard for me to imagine that a court
would rule that it was not a derived work.

Note we are talking copyright here and not patents, the distinction
if of course crucial in this case.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <x6yaj4k3y2.fsf@no-such-thing-as-a.free-lunch.demon.co.uk>,
  James Youngman <james@no-such-thing-as-a.free-lunch.demon.co.uk> wrote:
>  You are
> allowed to *use* but not *copy* GPLed software without taking on even
> a single one of the obligations mentioned in the GPL.
>

The GPL *never* obliges anyone to do anything. It merely sets forth
a set of conditions under which you have a license to copy a
copyrighted work. You meet these conditions, and you may make
and distribute copies, but the GPL never obliges you to make or
distribute copies.

It also does not forbid anything. For example, suppose you want
to include a GPL'ed program in your proprietary licensed work.
Well you need the copyright holder's permission to do it. You
have the GPL, but that does not license this particular usage.
It doesn't forbid it, it just does not permit it, there is a big
difference. It is the copyright that forbids the use you have in
mind unless you get an appropriate license.

There is of course legally nothing to stop you asking for permission,
and nothing to stop the copyright holder from granting it.

Suppose you wanted to put together a compiler using XYZ's proprietary
front end and the gcc backend. Well to distribute this derived work,
you need the permission of the Free Software Foundation. Very probably
they will refuse, as is their right as the copyright holder. But they
could give you special permission, and then you could distribute your
hybrid program as a proprietary program without violating the GPL, it
is just that you would not be relying on the GPL for your right to
distribute.

In fact this is far from a theoretical situation, it is quite fine
legally for someone to distribute GPL'ed software for price $x,
possibly $0, and charge you a bundle for a separate license that
allows you to use the same software in a proprietary context. Indeed
it is my understanding (and I apologize in advance if this is an
incorrect understanding) that Cygwin is distributed under EXACTLY
this split model.

One may argue over whether such a distribution model is a good idea
(at ACT we prefer a completely clean situation with no such dual
licensing, which is why we use the modified GPL for runtime stuff),
but it is absolutely 100% consistent with the GPL. Indeed I have
heard people argue that it is appropriate, because it gives full
access to those writing free software, and penalizes those writing
proprietary software, which to some people seems an appropriate
situation :-)

Robert Dewar
Ada Core Technologies

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <7gqjpa$9cp$1@eyry.econ>,
  hawk@eyry.econ.iastate.edu (Richard E. Hawkins Esq.) wrote:

> Also, RMS's opinion on what is valid weighs in at about five billion
> and one--after every one else on the planet. In interpreting
> a document, the opinion of the author is the *very* last thing
> to consider.  For that matter, general contract principles call
> for interpreting any ambiguity *against* the author.

Well I would not be so quick to assume this. After all the GPL
was reviewed by lawyers whose retainer *was* paid :-)

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <7gr768$kqe$1@Jupiter.mcs.net>,
  les@MCS.COM (Leslie Mikesell) wrote:

> There are lawsuits over nearly everything.  Is there any court
> decision that states that something you write is covered by
> someone else's copyright when it contains none of the other
> work?

If by "contains none of the other work" you mean contains no literal
characters copied from the other work, then of course the answer
is yes.

The process for evaluation of copyright violation involves looking
for protected elements at successive levels of abstraction. It is
pretty clear to me that this abstraction process in the case of
a patch file will necessarily involve looking at the combined work.

Note that software copright is definitely NOT simply a matter of
comparing two texts, it is a far more complicated process, as anyone
who has been involved in such cases is aware.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert A Duff]

Robert Dewar <robert_dewar@my-dejanews.com> writes:

> Suppose two companies want to work together on a super
> secret GCC implementation for a super secret chip (I will
> leave the imaginative reader to fill in appropriate
> details).
> 
> They cannot exchange derived versions of gcc, since
> these would have to be distributed under the GPL *without*
> any additional restrictions (like non-disclosure
> agreements).

Wouldn't this reasoning apply within a *single* company, too?  Because
the single company developing the super-secret chip presumably has
employees, and very well might want to both (1) let them see the derived
work, and (2) forbid them from telling secrets.  In other words,
counting both the company and its employees, there are always going to
be two or more legal entities involved.

- Bob
-- 
Change robert to bob to get my real email address.  Sorry.

Re: GPL and "free" software

[Barry Margolin]

In article <7gqrbf$mun$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <7gqjpa$9cp$1@eyry.econ>,
>  hawk@eyry.econ.iastate.edu (Richard E. Hawkins Esq.) wrote:
>
>> Also, RMS's opinion on what is valid weighs in at about five billion
>> and one--after every one else on the planet. In interpreting
>> a document, the opinion of the author is the *very* last thing
>> to consider.  For that matter, general contract principles call
>> for interpreting any ambiguity *against* the author.
>
>Well I would not be so quick to assume this. After all the GPL
>was reviewed by lawyers whose retainer *was* paid :-)

The GPL never specifically addresses the issue we're discussing.  It's all
post facto posturing by RMS, as he tries to figure out how to apply the GPL
to situations that didn't exist when it was written.  What he feels is a
logical extension of the spirit of the GPL may not in fact be supported by
copyright law.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Robert Dewar]

In article <r2lY2.447$jw4.33999@burlma1-snr2>,
  Barry Margolin <barmar@bbnplanet.com> wrote:
> In article <7gqrbf$mun$1@nnrp1.deja.com>,

> The GPL never specifically addresses the issue we're discussing.  It's all
> post facto posturing by RMS, as he tries to figure out how to apply the GPL
> to situations that didn't exist when it was written.  What he feels is a
> logical extension of the spirit of the GPL may not in fact be supported by
> copyright law.

Nor could it! The issue of whether the patch file is or is not a derived
work is a matter of copyright law, not something that can be addressed
by the GPL.

However, it is my opinion that indeed the current copyright law would
determine that the patch file was a derived work. I base this opinion
on the work I have done as an expert witness in software copyright
cases. Of course it would need a court ruling to determine this
precisely.

I can also tell you that lawyers at one large company definitely
agree there is a real problem in distributing patch files ...

As always, feel free to try it and see. Note that most situations
in which companies have pushed the GPL have resulted in quick
agreements to change behavior and have not ended up in court. I
am not sure how many companies would be happy to be in court right
now in a position which would make them seem to be undermining the
principles of free software!

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <wcc7lqm41oj.fsf@world.std.com>,
  Robert A Duff <bobduff@world.std.com> wrote:

> Wouldn't this reasoning apply within a *single* company, too?  Because
> the single company developing the super-secret chip presumably has
> employees, and very well might want to both (1) let them see the derived
> work, and (2) forbid them from telling secrets.  In other words,
> counting both the company and its employees, there are always going to
> be two or more legal entities involved.

No, the company employees in this situation are part of the single
entity. This is the same principle that allows a company to sign an
NDA that is binding on all its employees without them having to sign
specifically. Giving a copy of GPL'ed software to one of your
employees is not distribution within the meaning of the GPL.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <r2lY2.447$jw4.33999@burlma1-snr2>,
  Barry Margolin <barmar@bbnplanet.com> wrote:

> The GPL never specifically addresses the issue we're discussing.  It's all
> post facto posturing by RMS, as he tries to figure out how to apply the GPL
> to situations that didn't exist when it was written.  What he feels is a
> logical extension of the spirit of the GPL may not in fact be supported by
> copyright law.

By the way Barry, perhaps it comes as a surprise to you, but the concept
of patch files is very very old, probably nearly as old as the concept
of programs. And yes, of course the issue of patch files has been
considered from the start, there is nothing "post facto" about this,
and of course this situation has existed from the inception of the GPL.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Barry Margolin]

In article <7gsrjo$g7i$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <r2lY2.447$jw4.33999@burlma1-snr2>,
>  Barry Margolin <barmar@bbnplanet.com> wrote:
>
>> The GPL never specifically addresses the issue we're discussing.  It's all
>> post facto posturing by RMS, as he tries to figure out how to apply the GPL
>> to situations that didn't exist when it was written.  What he feels is a
>> logical extension of the spirit of the GPL may not in fact be supported by
>> copyright law.
>
>By the way Barry, perhaps it comes as a surprise to you, but the concept
>of patch files is very very old, probably nearly as old as the concept
>of programs. And yes, of course the issue of patch files has been
>considered from the start, there is nothing "post facto" about this,
>and of course this situation has existed from the inception of the GPL.

I wasn't referring only to patch files, but all the varieties of things
that have been discussed in this thread or which RMS considers to be
derived works, such as dynamically-linked objects, plug-ins, etc.  Some of
them are not necessarily new concepts (Multics has used dynamic linking
since the 60's) but didn't see widespread use until relatively recently, so
there hasn't been much experience with applying copyright law to them.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Robert Dewar]

In article <jHmY2.473$jw4.33999@burlma1-snr2>,
  Barry Margolin <barmar@bbnplanet.com> wrote:

> I wasn't referring only to patch files, but all the varieties of things
> that have been discussed in this thread or which RMS considers to be
> derived works, such as dynamically-linked objects, plug-ins, etc.  Some of
> them are not necessarily new concepts (Multics has used dynamic linking
> since the 60's) but didn't see widespread use until relatively recently, so
> there hasn't been much experience with applying copyright law to them.

This thread has been specifically about patch files, not about other varieties
of things. If you want to broaden it out fine, but probably you should start
another thread and restrict it to gnu.misc.discuss.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Barry Margolin]

In article <7gt0is$kq5$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <jHmY2.473$jw4.33999@burlma1-snr2>,
>  Barry Margolin <barmar@bbnplanet.com> wrote:
>
>> I wasn't referring only to patch files, but all the varieties of things
>> that have been discussed in this thread or which RMS considers to be
>> derived works, such as dynamically-linked objects, plug-ins, etc.  Some of
>> them are not necessarily new concepts (Multics has used dynamic linking
>> since the 60's) but didn't see widespread use until relatively recently, so
>> there hasn't been much experience with applying copyright law to them.
>
>This thread has been specifically about patch files, not about other varieties
>of things. If you want to broaden it out fine, but probably you should start
>another thread and restrict it to gnu.misc.discuss.

This thread has many branches discussing many different things.  I barely
remember what it was originally about (the original message has expired on
my server and I don't feel like searching DejaNews); I think this may be
the thread that started with a question about how the GPL related to Ada
and C++ templates (I think that's why it was cross-posted to
comp.lang.ada).

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[Robert Dewar]

In article <qXnY2.481$jw4.34640@burlma1-snr2>,
  Barry Margolin <barmar@bbnplanet.com> wrote:
> In article <7gt0is$kq5$1@nnrp1.deja.com>,
> Robert Dewar  <robert_dewar@my-dejanews.com> wrote:

> This thread has many branches discussing many different things.  I barely
> remember what it was originally about (the original message has expired on
> my server and I don't feel like searching DejaNews); I think this may be
> the thread that started with a question about how the GPL related to Ada
> and C++ templates (I think that's why it was cross-posted to
> comp.lang.ada).
>
> --
> Barry Margolin, barmar@bbnplanet.com

If you have trouble keeping track of threads, I strongly suggest using
dejanews to read your newsgroup traffic. The nice new thread following
software, with the graphical diagrams keeping track of subthreads, will
avoid the "barely remember" phenomenon, and help you to avoid going off
track :-)

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Roger Espel Llima]

In article <7gt7qu$rlo$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>
>If you have trouble keeping track of threads, I strongly suggest using
>dejanews to read your newsgroup traffic. The nice new thread following
>software, with the graphical diagrams keeping track of subthreads, will
>avoid the "barely remember" phenomenon, and help you to avoid going off
>track :-)

last time I tried reading a newsgroup on DejaNews, I couldn't find a way
to make it remember which posts I'd seen, from one time to the next.

(to be fair, it wouldn't be easy for a web-based newsreading service to
do that; they'd basically have to make you log in and keep your .newsrc-
equivalent on their server.  that, or give you a giant cookie, probably
big enough that the browser would throw it away)

anyway.. have they fixed that?

-- 
Roger Espel Llima, espel@llaic.u-clermont1.fr
http://www.eleves.ens.fr:8080/home/espel/index.html

Re: GPL and "free" software

[Robert Dewar]

In article <373248e6@eeyore.callnetuk.com>,
  "Nick Roberts" <nickroberts@callnetuk.com> wrote:

> I think the relevant point is that, in any eventuality, if a program uses a
> template or generic unit, then some object code corresponding to that
> template or generic unit will end up in the program's object code.  I feel
> sure it is the case that the source code for many templates and generic
> units will constitute an 'original and substantive work' susceptible to a
> claim of copyright.

Indeed! And in fact some vendors of proprietary software are not at all
happy with the model of instantiation by expansion, because it forces them
to reveal source code which they want not only to copyright, but also to
keep secret.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Robert Dewar]

In article <7gt9k8$2pl$1@nef.ens.fr>,
  espel@news.ens.fr (Roger Espel Llima) wrote:
> In article <7gt7qu$rlo$1@nnrp1.deja.com>,

> last time I tried reading a newsgroup on DejaNews, I couldn't find a way
> to make it remember which posts I'd seen, from one time to the next.
>
> (to be fair, it wouldn't be easy for a web-based newsreading service to
> do that; they'd basically have to make you log in and keep your .newsrc-
> equivalent on their server.  that, or give you a giant cookie, probably
> big enough that the browser would throw it away)

Absolutely, you log in to mydejanews, and it keeps track of what groups
you are subscribed to, what messages you have read, and gives you a personal
email account (I am using it to send this message).

I used to use a Unix based newsreader to access news, but I find the
current thread-based access in DN to be greatly superior to anything
else I have seen. So if you are even a little bit unhappy with your
news access, give DN a try (the price is right, $0, as seems to be
the custom for internet stuff :-)

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Barry Margolin]

In article <7gur7s$75n$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>I used to use a Unix based newsreader to access news, but I find the
>current thread-based access in DN to be greatly superior to anything
>else I have seen. So if you are even a little bit unhappy with your
>news access, give DN a try (the price is right, $0, as seems to be
>the custom for internet stuff :-)

I use trn, which provides a reasonable display of the thread history.  I
find DejaNews satisfactory for archive searching, but the performance is
often too slow for regular newsreading.

-- 
Barry Margolin, barmar@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.

Re: GPL and "free" software

[dennison]

In article <7gur7s$75n$1@nnrp1.deja.com>,
  Robert Dewar <robert_dewar@my-dejanews.com> wrote:
> In article <7gt9k8$2pl$1@nef.ens.fr>,
>   espel@news.ens.fr (Roger Espel Llima) wrote:
> > In article <7gt7qu$rlo$1@nnrp1.deja.com>,
>
> > last time I tried reading a newsgroup on DejaNews, I couldn't find a way
> > to make it remember which posts I'd seen, from one time to the next.
> >
> > (to be fair, it wouldn't be easy for a web-based newsreading service to
> > do that; they'd basically have to make you log in and keep your .newsrc-
> > equivalent on their server.  that, or give you a giant cookie, probably
> > big enough that the browser would throw it away)
>
> Absolutely, you log in to mydejanews, and it keeps track of what groups
> you are subscribed to, what messages you have read, and gives you a personal
> email account (I am using it to send this message).
>
> I used to use a Unix based newsreader to access news, but I find the
> current thread-based access in DN to be greatly superior to anything
> else I have seen. So if you are even a little bit unhappy with your
> news access, give DN a try (the price is right, $0, as seems to be
> the custom for internet stuff :-)

To be fair it still does have some kinks. It seems to randomly switch me from
threaded viewing to message-based viewing. Read messages reappear as unread
sometimes. It stops updating for hours. The latest trend is to tell me
"Access Denied" on messages. But the service interruptions you will get with
any news server, and it has a nice set of features that are not all available
in one other reader that I have used. (Cancelling messages, surfing to URL's,
searching Usenet history back to 1995, etc).

The best part (for me) is that it works from sites where the firewall has
every protocol other than SMTP and HTTP disabled.

--
T.E.D.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[Joshua E. Rodd]

Robert Dewar wrote:
> No, the company employees in this situation are part of the single
> entity. This is the same principle that allows a company to sign an
> NDA that is binding on all its employees without them having to sign
> specifically. Giving a copy of GPL'ed software to one of your
> employees is not distribution within the meaning of the GPL.

Yes, it is. You couldn't give an employee a copy of Winodws 95
to install on their PC unless you had paid for a licence because
that would infringe the copyright.

Likewise, you may only redistribute GPL'd works if you do not
place any restrictions upon that the recipient of the works can
do with those works. If you restrict your employee from posting
copies of the code you gave him or her to some random FTP site,
you're violating the licence, and have no right to redistribute
the GPL'd works with your changes to your employee at all.

Re: GPL and "free" software

[dewarr]

In article <37336A2E.774E0A80@iiinet.dhs.org>,
  "Joshua E. Rodd" <jerodd@iiinet.dhs.org> wrote:

>
> Yes, it is. You couldn't give an employee a copy of Winodws 95
> to install on their PC unless you had paid for a licence because
> that would infringe the copyright.
>
> Likewise, you may only redistribute GPL'd works if you do not
> place any restrictions upon that the recipient of the works can
> do with those works. If you restrict your employee from posting
> copies of the code you gave him or her to some random FTP site,
> you're violating the licence, and have no right to redistribute
> the GPL'd works with your changes to your employee at all.

Nope, this is wrong! It has been much discussed before, so consult
the archives for details.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[dewarr]

In article <OYGX2.350$jw4.27311@burlma1-snr2>,
  Barry Margolin <barmar@bbnplanet.com> wrote:

> They are, but in some cases copying small snippets out of a larger work
> might be fair use, and hence not infringing.

I know of no court cases that would back up this claim, can you give
a reference. It would definitely be of interest to know if any court
has established this principle clearly.

Robert Dewar

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: GPL and "free" software

[dewarr]

In article <7g7tlc$j5g$1@jetsam.uits.indiana.edu>,
  owinebar@ezinfo.ucs.indiana.edu (Lynn Winebarger) wrote:

>  Well, I think it would be legal for a company to distribute patches
> (under a stricter license even), as long as they did not also distribute
> the original work.  Not that I would want to encourage that behaviour.
>
> Lynn

It would be useful when people state legal opinions of this nature, if they
indicate the source (e.g. their own legal expertise, or court cases).
Otherwise it is very hard to identify substantive input from guesses about
what people think should be the case from an informal point of view. The
trouble with the latter is that so often "common sense" can mislead you when
it comes to legal issues. This is particularly true in the software copyright
business (I base this observation on my experience in acting as an expert
witness in several court cases involving alledged copyright infringement, and
on having read the principle court decisions in this area). In addition, the
field is changing very rapidly as congress takes action on a number of bills
that have a significant effect on copyrights in general, and software
copyrights in particular.

An example: If you have a (legitimate) copy of a program on your hard disk,
then does executing it involve making a copy? Tricky question, and courts
have disagreed, it is not a trivial question to answer at all! The point is
that in some sense there is a copy in memory .. what sense? Well that's the
question to be answered!

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: [O/T 4 cla] Re: GPL and "free" software

[Leslie Mikesell]

In article <7grvur$li2$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>
>> There are lawsuits over nearly everything.  Is there any court
>> decision that states that something you write is covered by
>> someone else's copyright when it contains none of the other
>> work?
>
>If by "contains none of the other work" you mean contains no literal
>characters copied from the other work, then of course the answer
>is yes.

No, it wasn't a trick question.  Obviously you could encode the
original work in some other representation and it would still be
covered.  However, the concept of a 'diff' is exactly the opposite.
It conveys, by definition, the parts that were not in the original.

>The process for evaluation of copyright violation involves looking
>for protected elements at successive levels of abstraction. It is
>pretty clear to me that this abstraction process in the case of
>a patch file will necessarily involve looking at the combined work.

If the only part copied is the patch, how can anything else be
involved?  A normal patch file does contain bits of the surrounding
context; even in the ed form there will be original content as parts
of some lines.  I'd think this would be 'fair use' particularly
since the recipient must already have his own copy of the original
to use the patch.  However that objection could be eliminated by
a different patch format that used character offsets with the lines
to anchor the changes instead of containing any of the original content.

>Note that software copright is definitely NOT simply a matter of
>comparing two texts, it is a far more complicated process, as anyone
>who has been involved in such cases is aware.

Like they say, if the two lawyers can't agree, how are the jurors
supposed to...

  Les Mikesell
     les@mcs.com

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <7h2aca$2mqq$1@Mercury.mcs.net>,
  les@MCS.COM (Leslie Mikesell) wrote:
> Like they say, if the two lawyers can't agree, how are the jurors
> supposed to...

I really strongly suggest that you look up the case law here
to better understand the abstract-filter-compare-iterate
methodology, which is fundamental for judging software copyright
issues.

-----------== Posted via Deja News, The Discussion Network ==----------
http://www.dejanews.com/       Search, Read, Discuss, or Start Your Own    

Re: [O/T 4 cla] Re: GPL and "free" software

[Leslie Mikesell]

In article <7h83tf$150$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <7h5rf0$i3t$1@Mercury.mcs.net>,
>  les@MCS.COM (Leslie Mikesell) wrote:
>> Can you suggest a starting point where (a) the parties had not
>> agreed to restrictive licensing terms and (b) the material
>> copied
>> contained only changes to the covered work and none of the
>> original?  Any other type of case would not have much
>> relevance.
>
>I disagree, if one followed the filter-compare-abstract cycle
>with a patch, it is pretty clear that the very first level
>of abstraction (going from the specific text to the abstracted
>meaning) would immediately involve the original work. I don't
>see any way to follow the required FCA methodology that would
>avoid this.

Given that the recipient of the patch already has his own copy
of the original material obtained by other means, I'm not quite
sure why this would be relevant.  The abstract meaning is all
different from the original in any case.  The way you represent
'copy the original' in a patch file is to omit any mention
of it.

>If you can see how the FCA approach would apply in some other
>manner to a patch file, please elucidiate! Certainly if I was
>an expert in such a case, and I was asked to abstract a patch
>file, I would have to see the original file that was being
>patched, I can't see any other way to proceed.

I think of a patch as a mechanical abstraction of someone retyping
a set of changes, equivalent to the author of the changes dictating
the keystrokes to you to reproduce his modifications.  Yes there
is some relationship to the original work, but at the patch/modification
level that relationship is just an interface to anchor the position
of the changes.  Even where patches contain substantial portions
of the original work, the function of those parts is still just
mechanical, as a positioning aid, not as content.  The part being
conveyed is by definition, differences.

Is the orginal modification legal?  Would it be legal for the
author of the modification to come over and retype it for you?
How about dictating the edit commands to you over the phone as you
modify your own copy?  This is exactly the function the patch file
performs.  I don't see how one way of performing these changes
can be treated any differently than the others - they are really
all the same.

  Les Mikesell
    les@mcs.com

Re: [O/T 4 cla] Re: GPL and "free" software

[Leslie Mikesell]

In article <7h840p$15i$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <7h5rf0$i3t$1@Mercury.mcs.net>,
>> >> Like they say, if the two lawyers can't agree, how are
>> >> the jurors supposed to...
>
>By the way, I don't know who "they" is here, but this does not
>make much sense. The two lawyers in a trial will ALWAYS
>disagree, they are basically required to do so, but juries can
>still often reach a unanimous verdict :-)

It was an extract from a joke that I assumed everyone had 
heard.  Oh, well.

  Les Mikesell
    les@mcs.com

Re: [O/T 4 cla] Re: GPL and "free" software

[Leslie Mikesell]

In article <7h2i00$adl$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:

>> Like they say, if the two lawyers can't agree, how are the jurors
>> supposed to...
>
>I really strongly suggest that you look up the case law here
>to better understand the abstract-filter-compare-iterate
>methodology, which is fundamental for judging software copyright
>issues.

Can you suggest a starting point where (a) the parties had not
agreed to restrictive licensing terms and (b) the material copied
contained only changes to the covered work and none of the 
original?  Any other type of case would not have much relevance.

  Les Mikesell
    les@mcs.com

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <7h8aqf$4nm$1@Mercury.mcs.net>,
  les@MCS.COM (Leslie Mikesell) wrote:
> Is the orginal modification legal?  Would it be legal for the
> author of the modification to come over and retype it for you?
> How about dictating the edit commands to you over the phone as
> you modify your own copy?  This is exactly the function the
> patch file performs.  I don't see how one way of performing
> these changes can be treated any differently than the others -
> they are really all the same.

I assume you are speaking as an attorney, or at least an expert
in the copyright field, so it would be useful if you would cite
some basis for your legal opinions.

I continue to think that the abstraction process would
immediately indicate that the patch file was a derived work
in the normal sense of the meaning of this term. Once again,
I do not see how the filter-compare-abstraction cycle could
be applied to the patch file without reference to the original.
Remember that you have to get all the way to the top in this
cycle -- to something like "write a compiler" ...

Can you coordinate your viewpoint with the specific court cases
involved here (I assume you are familiar with them ...)


--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---

Re: [O/T 4 cla] Re: GPL and "free" software

[Leslie Mikesell]

In article <7h97qi$t1j$1@nnrp1.deja.com>,
Robert Dewar  <robert_dewar@my-dejanews.com> wrote:
>In article <7h8aqf$4nm$1@Mercury.mcs.net>,
>  les@MCS.COM (Leslie Mikesell) wrote:
>> Is the orginal modification legal?  Would it be legal for the
>> author of the modification to come over and retype it for you?
>> How about dictating the edit commands to you over the phone as
>> you modify your own copy?  This is exactly the function the
>> patch file performs.  I don't see how one way of performing
>> these changes can be treated any differently than the others -
>> they are really all the same.
>
>I assume you are speaking as an attorney, or at least an expert
>in the copyright field, so it would be useful if you would cite
>some basis for your legal opinions.

Of course not - this is usenet.  Anything goes here.  But I
do have a pretty good understanding of what a patch actually
does.

>I continue to think that the abstraction process would
>immediately indicate that the patch file was a derived work
>in the normal sense of the meaning of this term.

How?  After applying the patch you would have a derived
work.  The patch itself is just replaying the keystrokes used
to type in the modifications.

>Once again,
>I do not see how the filter-compare-abstraction cycle could
>be applied to the patch file without reference to the original.
>Remember that you have to get all the way to the top in this
>cycle -- to something like "write a compiler" ...

Sure, but a patch that represents the original best is
an empty file.  How can an empty file be illegal to copy?
The non-empty portion of a patch file does not represent
the original content.

>Can you coordinate your viewpoint with the specific court cases
>involved here (I assume you are familiar with them ...)

No, I don't think there has been anything similar.  The GPL is
a rather strange beast and when you combine that with the fact
that people are allowed to get their own copies without first
agreeing to a license you have a unique situation. 

  Les Mikesell
   les@mcs.com

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <7h5rf0$i3t$1@Mercury.mcs.net>,
  les@MCS.COM (Leslie Mikesell) wrote:
> Can you suggest a starting point where (a) the parties had not
> agreed to restrictive licensing terms and (b) the material
> copied
> contained only changes to the covered work and none of the
> original?  Any other type of case would not have much
> relevance.

I disagree, if one followed the filter-compare-abstract cycle
with a patch, it is pretty clear that the very first level
of abstraction (going from the specific text to the abstracted
meaning) would immediately involve the original work. I don't
see any way to follow the required FCA methodology that would
avoid this.

If you can see how the FCA approach would apply in some other
manner to a patch file, please elucidiate! Certainly if I was
an expert in such a case, and I was asked to abstract a patch
file, I would have to see the original file that was being
patched, I can't see any other way to proceed.



--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <7h5rf0$i3t$1@Mercury.mcs.net>,
> >> Like they say, if the two lawyers can't agree, how are
> >> the jurors supposed to...

By the way, I don't know who "they" is here, but this does not
make much sense. The two lawyers in a trial will ALWAYS
disagree, they are basically required to do so, but juries can
still often reach a unanimous verdict :-)



--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---

Re: [O/T 4 cla] Re: GPL and "free" software

[Robert Dewar]

In article <7ha31n$htn$1@Mercury.mcs.net>,
  les@MCS.COM (Leslie Mikesell) wrote:
> Of course not - this is usenet.  Anything goes here.

I am not quite sure what that means. It is always important
to understand what kind of knolwedge and experience goes into
usenet posts. The "of course not" here is not appropriate. Yes
there are plenty of people posting who do not have the knowledge
to post but post anyway, but there are also many experts who
do speak as experts (we have had at least one person whom I
assumed was an attorney or at least very knowledgable in the
law posting here, and up to the above, I had assumed that you
had similar expertise).

> But I do have a pretty good understanding of what a patch
> actually does.

I think you may assume that we all know what a patch is (says
he trying to keep from making some sarcastic comment :-)
Seriously, of COURSE we all know what a patch is, but that
has little to do with the discussion at hand, which is not
about patches but about software copyright.

The issue is knowing how software copyright works in terms
of case law. The case law here is not inaccessible, in fact
it makes quite interesting technical reading.

You can't figure out what the law is by thinking about what
it *should* be according to your common sense view, you have
to look at what the statutes actually say, and in this
particular case, understand the abstraction process, which is
the key here, and which I think you are not understanding as
best I can read your posts.

Once again you do NOT simply copyright text when it comes
to software, you copyright protectable elements of the design
at all levels of abstraction.

Could you copyright an empty file. I would say that you can
certainly see circumstances in which an empty file could be
copyrighted.

Suppose a file represented some particular option in a standard
algorithm.

You discovered a new more efficient form in which this option
was not needed.

You program the result, and the file for the option is now null
since the option is not needed.

I would guess that this would be a protectable element.
But you would have to look at the exact circumstances to
actually make a judgment.

Software copyright is not a simple issue, as many recent
court cases have shown.




--== Sent via Deja.com http://www.deja.com/ ==--
---Share what you know. Learn what you don't.---

Re: GPL and "free" software

[Joshua E. Rodd]

dewarr@my-dejanews.com wrote:
> > Yes, it is. You couldn't give an employee a copy of Winodws 95
> > to install on their PC unless you had paid for a licence because
> > that would infringe the copyright.
> > Likewise, you may only redistribute GPL'd works if you do not
> > place any restrictions upon that the recipient of the works can
> > do with those works. If you restrict your employee from posting
> > copies of the code you gave him or her to some random FTP site,
> > you're violating the licence, and have no right to redistribute
> > the GPL'd works with your changes to your employee at all.
> Nope, this is wrong! It has been much discussed before, so consult
> the archives for details.

Can you elaborate?