Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Linux is a GNU system and the DWARF support

97 views
Skip to first unread message
Message has been deleted

Joe Buck

unread,
Sep 7, 1994, 1:52:31 AM9/7/94
to
h...@nynexst.com (H.J. Lu) writes:
>I still think the commercial softwares should be encouraged under
>Linux. There are places for everyone under Linux. The kernel, the C
>library and other basic components should remain free. The user can make
>choices on the commercial applications.

Certainly. And RMS and MIB have also said that commercial/binary-only
stuff will be allowed on the Hurd.

Note, however, that crt0.S could have used the "as a special exception"
copyright used by the stuff in libgcc.a: the file itself is under the
GPL, but as a special exception, applications built with gcc and linked
to it don't fall under the GPL. This is how the FSF and Cygnus folk
have been addressing the issue of key files that always have to be
linked in.

--
-- Joe Buck <jb...@synopsys.com>
Posting from but not speaking for Synopsys, Inc.
***** Stamp out junk e-mail spamming! If someone sends you a junk e-mail
***** ad just because you posted in comp.foo, boycott their company.

Message has been deleted

Michael I Bushnell

unread,
Sep 8, 1994, 10:14:56 AM9/8/94
to
In article <Cvt4F...@info.swan.ac.uk> iia...@iifeak.swan.ac.uk (Alan Cox) writes:

The FSF seem to have a habit of trying to use their (as opposed to other
peoples) GPL'd code as a political lever at time (eg the RSAREF thing).
Thats a pity and detracts from the basic aim.

Apparently you haven't understood the basic aim of the FSF. Far from
detracting from the basic aim, that *IS* the basic aim.

--
+1 617 623 3248 (H) | En arche en ho logos,
+1 617 253 8568 (W) -+- kai ho logos en pros ton theon,
1105 Broadway | kai theos en ho logos.
Somerville, MA 02144 | Kai ho logos sarx egeneto,
m...@gnu.ai.mit.edu | kai eskenosen en hemin.

Sven Goldt

unread,
Sep 6, 1994, 11:00:48 AM9/6/94
to
H.J. Lu (h...@nynexst.com) wrote:
: which is due in a few weeks. I think noone should call GNU is a vaporware
: now since the Linux/GNU system is running on x86. I was told if Linux had

And on the linux channel NORMAL he wrote:
:It looks like Linux will be/is pretty much a part of the GNU project.
:Please read my article in comp.os.linux.misc for details.

:We seem to have to work on some copyright issues on the Linux C
:library. If yuou have contrbuted anything to the Linux C library,
:please get in touch with me ASAP.

So then please keep that in mind which can be found in the linux libc
sources in a file called crt0.S:

/* Notice of general intent:
*
* The linux operating system generally contains large amounts of code
* that fall under the GNU General Public License, or GPL for short.
* This file contains source code that by it's very nature would always
* be linked with an application program, and because of this a GPL type
* of copyright on this file would place restrictions upon the
* distribution of binary-only commercial software. Since the goal of the
* Linux project as a whole is not to discourage the development and
* distribution of commercial software for Linux, this file has been placed
* under a more relaxed BSD-style of copyright.
*
* It is the general understanding of the above contributors that a
* program executable linked to a library containing code that falls
* under the GPL or GLPL style of license is not subject to the terms of
* the GPL or GLPL license if the program executable(s) that are supplied
* are linked to a shared library form of the GPL or GLPL library, and as long
* as the form of the shared library is such that it is possible for
* the end user to modify and rebuild the library and use it in
* conjunction with the program executable.
*/


--
*****************************************************************************
* # THE MOST IMPORTANT FINANCIAL QUESTION IS: Where is the money ? # *
*****************************************************************************

Matt Welsh

unread,
Sep 8, 1994, 10:22:06 AM9/8/94
to
In article <34l5qb$d...@geraldo.cc.utexas.edu> li...@slip-14-1.ots.utexas.edu (Dances With Geeks) writes:
>There are licensing problems in the standard GNU approach,
>again IMO, involving the linkage of libraries into applications. I'm not
>saying the GNU approach is wrong, just that it's not as open as it might be.

Sorry, but you're stuck with the "GNU approach" (whatever that means)
because you use software and libraries covered by the GPL. Any "problems"
perceived with GNU software applies equally to Linux.

RMS's idea (which I have heard first-hand) is that Linux systems
should be considered GNU systems with Linux as the kernel. This
might be an over-generalization, but you get the idea. Linux systems
don't use exclusively GNU software, and don't adhere 100% to the
GPL ideal. However, the _guts_ of the system (kernel, libraries,
all of the basic and no-so-basic binaries, and so forth) are all
covered by the GPL. Most of these were developed by the GNU project.
Why shouldn't GNU receive recognition for this? RMS is simply stating
the status quo in a different way.

The Debian Linux Association is working with the FSF in order to
develop a good working relationship with them, as well as to
assist each other through shared resources. Case in point: I'm
going to Cambridge (perhaps next week) to install Debian on a machine
on the FSF's network. The GNU project has donated the machine and
network connectivity for Debian's use---something that we can all
benefit from.

Calm down. The FSF isn't the Borg. They are not out to assimilate Linux.

M. Welsh

Dances With Geeks

unread,
Sep 7, 1994, 3:54:51 PM9/7/94
to
On 5 Sep 1994 22:04:13 GMT, H.J. Lu (h...@nynexst.com) wrote:

> Today after I talked with RMS on the phone, I learned that the GNU project
> is preparing for a Linux/GNU distribution based on the Debian distribution,


> which is due in a few weeks. I think noone should call GNU is a vaporware
> now since the Linux/GNU system is running on x86. I was told if Linux had

> came along be fore hurd was started Linux would have been used as the GNU
> kernel. Now since hurd is well on the way, it should not be aborted. It looks
> like porting Linux to other platforms may pick up some steams as well as the
> work on the Linux kernel. I am glad to see it finnally happen.

I don't want to rain on the parade, but I hope we do *not* start considering
ourselves a GNU system. We use the GNU software, it's wonderful stuff, I
like it, but we also use whatever fits in with the Linux philosophy, which
is much more unconstrained and much less ideological than the GNU
philosophy, IMO. There are licensing problems in the standard GNU approach,


again IMO, involving the linkage of libraries into applications. I'm not
saying the GNU approach is wrong, just that it's not as open as it might be.

I'm glad to see that the Debian release is moving along--but not because I
think Linux development should be subsumed into it. On the contrary, I
think the strength of development for Linux is that there are a number of
competing, independent threads which serve to strengthen the whole tapestry.
I would hate to see that change, as it would inevitably do if we became part
of a less-flexible structure.

I very much appreciate the amount of work that has gone into GNU software,
and the quality of that software. I use it all the time! But I don't think
that recognition by FSF is the primary measure of the success of Linux.
They are different things entirely.

Mainstreaming support for Linux in GCC is another issue entirely, BTW....if
Linux is used extensively in the community, it probably makes sense for both
Linux people and GNU people for both groups to work closely together....

Anyway, if my message isn't reasonably diplomatic, I better apologize in
advance, I've had nine hours of sleep in two days and really need about
twice that :) :)


lilo

Joe Buck

unread,
Sep 7, 1994, 8:21:55 PM9/7/94
to
li...@slip-14-1.ots.utexas.edu (Dances With Geeks) writes:
>There are licensing problems in the standard GNU approach,
>again IMO, involving the linkage of libraries into applications. I'm not
>saying the GNU approach is wrong, just that it's not as open as it might be.

If there are licensing problems in the GNU approach, then Linux suffers
from them. The kernel is under the GPL, the C library is under the LGPL,
certain key files always linked in, like libgcc.a and crt0.S have special
exceptions to avoid bringing whole applications under the GPL. These
statements are true of both Linux and the future Hurd. There is no
difference, Linux is using every single GNU licensing convention exactly
as the FSF intended.

The only difference I can see is attitude: the FSF people talk about
software hoarders, grumble a lot but then adjust their licenses where
needed (example: the stream classes in libg++ moved from the LGPL to the
"as a special exception" no-restrictions license), and the Linux people
are more relaxed about the whole thing. But legally, it's *exactly* the
same: if you have problems with the GNU licensing approach you'll need
to abandon Linux and join up with the BSD folks.

Alan Cox

unread,
Sep 8, 1994, 6:43:07 AM9/8/94
to
In article <34llf3$s...@hermes.synopsys.com> jb...@synopsys.com (Joe Buck) writes:
>certain key files always linked in, like libgcc.a and crt0.S have special
>exceptions to avoid bringing whole applications under the GPL. These
>statements are true of both Linux and the future Hurd. There is no

Note the kernel syscall interface is also explicitly removed from the list.
If I wanted to be paranoid, I'd download the dll tools generate my own
dynamic library binding files (so they are mine not GNU's) and have a quick
check that none of the headers used >10 lines of code. Structures then don't
matter here (interface copyright stuff).

>The only difference I can see is attitude: the FSF people talk about
>software hoarders, grumble a lot but then adjust their licenses where
>needed (example: the stream classes in libg++ moved from the LGPL to the

The FSF seem to have a habit of trying to use their (as opposed to other


peoples) GPL'd code as a political lever at time (eg the RSAREF thing).
Thats a pity and detracts from the basic aim.

Alan
[GPL but not FSF supporter...]
--
..-----------,,----------------------------,,----------------------------,,
// Alan Cox // iia...@www.linux.org.uk // GW4PTS@GB7SWN.#45.GBR.EU //
``----------'`----------------------------'`----------------------------''

Jay Maynard

unread,
Sep 8, 1994, 7:54:46 PM9/8/94
to
In article <1994Sep8.1...@cs.cornell.edu>,

Matt Welsh <m...@cs.cornell.edu> wrote:
>Calm down. The FSF isn't the Borg. They are not out to assimilate Linux.

Of course not. As you point out earlier in your message, they've already
done it.
--
Jay Maynard, EMT-P, K5ZC, PP-ASEL | Never ascribe to malice that which can
jmay...@admin5.hsc.uth.tmc.edu | adequately be explained by stupidity.
The US Constitution: 1789-1994. RIP.

Pete Bergstrom

unread,
Sep 9, 1994, 6:05:14 PM9/9/94
to
>>>>> On 8 Sep 1994 17:54:46 -0600, jmay...@nyx10.cs.du.edu (Jay Maynard) said:
>>Calm down. The FSF isn't the Borg. They are not out to assimilate Linux.

> Of course not. As you point out earlier in your message, they've already
> done it.

Wrong. The implementers of Linux actively decided to use FSF-produced
tools which fall under the GPL and also decided to place the (majority
of) the kernel under the GPL. The FSF really didn't have any say in
the matter (and certainly your statement that "they've already done
it" is incorrect - "they" didn't do anything of the sort).

Pete

Orc

unread,
Sep 9, 1994, 5:35:17 PM9/9/94
to
In article <1994Sep8.1...@cs.cornell.edu>,
Matt Welsh <m...@cs.cornell.edu> wrote:

>RMS's idea (which I have heard first-hand) is that Linux systems
>should be considered GNU systems with Linux as the kernel.

Well, that certainly puts a new twist on the GPL. An unpleasant
one; GPLing things makes them part of the GNU project.

[...]


>Why shouldn't GNU receive recognition for this?

No reason. But since my first interpretation of RMS's idea is
that the FSF will take credit for work they've not done, it seems
like it's not the most politic way to ask for recognition.

____
david parsons \bi/ who won't be GPLing any more of his code.
\/

Dances With Geeks

unread,
Sep 10, 1994, 8:12:11 PM9/10/94
to
On 8 Sep 1994 00:21:55 GMT, Joe Buck (jb...@synopsys.com) wrote:

> li...@slip-14-1.ots.utexas.edu (Dances With Geeks) writes:
> >There are licensing problems in the standard GNU approach,
> >again IMO, involving the linkage of libraries into applications. I'm not
> >saying the GNU approach is wrong, just that it's not as open as it might
> >be.

> If there are licensing problems in the GNU approach, then Linux suffers
> from them. The kernel is under the GPL, the C library is under the LGPL,
> certain key files always linked in, like libgcc.a and crt0.S have special
> exceptions to avoid bringing whole applications under the GPL. These
> statements are true of both Linux and the future Hurd. There is no
> difference, Linux is using every single GNU licensing convention exactly
> as the FSF intended.

Well, note that certain additional disclaimers have been added in the
kernel. I think it's a key point.

> The only difference I can see is attitude: the FSF people talk about
> software hoarders, grumble a lot but then adjust their licenses where
> needed (example: the stream classes in libg++ moved from the LGPL to the
> "as a special exception" no-restrictions license), and the Linux people
> are more relaxed about the whole thing. But legally, it's *exactly* the
> same: if you have problems with the GNU licensing approach you'll need
> to abandon Linux and join up with the BSD folks.

I believe, even more than the legal issues (and they are there,
indiscriminate application of an unmodified GPL or LGPL would certainly
bring them out), I do have problems with depending on the "largesse" of the
FSF people to somehow always make exceptions. It's a bad idea. Having the
FSF "take over" Linux, or WINE, or what-have-you, conveys no real advantages
and has the disadvantage of involving what might be well on its way towards
becoming a software bureaucracy. No offense intended to anyone. :)


lilo

Dances With Geeks

unread,
Sep 10, 1994, 8:23:25 PM9/10/94
to
On Thu, 8 Sep 1994 14:22:06 GMT, Matt Welsh (m...@cs.cornell.edu) wrote:
> In article <34l5qb$d...@geraldo.cc.utexas.edu> li...@slip-14-1.ots.utexas.edu (Dances With Geeks) writes:
> >There are licensing problems in the standard GNU approach,
> >again IMO, involving the linkage of libraries into applications. I'm not
> >saying the GNU approach is wrong, just that it's not as open as it might be.

> Sorry, but you're stuck with the "GNU approach" (whatever that means)
> because you use software and libraries covered by the GPL. Any "problems"
> perceived with GNU software applies equally to Linux.

Sorry, but you're wrong. The Linux kernel, for example, adds additional
disclaimers which modify the GNU-format license it is used under.

Also, a GNU license which is owned by FSF would be enforced by FSF (at least
FSF would attempt to enforce its license as it saw fit, which could
certainly be a nuisance at times, and which would probably change the
emphasis of Linux licensing.).

> RMS's idea (which I have heard first-hand) is that Linux systems
> should be considered GNU systems with Linux as the kernel. This
> might be an over-generalization, but you get the idea. Linux systems
> don't use exclusively GNU software, and don't adhere 100% to the
> GPL ideal. However, the _guts_ of the system (kernel, libraries,
> all of the basic and no-so-basic binaries, and so forth) are all
> covered by the GPL. Most of these were developed by the GNU project.
> Why shouldn't GNU receive recognition for this? RMS is simply stating
> the status quo in a different way.

GNU software *has* received recognition for being GNU software.
Nonetheless, the Linux itself was not written by FSF, has a modified GPL,
and could be packaged in a distribution which did not include FSF software
(and, I think, has been, at early stages of its development). It would
still remain Linux. FSF software without Linux would just be, um, FSF
software, wouldn't it? ;)

I have no problem with RMS stating his views in any way he sees fit.
Nonetheless, I believe too close an association with FSF would not
especially profit Linux.

> The Debian Linux Association is working with the FSF in order to
> develop a good working relationship with them, as well as to
> assist each other through shared resources. Case in point: I'm
> going to Cambridge (perhaps next week) to install Debian on a machine
> on the FSF's network. The GNU project has donated the machine and
> network connectivity for Debian's use---something that we can all
> benefit from.

That's nice. I'm sure everyone will benefit from the competition, as long
as we don't all decide that all we need is One Really Good Distribution. :)
:)

> Calm down. The FSF isn't the Borg. They are not out to assimilate Linux.

Oh, I'm perfectly calm. Thank you for your solicitous interest *innocent
look*. ;)

> M. Welsh


lilo

Matt Welsh

unread,
Sep 11, 1994, 4:08:42 PM9/11/94
to
In article <CvvtA...@pell.com> o...@pell.com (Orc) writes:
>In article <1994Sep8.1...@cs.cornell.edu>,
>Matt Welsh <m...@cs.cornell.edu> wrote:
>>RMS's idea (which I have heard first-hand) is that Linux systems
>>should be considered GNU systems with Linux as the kernel.
>
> Well, that certainly puts a new twist on the GPL. An unpleasant
>one; GPLing things makes them part of the GNU project.

This is absolutely wrong. This viewpoint has nothing to do with
the fact that Linux is covered by the GPL. If you had bothered
to read my message, you would have known that. It has to do
with the fact that Linux systems use GNU software for nearly
all basic utilities, including the libraries.

> No reason. But since my first interpretation of RMS's idea is
>that the FSF will take credit for work they've not done, it seems
>like it's not the most politic way to ask for recognition.

The FSF is not taking credit for anything that they have not
done. What they HAVE done is provide a damn fine base upon
which to build a complete UNIX system such as Linux. Or did
believe that it was the Linux development team which wrote gcc,
libc, and the dozens of other software tools which your Linux
system depends upon to run?

M. Welsh

Matt Welsh

unread,
Sep 11, 1994, 4:12:08 PM9/11/94
to
In article <34tilt$k...@geraldo.cc.utexas.edu> li...@slip-1-72.ots.utexas.edu (Dances With Geeks) writes:
>On Thu, 8 Sep 1994 14:22:06 GMT, Matt Welsh (m...@cs.cornell.edu) wrote:
>> Sorry, but you're stuck with the "GNU approach" (whatever that means)
>> because you use software and libraries covered by the GPL. Any "problems"
>> perceived with GNU software applies equally to Linux.
>
>Sorry, but you're wrong. The Linux kernel, for example, adds additional
>disclaimers which modify the GNU-format license it is used under.

Those "disclaimers" don't relieve the fundamental problems that
people perceive with the GPL, namely, the fact that (a) source
mustg be provided, and (b) modifications must be copylefted as
well.

M. Welsh

Alan Cox

unread,
Sep 12, 1994, 12:15:07 PM9/12/94
to
In article <CvvtA...@pell.com> o...@pell.com (Orc) writes:
>In article <1994Sep8.1...@cs.cornell.edu>,
>Matt Welsh <m...@cs.cornell.edu> wrote:
>>RMS's idea (which I have heard first-hand) is that Linux systems
>>should be considered GNU systems with Linux as the kernel.
> Well, that certainly puts a new twist on the GPL. An unpleasant
>one; GPLing things makes them part of the GNU project.

Chuckle.. poor Mr Stallman I guess the Linux community will have to absorb
him rather than the reverse.

>>Why shouldn't GNU receive recognition for this?
> No reason. But since my first interpretation of RMS's idea is
>that the FSF will take credit for work they've not done, it seems
>like it's not the most politic way to ask for recognition.

RMS is RMS. He can think what he likes. Do you think anyone cares ? I don't
see what the difference between GNU using Linux and anyone else using Linux
is apart from its more good publicity for us and makes their hurd project
look rather silly 8)

Alan

Orc

unread,
Sep 12, 1994, 11:04:47 PM9/12/94
to
In article <1994Sep11.2...@cs.cornell.edu>,

Matt Welsh <m...@cs.cornell.edu> wrote:
>In article <CvvtA...@pell.com> o...@pell.com (Orc) writes:
>>In article <1994Sep8.1...@cs.cornell.edu>,

>> No reason. But since my first interpretation of RMS's idea is


>>that the FSF will take credit for work they've not done, it seems
>>like it's not the most politic way to ask for recognition.
>
>The FSF is not taking credit for anything that they have not
>done.

True. note that I said "first interpretation": when you said

> RMS's idea (which I have heard first-hand) is that Linux systems
> should be considered GNU systems with Linux as the kernel.

the very first thing that popped into my head, even armed with
the knowledge that RMS likes to make Pronouncements, was that FSF
was trying to take credit for Linux. And I *like* the FSF, and
support the work they're doing.

>Or did
>believe that it was the Linux development team which wrote gcc,
>libc, and the dozens of other software tools which your Linux
>system depends upon to run?

Now I'd have to disagree with this. Aside from gcc and libc, I
can get stuff from multiple sources, not just the FSF. I have at
least as much Berkeley code on my linux box as I do FSF code --
if I was to convert over to FreeBSD, the only substantive change
would be the kernel, and that's certainly not a "GNU system with
BSD as the kernel"

The GNU stuff is good, with very few bugs and gotchas, and the
development of Linux would be quite different without having gcc
around. And the [L]GPL is a nice idea, which I approve of, and
would be happy to use if it wasn't for this recent exchange of
messages. But using <x> tools doesn't make what you wrote also an
<x> tool.

____
david parsons \bi/ o...@pell.com
\/

Stephen Harris

unread,
Sep 12, 1994, 5:05:29 PM9/12/94
to
Matt Welsh (m...@cs.cornell.edu) wrote:
: This is absolutely wrong. This viewpoint has nothing to do with

: the fact that Linux is covered by the GPL. If you had bothered
: to read my message, you would have known that. It has to do
: with the fact that Linux systems use GNU software for nearly
: all basic utilities, including the libraries.

Here is where one problem in this thread is happening:
LINUX can refer to
1. The Linux kernel.
This is GPL'd but is _not_ GNU, and FSF have no claim on it (and nor do
I believe they are trying to do so).
2. A Linux distribution.
Generally these are based on GNU tools, but there is _no_ requirement
that this has to be so (excepting gcc I guess). The majority of
programs could be replaced with BSD ones. Even the libs could be
replaced with BSD libc if you wanted to go to that effort.

Now, there is nothing to stop the FSF providing a Linux distribution of their
own, and calling it GNU-Linux. Maybe Debian would even let their stuff be
called Debian GNU-Linux, and the FSF could distribute that. Those tools with
sources based on non-GPL code might cause them some headaches unless they are
willing to distribute non-GPL source as well :-)

The one thing that FSF should be aware of is the conflicting usage of the
term 'Linux' and be guided accordingly.
--

rgds
Stephen

Dances With Geeks

unread,
Sep 14, 1994, 7:55:35 AM9/14/94
to

> M. Welsh

Um, perhaps you weren't listening. Someone may see the problems you listed
as fundamental problems; not all of us do. Some of us see the "fundamental
problems as being elsewhere, and having been addressed by the additional
disclaimers.... ;)

Oh, BTW, these arguments might go on less interminably if everyone adopted a
less confrontational style. ;)


lilo

Marcus Daniels

unread,
Sep 15, 1994, 9:30:15 AM9/15/94
to
In-reply-to: iia...@iifeak.swan.ac.uk's message of Thu, 15 Sep 1994 11:52:11 GMT

>>>>> "Alan" == Alan Cox <iia...@iifeak.swan.ac.uk> writes:
In article <Cw66A...@info.swan.ac.uk> iia...@iifeak.swan.ac.uk (Alan Cox) writes:

[]
Alan> - How do inline functions in include files related to the GPL

Alan> - What about structures from include files

Alan> - When is a dynamic link a dynamic link
[]
Alan> A GPL v3 seems needed for at least one of these issues.

Indeed, it may be necessary to address these issues in a broad way.
However, it becomes easy to paint yourself into a corner by specifying legal
interpretations for each and every techinical convention. It is easy
to see why the FSF has avoided this level of detail.

The examples you cite are specific to the C/object-module
world. These terms don't apply so well to reflective development environments.

Developers should be able to tune the L/GPL to their personal intent, making
minor exceptions just as the FSF does. I think the way to do this is with
explicit agreements with users.

Joe Buck

unread,
Sep 15, 1994, 2:48:21 PM9/15/94
to
m...@churchy.gnu.ai.mit.edu (Michael I Bushnell) writes:
> - What about structures from include files
>
>Again, this depends on what the courts say is a "derived work". It
>seems likely that use of data structure layouts does not constitute
>derivation, at the very least, if you compile without symbols.

In the AT&T vs BSDI/UCB case, the judge found that use of identical
structures was not a copyright violation because it was basically
a compatibility requirement. Since this was only preliminary it's not an
official precedent. Anyway, for FSF to assert control over data
structures even in GPLed code would be to assert an interface copyright.


>With the LGPL, there is a special exemption for data structure layouts.


>
> - When is a dynamic link a dynamic link
>

>The FSF's legal documents never use this term, so it's definition
>doesn't matter.

In a way, this is unfortunate. A lot of the uncertainty over the LGPL
is because people say "well, it looks like a dynamic link would satisfy
the requirements, but it doesn't say so specifically so I'm not sure".

MacGyver

unread,
Sep 16, 1994, 4:43:13 AM9/16/94
to
I've been following this thread since its inception, and I really would like
to know something -- do we REALLY need all these restrictions and license
agreements? Personally, I think the notion has been blown WAY out of
proportion by the folks at FSF and a couple of other places. Placing a
copyright on code you wrote is definitely something worthwhile and its not
something most people dispute, but, to FORCE people who use a certain package
to have to abide by any special rules is completely ridiculous. Such practices
force people to look elsewhere for packages or libraries, often wasting many
precious man hours because a few people decided that they released something
and put it under licensing terms that essentially enforce their views on
source availiblity, and they get away with it!

I would argue against most forms of licensing, especially the GPL/LGPL. I feel
that if someone releases something with source code (unless it's specifically
Shareware or such), they are effectively saying 'Ok, I designed this package,
feel free to use it/modify it, but give credit to the author(s) if you use this
package/library'. Such an attitude doesn't, in my mind, infringe on the
author's rights, so long as he/she is give the appropriate credit. It's
simple, it's clean, and it makes a lot of sense, so then why do people at FSF
insist on making things difficult?

HJD.

Alan Cox

unread,
Sep 15, 1994, 7:52:11 AM9/15/94
to
In article <1994Sep11.2...@cs.cornell.edu> m...@cs.cornell.edu (Matt Welsh) writes:
>Those "disclaimers" don't relieve the fundamental problems that
>people perceive with the GPL, namely, the fact that (a) source
>mustg be provided, and (b) modifications must be copylefted as
>well.

Well the problem is what constitutes an interface and thats something GNU
do need to clean up. There are numerous other little problems with it like
any other license

eg - Can BSD code be incorporated and redistributed as part
of a GPL program: Specifically does the BSD license
requirement for credits in the documentation etc count
as an 'additional restriction' on distribution as the GPL
requires there are none. If it does (as my legal info says)
then you can't mix BSD with GPL and distribute the result.
If it doesn't can I therefore sell software GPL'd but with
a documentation requirement of must come with this expensive
booklet I've written ?

- How do inline functions in include files related to the GPL

- What about structures from include files

- When is a dynamic link a dynamic link

- Interface issues. If I build a system that can use an LGPL
shared library I don't have to give a damn about the LGPL -
It hasn't got any gnu code in it and if people choose to
plug the two together its up to them.

A GPL v3 seems needed for at least one of these issues.

Alan

Stephen J Bevan

unread,
Sep 16, 1994, 8:56:36 AM9/16/94
to
In article <35blr1$c...@Venus.mcs.com> macg...@MCS.COM (MacGyver) writes:
I've been following this thread since its inception, and I really would like
to know something -- do we REALLY need all these restrictions and license
^^^^^^^^^^^^
agreements? ...

... I feel that if someone releases something with source


code (unless it's specifically Shareware or such), they are
effectively saying 'Ok, I designed this package, feel free to use
it/modify it, but give credit to the author(s) if you use this

^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
package/library'.
^^^^^^^^^^^^^^^

Am I right in assuming that the "give credit" part is just a
suggestion and not a restriction i.e. it is ok by you if someone takes
the source and passes it off as their own work?

Michael I Bushnell

unread,
Sep 15, 1994, 11:24:55 AM9/15/94
to
In article <Cw66A...@info.swan.ac.uk> iia...@iifeak.swan.ac.uk (Alan Cox) writes:

eg - Can BSD code be incorporated and redistributed as part
of a GPL program: Specifically does the BSD license
requirement for credits in the documentation etc count
as an 'additional restriction' on distribution as the GPL
requires there are none. If it does (as my legal info says)
then you can't mix BSD with GPL and distribute the result.
If it doesn't can I therefore sell software GPL'd but with
a documentation requirement of must come with this expensive
booklet I've written ?

In the United States, the anwser is that there is apparently no
problem, because the BSD credit requirement is not thought to be
legally enforceable by most attorneys. However, this is possibly not
the case in other countries (notably, the UK).

It *is* an additional restriction (if legally binding) and it does
pose a problem. Given the good will of Berkeley, while it's a problem
that needs to be addressed (and rms is looking into various ways of
doing so); nobody is going to get sued on either side.

People might think that the GPL should make a special exception for
this. (I did at first.) But actually, the Berkeley requirement (if
enforceable) really is inimical to the GPL. Imagine the situation
after twenty--or a hundred--people have added their own credit
requirements. It would become essentially impossible to advertise GNU
software at all.

- How do inline functions in include files related to the GPL

GPL: If the inline functions are big/complex enough to make the result
a derived work, then the GPL applies as a whole to the resulting work.
The meanings of "big" and "complex" are undecided by the law at this
point.

LGPL: The same thing applies, except that the LGPL (*not* the GPL)
excludes from worry functions less than ten lines long, whether or not
they cause derivative status.

- What about structures from include files

Again, this depends on what the courts say is a "derived work". It


seems likely that use of data structure layouts does not constitute
derivation, at the very least, if you compile without symbols.

With the LGPL, there is a special exemption for data structure layouts.

- When is a dynamic link a dynamic link

The FSF's legal documents never use this term, so it's definition
doesn't matter.

- Interface issues. If I build a system that can use an LGPL


shared library I don't have to give a damn about the LGPL -
It hasn't got any gnu code in it and if people choose to
plug the two together its up to them.

If you distribute .o files (but not fully linked executables) which
only include data structure layouts and short (< 10 lines) functions
from header files, then you are on safe ground, with an LGPL'd
library.

-mib

MacGyver

unread,
Sep 16, 1994, 2:32:35 PM9/16/94
to
Stephen J Bevan (be...@cs.man.ac.uk) wrote:

Not at all. I would say that they MUST, in fact, give the proper
acknowledgement. I should have perhaps rephrased what I said to something
like 'It is reasonable, and I would say in fact required, to give appropriate
acknowledgement to the author'. My apologies for the abiguity.

HJD.

Anselm Lingnau

unread,
Sep 16, 1994, 6:13:28 AM9/16/94
to
In article <35blr1$c...@venus.mcs.com>, MacGyver <macg...@MCS.COM> wrote:

> I've been following this thread since its inception, and I really would like
> to know something -- do we REALLY need all these restrictions and license

> agreements?

Yes, we do need license agreements. International copyright (as per the
Berne Convention) says that stuff like books, articles, ... is
copyrighted as soon as it is written, and at least over here in Germany
programs are mostly like literary works for the purposes of copyright.

So if you write a program, the copyright is yours. You may keep the
code, give it away or sell it, but the people who received the program
from you may not do *anything at all*, especially in the way of
redistributing it, without you *allowing* them to do so. That is
basically what a `license' is all about -- the word derives from the
Latin where `licet' means `it is allowed'. The fact that most commercial
license agreements are about forbidding redistribution is beside the
point; they go through intricate gyrations to make sure you're only
allowed to *use* the stuff but not to pass it on.

Anselm
--
Anselm Lingnau ......................... lin...@tm.informatik.uni-frankfurt.de
When I am working on a problem, I never think about beauty. I think only how to
solve the problem. But when I have finished, if the solution is not beautiful,
I know it is wrong. --- R. Buckminster Fuller

Marcus Daniels

unread,
Sep 16, 1994, 6:04:53 AM9/16/94
to
In-reply-to: macg...@MCS.COM's message of 16 Sep 1994 03:43:13 -0500


>>>>> "mg" == MacGyver <macg...@MCS.COM> writes:
In article <35blr1$c...@Venus.mcs.com> macg...@MCS.COM (MacGyver) writes:

[ It does occur to me this post may be from the keyboard of a
devilishly dry wit, but, sorry, I simply lack the humor to let it slide. ]

mg> I've been following this thread since its inception, and I really
mg> would like to know something -- do we REALLY need all these
mg> restrictions and license agreements?

The intent of all GPL developers is to encourage cooperation.
To further this goal, full source distribution is necessary.
Although there is valid debate about what exact terms are legal and
realistic, there are apparently a growing number of users who
Just Don't Get It.

mg> Personally, I think the
mg> notion has been blown WAY out of proportion by the folks at FSF
mg> and a couple of other places. Placing a copyright on code you
mg> wrote is definitely something worthwhile and its not something
mg> most people dispute, but, to FORCE people who use a certain
mg> package to have to abide by any special rules is completely
mg> ridiculous.

Indeed! How absurd to think that people who devote hundreds or even
thousands hours of their own time to projects they get little or no
remuneration for should have any say about how their software is used!

What are these wiggy FSF people thinking?! Can you _imagine_?

mg> Such practices force people to look elsewhere for
mg> packages or libraries, often wasting many precious man hours

Shrug. Nine times out of ten, I'm more than happy to participate in the
inhibition of individuals who would be actively seeking to circumvent
licenses like the GPL. Handy litmus test.

mg> because a few people decided that they released something and put
mg> it under licensing terms that essentially enforce their views on
mg> source availiblity, and they get away with it!

;(

Marcus Daniels

unread,
Sep 16, 1994, 9:55:35 PM9/16/94
to
macg...@MCS.COM (MacGyver) writes:

>Encouraging cooperation is fine and dandy, HOWEVER, I do not feel that
>enforcing restrictions like 'you link with our library and you must
>release the source' fosters cooperation at all. Cooperation should NOT
>be dependant on the release of source, mandated by the GPL.

Source code is a necessary but not sufficient condition for
cooperation to occur. How is a developer contributing to the success
of other developers unless they volunteer their sources?
The only way I can think of is to make them pay.
Perhaps a pay-through-the-teeth GHL (GNU Hoarding License) is in order.

>Cooperation
>should be collaboration and discussion among developers, who share source
>code on their own, without the gun of the GPL in their backs.

Implicit in this kind of attitude is that free software developers
are complusive programmers, and that they will Create whether
or not cooperation occurs. Hence, they can safely be exploited.

>Even
>releasing the source is no guarantee of cooperation for the simple reason that
>some people's knowledge of whatever language was used to develop the project
>might not be as well developed as others, and it does them no good to have
>the source, but not understand a thing about it.

There will always be the interminably ignorant people who simply
refuse to learn. So? Source code levels the playing field.

Marcus Daniels

unread,
Sep 17, 1994, 4:31:01 AM9/17/94
to
du...@telerama.lm.com (Jim Duke) writes:

>We (the unix community in general) are fortunate that people such as the
>MIT X Windows Developers (now the X Consortium) did not take such a
>limiting position about their software or most unix user would still
>be using character systems.

Or, more optimistically, Unix and X11 would have died, and we'd all be
using an O.S. and GUI based on Smalltalk or Scheme.

Or, supposing X Windows was GPL'ed, it might have turned more
people on to Free software. More time and money would be spent on
development and support of Free software, and less time and money on
beating each other up.

My point? We are both speculating. Moot point.

>> mg> Such practices force people to look elsewhere for
>> mg> packages or libraries, often wasting many precious man hours

>> Shrug. Nine times out of ten, I'm more than happy to participate in the
>> inhibition of individuals who would be actively seeking to circumvent
>> licenses like the GPL. Handy litmus test.

>mg - How true, I can't recall how many times I have seen something useful
>that can't be used by me because of it's GPL or LPGL licensing. I can
>understand the reasons for their (FSF) terms but feel that in general
>GPL and LGPL have stifled at least as much technology as they have fostered.

>Case in point, I have many useful libraries, utilities, programs etc
>(Mostly X/Motif stuff) which I have developed over the years which
>have been limited by GPL in a couple of ways.

> 1) All home development is for the eventual purpose of enhancing
> my work abilities, therfore any GPL or LGPL stuff cannot be
> incorporated into my home work (I do not have time to learn
> anything that will not transfer directly to work).

You say `learn'? Curious. Do you think you are the only one
with limited time? Don't you think that everyone has a body
of their own professional work that couldn't be GPL'ed (because
you don't own it, or you don't have time to package it properly,
or it just is idiosyncratic and of low quality, etc.).

> 2) Just as FSF cannot in good consience let their work be used
> for un-restricted commercial use (such as the way X is used)
> I can't allow anything that I would create be used to further
> the goals of FSF or GPL.

Okay, like many of us, you rely on commercial exploitation to live.
Come on -- kind of transparent.

>Why should FSF care... well they shouldn't they don't lose out, the net
>community in general does.

I've won in major way. Linux actually gives Unix a good name. I've
got a solid development environment (Emacs, GCC, CLISP, etc.).
All with complete consistent source (thanks to the FSF's leadership).
A true joy.

The GPL doesn't say anything about technical superiority. It doesn't
need to be said. The personality trait that makes hackers want it
`perfect' is so common it doesn't need to be encouraged further.

>Actually I have had an Idea recently that people that feel as strongly
>opposed to the GPL type licensing as FSF does about the unrestricted
>licensing should start releaseing things with an anti-GPL. It would
>be allowed to be distributed and used freely, with or without source
>for, or not for profit, the only restriction would be that it could
>not be used in conjunction with any GPL tools/systems/utilities.

Where is that Barney animal when you need hir?

Terry Lambert

unread,
Sep 17, 1994, 6:40:51 AM9/17/94
to
In article <MARCUS.94S...@tdb.ee.pdx.edu> mar...@ee.pdx.edu writes:
] The intent of all GPL developers is to encourage cooperation.

] To further this goal, full source distribution is necessary.
] Although there is valid debate about what exact terms are legal and
] realistic, there are apparently a growing number of users who
] Just Don't Get It.

I think people are generally talking about three things:

1) The intent of the license
2) The wording of the license
3) The release of software utilizing the license as an instrument


In doing this, they complain about:

A) Thier personal diesenfranchisement with #1, #2, or #3.
B) Methods whereby #3 may be achieved at the same time as complying
with #2 in the pursuit of a non-aligned side goal.
C) RMS and/or the FSF (Ad Hominim attacks).
D) Apparent conflicts between #1 and #2.


Complaints (A) and (C) are clearly specious.

Complaint (B) is in violation of item (1) but potentially not in
violation of item (2). People who would argue thusly must admit
to the potential validity of complaint (D).

I, for my part, am making complaint (D) and being accused of (A), (B),
and (C). 8-).

I would agree that people makinging complaints (A), (B), or (C) "Just
Don't Get It". Thank you for recognizing complaint (D) as having
potential validity.


Regards,
Terry Lambert
te...@cs.weber.edu
---
Any opinions in this posting are my own and not those of my present
or previous employers.

MacGyver

unread,
Sep 16, 1994, 2:50:34 PM9/16/94
to
Anselm Lingnau (lin...@informatik.uni-frankfurt.de) wrote:

: Yes, we do need license agreements. International copyright (as per the


: Berne Convention) says that stuff like books, articles, ... is
: copyrighted as soon as it is written, and at least over here in Germany
: programs are mostly like literary works for the purposes of copyright.

: So if you write a program, the copyright is yours. You may keep the
: code, give it away or sell it, but the people who received the program
: from you may not do *anything at all*, especially in the way of
: redistributing it, without you *allowing* them to do so. That is
: basically what a `license' is all about -- the word derives from the
: Latin where `licet' means `it is allowed'. The fact that most commercial
: license agreements are about forbidding redistribution is beside the
: point; they go through intricate gyrations to make sure you're only
: allowed to *use* the stuff but not to pass it on.

Thank you for the clarification. In that case, I feel that licenses should
ONLY speak to the point of redistribution and not actual use. By
redistribution, I mean actually distributing that package, rather than say
developing an application and releasing the binary which happens to be
linked with a GPLd library as opposed to someone CHANGING the library and
redistribute it. In the latter case, restrictions and such are very
reasonable in my mind, but, I don't believe that licenses have any place in
dictating the former case.

HJD.

MacGyver

unread,
Sep 16, 1994, 2:46:01 PM9/16/94
to
Marcus Daniels (mar...@ee.pdx.edu) wrote:
: In-reply-to: macg...@MCS.COM's message of 16 Sep 1994 03:43:13 -0500


: >>>>> "mg" == MacGyver <macg...@MCS.COM> writes:
: In article <35blr1$c...@Venus.mcs.com> macg...@MCS.COM (MacGyver) writes:

: [ It does occur to me this post may be from the keyboard of a
: devilishly dry wit, but, sorry, I simply lack the humor to let it slide. ]

Not humor, nor wit, it was a serious response, and I feel a valid point.

: mg> I've been following this thread since its inception, and I really


: mg> would like to know something -- do we REALLY need all these
: mg> restrictions and license agreements?

: The intent of all GPL developers is to encourage cooperation.
: To further this goal, full source distribution is necessary.
: Although there is valid debate about what exact terms are legal and
: realistic, there are apparently a growing number of users who
: Just Don't Get It.

Encouraging cooperation is fine and dandy, HOWEVER, I do not feel that


enforcing restrictions like 'you link with our library and you must
release the source' fosters cooperation at all. Cooperation should NOT

be dependant on the release of source, mandated by the GPL. Cooperation


should be collaboration and discussion among developers, who share source

code on their own, without the gun of the GPL in their backs. Even


releasing the source is no guarantee of cooperation for the simple reason that
some people's knowledge of whatever language was used to develop the project
might not be as well developed as others, and it does them no good to have
the source, but not understand a thing about it.

: mg> Personally, I think the


: mg> notion has been blown WAY out of proportion by the folks at FSF
: mg> and a couple of other places. Placing a copyright on code you
: mg> wrote is definitely something worthwhile and its not something
: mg> most people dispute, but, to FORCE people who use a certain
: mg> package to have to abide by any special rules is completely
: mg> ridiculous.

: Indeed! How absurd to think that people who devote hundreds or even
: thousands hours of their own time to projects they get little or no
: remuneration for should have any say about how their software is used!

Even though this was a rather dry attempt at sarcasm, it is EXACTLY my
point. No one forces them to develop the software, or to release it. Lets
face it, nearly everything you get off the net is in source form because there
are tons of different platforms out there, when software IS released, it is,
in effect, in the public domain (copyright infringement and accreditation being
the obvious exception), and they really SHOULD NOT be able to dictate how it
is used to others. No one forces anybody to contribute anything.

: What are these wiggy FSF people thinking?! Can you _imagine_?

The FSF has a very specific, politcal, agenda, and it is getting to the point
where everyone is afraid to code ANYTHING lest they hear that lawyer knocking
on their door.

: mg> Such practices force people to look elsewhere for


: mg> packages or libraries, often wasting many precious man hours

: Shrug. Nine times out of ten, I'm more than happy to participate in the
: inhibition of individuals who would be actively seeking to circumvent
: licenses like the GPL. Handy litmus test.

Litmus test: The only person in the slough of responses to answer in a
condescending, childlike manner, instead of giving an
intellegent rebutle and resorting to pathetic tactics.

HJD.

MacGyver

unread,
Sep 17, 1994, 2:53:31 AM9/17/94
to
Marcus Daniels (mar...@ee.pdx.edu) wrote:
: macg...@MCS.COM (MacGyver) writes:

: >Encouraging cooperation is fine and dandy, HOWEVER, I do not feel that
: >enforcing restrictions like 'you link with our library and you must
: >release the source' fosters cooperation at all. Cooperation should NOT
: >be dependant on the release of source, mandated by the GPL.

: Source code is a necessary but not sufficient condition for
: cooperation to occur. How is a developer contributing to the success
: of other developers unless they volunteer their sources?
: The only way I can think of is to make them pay.
: Perhaps a pay-through-the-teeth GHL (GNU Hoarding License) is in order.

I don't see release of source code as being necessary for cooperation at all.
Sharing IDEAS is. Implementation is something that varies from one individual
to another, so how is source code necessary for it to occur?

: >Cooperation


: >should be collaboration and discussion among developers, who share source
: >code on their own, without the gun of the GPL in their backs.

: Implicit in this kind of attitude is that free software developers
: are complusive programmers, and that they will Create whether
: or not cooperation occurs. Hence, they can safely be exploited.

True enough. To that, I've had several nasty experiences myself.

: >Even


: >releasing the source is no guarantee of cooperation for the simple reason that
: >some people's knowledge of whatever language was used to develop the project
: >might not be as well developed as others, and it does them no good to have
: >the source, but not understand a thing about it.

: There will always be the interminably ignorant people who simply
: refuse to learn. So? Source code levels the playing field.

Yes, it does level the field, but GPL/FSF is trying to essentially create
an interface copyright, so that they can force their agenda on the public.

HJD.

Marcus Daniels

unread,
Sep 17, 1994, 3:56:29 AM9/17/94
to
macg...@MCS.COM (MacGyver) writes:

>I don't see release of source code as being necessary for cooperation at all.

Umm, okay.

>Yes, it does level the field, but GPL/FSF is trying to essentially create
>an interface copyright, so that they can force their agenda on the public.

Folks keep talking about this, but I have as yet to any complaints
from GPL developers who feel their hard work was stolen. I have,
however, heard many complaints from individuals who want to use
GPL'ed code, but are oddly upset when the authors anticipated
this unfortunate urge some people have to hoard. Quite franky,
IMHO, I think the only thing coercing these unhappy people is the
greed they won't even acknowledge to themselves.

Although I very much hope the FSF can find legal language to secure
GPL licensing terms for future circumstances like PGP/gmp, this is a
issue that really has nothing to do with the GPL itself.

The GPL is straightforward.
Furthermore, the goals of the FSF are self-advertised.

Jim Duke

unread,
Sep 16, 1994, 11:24:31 PM9/16/94
to
Marcus Daniels (mar...@ee.pdx.edu) wrote:
> In-reply-to: macg...@MCS.COM's message of 16 Sep 1994 03:43:13 -0500

> mg> and a couple of other places. Placing a copyright on code you
> mg> wrote is definitely something worthwhile and its not something
> mg> most people dispute, but, to FORCE people who use a certain
> mg> package to have to abide by any special rules is completely
> mg> ridiculous.

> Indeed! How absurd to think that people who devote hundreds or even
> thousands hours of their own time to projects they get little or no
> remuneration for should have any say about how their software is used!

> What are these wiggy FSF people thinking?! Can you _imagine_?

We (the unix community in general) are fortunate that people such as the


MIT X Windows Developers (now the X Consortium) did not take such a
limiting position about their software or most unix user would still
be using character systems.

> mg> Such practices force people to look elsewhere for


> mg> packages or libraries, often wasting many precious man hours

> Shrug. Nine times out of ten, I'm more than happy to participate in the
> inhibition of individuals who would be actively seeking to circumvent
> licenses like the GPL. Handy litmus test.

mg - How true, I can't recall how many times I have seen something useful


that can't be used by me because of it's GPL or LPGL licensing. I can
understand the reasons for their (FSF) terms but feel that in general
GPL and LGPL have stifled at least as much technology as they have fostered.

Case in point, I have many useful libraries, utilities, programs etc
(Mostly X/Motif stuff) which I have developed over the years which
have been limited by GPL in a couple of ways.

1) All home development is for the eventual purpose of enhancing
my work abilities, therfore any GPL or LGPL stuff cannot be
incorporated into my home work (I do not have time to learn
anything that will not transfer directly to work).

2) Just as FSF cannot in good consience let their work be used


for un-restricted commercial use (such as the way X is used)
I can't allow anything that I would create be used to further
the goals of FSF or GPL.

Why should FSF care... well they shouldn't they don't lose out, the net
community in general does.

Actually I have had an Idea recently that people that feel as strongly


opposed to the GPL type licensing as FSF does about the unrestricted
licensing should start releaseing things with an anti-GPL. It would
be allowed to be distributed and used freely, with or without source
for, or not for profit, the only restriction would be that it could
not be used in conjunction with any GPL tools/systems/utilities.

(Note: I hope this message does not seem to inflamatory, I am trying
my best to express my feelings, and maybe to generate some useful
dialog without starting all out flameing)

If you must flame please do so by e-mail as to limit bandwidth waste
(I have already contributed enough to that)


--
Jim Duke | These opinions are my own, any
du...@telerama.lm.com | resemblance to the opinions of
Unix/X/Motif Consultant | others is strictly coincidental.

Tim Smith

unread,
Sep 17, 1994, 10:38:20 PM9/17/94
to
Michael I Bushnell <m...@churchy.gnu.ai.mit.edu> wrote:
>In the United States, the anwser is that there is apparently no
>problem, because the BSD credit requirement is not thought to be
>legally enforceable by most attorneys. However, this is possibly not

Why would it not be enforceable?

--Tim Smith

Juana Moreno

unread,
Sep 18, 1994, 6:00:04 AM9/18/94
to
du...@telerama.lm.com (Jim Duke) writes:

>Actually I have had an Idea recently that people that feel as strongly
>opposed to the GPL type licensing as FSF does about the unrestricted
>licensing should start releaseing things with an anti-GPL. It would
>be allowed to be distributed and used freely, with or without source
>for, or not for profit, the only restriction would be that it could
>not be used in conjunction with any GPL tools/systems/utilities.

I like your idea. If you release something under this anti-GPL license,
please let me know. I will take it, make a few modifications, sell it
AND use the modified version along with GPL-ed programs. And if you
really want to enforce your license, you'll need to make your wording as
much or even more bizarre than the GPL one. So, just hire a good lawyer and
do it, please. Maybe even someone could write an anti-(anti-GPL) license?

Seriously now. If you like the GPL, then use it. If not, leave them alone
and mind your own business. Stop doing politics in a comp. group.

Marcus Daniels

unread,
Sep 18, 1994, 8:53:21 AM9/18/94
to
Developer: "The GPL and the FSF have absurd goals that I don't care about."

GPL'er: "Then don't derive your work from GPL'ed software."

Developer: "But commercial software is too expensive and is inflexible."

GPL'er: "Then help to stamp out software hoarding."

Developer: "The GPL and FSF have absurd goals that I don't care about."

Rinse, repeat.

Patrick J. Volkerding

unread,
Sep 18, 1994, 1:52:30 PM9/18/94
to
In article <35h334$g...@gandalf.rutgers.edu>,

Juana Moreno <mad...@gandalf.rutgers.edu> wrote:
>I like your idea. If you release something under this anti-GPL license,
>please let me know. I will take it, make a few modifications, sell it
>AND use the modified version along with GPL-ed programs. And if you
>really want to enforce your license, you'll need to make your wording as
>much or even more bizarre than the GPL one. So, just hire a good lawyer and
>do it, please. Maybe even someone could write an anti-(anti-GPL) license?

This little blurb from the "file" source is about the closest I've even
seen. I'll reproduce part of it here for everyone's enlightenment:

>This software is not subject to and may not be made subject to any
>license of the American Telephone and Telegraph Company (AT&T Inc.),
>UNIX System Laboratories (USL Inc.), Novell Inc., Sun Microsystems
>Inc., Digital Equipment Inc., Lotus Development Inc., the Regents of
>the University of California, The X Consortium or MIT, or The Free
>Software Foundation.

Clearly there are people out there who aren't big fans of the GPL, and I
guess if I released something under a different (less restrictive?) license
and someone changed it a little and then GPLed it, I'm not sure how I'd
like it. FWIW, I have GPLed some of my own software before, but I think the
author has the right to say "derivatives of this software must *never*
be GPLed" just like the FSF says "derivatives of GPLed software must
*always* be GPLed".

Pat

Per Abrahamsen

unread,
Sep 18, 1994, 4:44:49 PM9/18/94
to

>>>>> "Marcus" == Marcus Daniels <mar...@ee.pdx.edu> writes:

Marcus> Folks keep talking about this, but I have as yet to any complaints
Marcus> from GPL developers who feel their hard work was stolen. I have,
Marcus> however, heard many complaints from individuals who want to use
Marcus> GPL'ed code, but are oddly upset when the authors anticipated
Marcus> this unfortunate urge some people have to hoard. Quite franky,
Marcus> IMHO, I think the only thing coercing these unhappy people is the
Marcus> greed they won't even acknowledge to themselves.

The package in question was itself freeware, and I doubt greed was a
motive for optionally making it support the GMP package, nor do I
think it was intended as a smart way to get around the GPL.

Whatever can be said about FSF's legal merit in the case, tactically
and politically it was incredible stupid. They would have found much
more sympathy if they have chosen to go after Pearl Software instead.

[ Pearl Software sells a modified version of Emacs for Windows that
uses a proprietary X11 emulation DDL also from Pearl. You can get the
source for the modified Emacs, but not for the DDL. ]

Nate Williams

unread,
Sep 18, 1994, 5:52:29 PM9/18/94
to
[ Jumping into the fray. ]

In article <MIB.94Se...@churchy.gnu.ai.mit.edu>,


Michael I Bushnell <m...@churchy.gnu.ai.mit.edu> wrote:

>In article <Cw66A...@info.swan.ac.uk> iia...@iifeak.swan.ac.uk (Alan Cox) writes:
>
> eg - Can BSD code be incorporated and redistributed as part
> of a GPL program: Specifically does the BSD license
> requirement for credits in the documentation etc count
> as an 'additional restriction' on distribution as the GPL
> requires there are none. If it does (as my legal info says)
> then you can't mix BSD with GPL and distribute the result.
> If it doesn't can I therefore sell software GPL'd but with
> a documentation requirement of must come with this expensive
> booklet I've written ?
>
>In the United States, the anwser is that there is apparently no
>problem, because the BSD credit requirement is not thought to be
>legally enforceable by most attorneys. However, this is possibly not
>the case in other countries (notably, the UK).

Having the utmost respect for Michael and the work he has done, I must
jump in here and ask this question.

In many articles people have said that "many lawyers consider portions
of the GPL/LGPL unenforceable', and you counter with "Obviously, the FSF
lawyers and others have not seen fit to test it's legality out, so it
must be enforceable, yet in this article you present the claim the same
accusations on the BSD copyright, which was developed by lawyers of
Berkeley and the X Consortium. By stating that you don't believe it to
be enforceable, you are (IMHO) invalidating your defense of the GPL.

If that offense/defense of the GPL/BSD copyright is invalid in one case,
then it must be invalid in all cases if the situation is similar, which
I believe it is. Both groups hired lawyers to setup the clause, and the
BSD copyright was 'sort-of' tested in the counter-suit filed by UCB against
USL in the recent BSDI-USL copyright suit. UCB's suit was that USL did not
give BSD due credit where the licensing scheme required it.

All I ask for is consistancy in your defense and accusations.


Nate

--
na...@bsd.coe.montana.edu | FreeBSD core member and all around tech.
na...@cs.montana.edu | weenie.
work #: (406) 994-4836 |
home #: (406) 586-0579 | Available for contract/otherwise work.

Marcus Daniels

unread,
Sep 18, 1994, 6:10:45 PM9/18/94
to
abr...@iesd.auc.dk (Per Abrahamsen) writes:

>Whatever can be said about FSF's legal merit in the case, tactically
>and politically it was incredible stupid. They would have found much
>more sympathy if they have chosen to go after Pearl Software instead.

The former had legal merit and the latter didn't. A DLL that
uses an X11 interface. What is the problem? As I recall there
was considerable dialogue about the Perl Software Emacs.

Alan Cox

unread,
Sep 19, 1994, 7:12:13 AM9/19/94
to
In article <35dnhf$r...@terrazzo.lm.com> du...@telerama.lm.com (Jim Duke) writes:
>We (the unix community in general) are fortunate that people such as the
>MIT X Windows Developers (now the X Consortium) did not take such a
>limiting position about their software or most unix user would still
>be using character systems.

I doubt it. It would probably have kicked the kinks out of the GPL/LGPL but
it would have stopped people hiding tricks and dealt with some of the
'we don't tell you how our video card works' issues.

>mg - How true, I can't recall how many times I have seen something useful
>that can't be used by me because of it's GPL or LPGL licensing. I can
>understand the reasons for their (FSF) terms but feel that in general
>GPL and LGPL have stifled at least as much technology as they have fostered.

How shall I put this. If the person who wrote it believes in the GPL concept
they can use it. If you don't like that fact then negotiate an alternative
license with them and pay them lots of money.

> 2) Just as FSF cannot in good consience let their work be used
> for un-restricted commercial use (such as the way X is used)
> I can't allow anything that I would create be used to further
> the goals of FSF or GPL.

Fine. You aren't helping the FSF, they aren't helping you. Seems a fair
nonexchange.

>Actually I have had an Idea recently that people that feel as strongly
>opposed to the GPL type licensing as FSF does about the unrestricted
>licensing should start releaseing things with an anti-GPL. It would
>be allowed to be distributed and used freely, with or without source
>for, or not for profit, the only restriction would be that it could
>not be used in conjunction with any GPL tools/systems/utilities.

This has been suggested before. It wouldn't get you very far. Nobody
stops you using different licenses for stuff _YOU_ wrote. When it comes
to other peoples software play by their rules as you expect them to play
by yours.

ron house

unread,
Sep 19, 1994, 10:22:19 PM9/19/94
to
macg...@MCS.COM (MacGyver) writes:

>I've been following this thread since its inception, and I really would like
>to know something -- do we REALLY need all these restrictions and license
>agreements?

I think we _do_ need protection in the form of licence agreements to protect
access to source. My objection is only to the extreme form of
the GNU licence, not to the basic idea of licences.

>I feel
>that if someone releases something with source code (unless it's specifically
>Shareware or such), they are effectively saying 'Ok, I designed this package,
>feel free to use it/modify it, but give credit to the author(s) if you use this
>package/library'.

What they are saying is up to them, not you. That's why there are licences -
to clarify what they are saying. But what I ask authors to do is ask
themselves whether they really _want_ to say what the GNU licences say,
given their basis in a very strange concept of (lack of) property rights.

--

Ron House. USQ
(ho...@usq.edu.au) Toowoomba, Australia.

Michael I Bushnell

unread,
Sep 20, 1994, 12:25:39 PM9/20/94
to

I don't know the specific reasons; perhaps someone else could explain.
It also might only apply in the US, so it's either a question of
public policy or perhaps a first amendment restriction.

Michael I Bushnell

unread,
Sep 20, 1994, 12:30:46 PM9/20/94
to
In article <35icqt$a...@pdq.coe.montana.edu> na...@bsd.coe.montana.edu (Nate Williams) writes:

In many articles people have said that "many lawyers consider portions
of the GPL/LGPL unenforceable', and you counter with "Obviously, the FSF
lawyers and others have not seen fit to test it's legality out, so it
must be enforceable, yet in this article you present the claim the same
accusations on the BSD copyright, which was developed by lawyers of
Berkeley and the X Consortium. By stating that you don't believe it to
be enforceable, you are (IMHO) invalidating your defense of the GPL.

If that offense/defense of the GPL/BSD copyright is invalid in one case,
then it must be invalid in all cases if the situation is similar, which
I believe it is. Both groups hired lawyers to setup the clause, and the
BSD copyright was 'sort-of' tested in the counter-suit filed by UCB against
USL in the recent BSDI-USL copyright suit. UCB's suit was that USL did not
give BSD due credit where the licensing scheme required it.

All I ask for is consistancy in your defense and accusations.

I'm sorry. It's not the general idea of attaching conditions like
this to the copyright. It's very specifically the "give credit in all
advertisement" clause. This one clause is probably unenforceable,
according to the FSF's (and AT&T's, as it happens) lawyers.

If a license said "this software may not be distributed to blacks",
that would be an unenforceable clause (on the grounds that it is
contrary to public policy) but not because attaching conditions is
bad--just because the of the specific condition.

I don't know if public policy is the reason the credit requirement is
believed to be unenforceable; I just bring this up to explain that the
offense is not the manner in which the thing is done, but the nature
of the thing that is done.

As for UCB's suit, notice that they didn't win: under the consent
decree AT&T voluntarily agreed to give credit, but there was no hint
of a finding that they had to.

J.J. Paijmans

unread,
Sep 20, 1994, 3:49:17 PM9/20/94
to
In article <MIB.94Se...@churchy.gnu.ai.mit.edu> m...@churchy.gnu.ai.mit.edu (Michael I Bushnell) writes:
...

>
>
>--
>+1 617 623 3248 (H) | En arche en ho logos,
>+1 617 253 8568 (W) -+- kai ho logos en pros ton theon,
>1105 Broadway | kai theos en ho logos.
>Somerville, MA 02144 | Kai ho logos sarx egeneto,
>m...@gnu.ai.mit.edu | kai eskenosen en hemin.

With regards to your signature: Gamo to theo sou;
(sorry, couldn't let that one pas..."
Paai.

Michael I Bushnell

unread,
Sep 21, 1994, 12:46:19 AM9/21/94
to
Distribution:

I copied my signature from my Nestle-Aland Greek New Testament. But
my Greek is exceedingly poor. So what is a translation of your bit?

Alan Cox

unread,
Sep 21, 1994, 6:30:10 AM9/21/94
to
In article <35huou$4...@nkosi.well.com> go...@magnet.mednet.net (Patrick J. Volkerding) writes:
>like it. FWIW, I have GPLed some of my own software before, but I think the
>author has the right to say "derivatives of this software must *never*
>be GPLed" just like the FSF says "derivatives of GPLed software must
>*always* be GPLed".

Unless you explicitly permit a license that is GPL or less restrictive than
the GPL then people can't GPL your material. For example if you said
'May not be used for commercial purposes' then it couldn't get mixed with
GPL'd code. Other variants I've see are 'May not be distributed under any
license other than this one' - which is close to what the GPL does save
that it permits you to extract freer than GPL components that are
clearly seperable (eg a PD module in a Linux kernel).

Chris Bitmead

unread,
Sep 23, 1994, 10:48:23 AM9/23/94
to
In article <35huou$4...@nkosi.well.com> go...@magnet.mednet.net (Patrick J. Volkerding) writes:

If you did this then I would just make a minor modification to the GPL,
and call it say the "APL" and put you software under that.

Stephen J Bevan

unread,
Sep 25, 1994, 4:20:21 AM9/25/94
to
[ Sorry for the slow response, I'm just getting articles over a week
old - bevan ]

In article <35coc3$k...@Venus.mcs.com> macg...@MCS.COM (MacGyver) writes:
: Am I right in assuming that the "give credit" part is just a
: suggestion and not a restriction i.e. it is ok by you if someone takes
: the source and passes it off as their own work?

Not at all. I would say that they MUST, in fact, give the proper
acknowledgement. I should have perhaps rephrased what I said to something
like 'It is reasonable, and I would say in fact required, to give
appropriate acknowledgement to the author'. My apologies for the
abiguity.

I still don't understand how this restriction is consistent with what
you wrote in your original message, the relevant part being :-

>... Placing a copyright on code you wrote is definitely something
>worthwhile and its not something most people dispute, but, to FORCE
>people who use a certain package to have to abide by any special
>rules is completely ridiculous. ...

Isn't your restriction a "special rule" which is "completely ridiculous"?

0 new messages