gnu-misc-discuss
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: GPL and other licences


From: Alexander Terekhov
Subject: Re: GPL and other licences
Date: Tue, 14 Feb 2006 21:30:18 +0100

Alexander Terekhov wrote:
> 
> [... "derived work" (i.e. "derivative work" under GNU law) ...]
> 
> I suppose that id "lrosen" belongs to http://www.rosenlaw.com/rosen.htm.
> 
> Nice to see both Hollaar and Rosen commenting GNU legal nonsense
> version three. (Note that the GPLv2 contains the same GNU definition of
> "derivative work".)

Interesting things are going on over there at gplv3.fsf.org.

http://gplv3.fsf.org/comments/rt/readsay.html?Query=%20Creator%20=%20'lrosen'%20%20AND%20'CF.NoteUrl'%20LIKE%20'gplv3-draft-1'%20&Order=DESC&OrderBy=id&Rows=

It appears that Rosen was (?is?) on the Committee A. He identified a 
bunch of issues (including his comments and a bunch of comments made
by others) and claimed them for the Committee A. Now, just a few days 
later someone "fontana" downgraded and removed all that stuff from 
docket for Committee A. 

I suppose that "fontana" belongs to Moglen's underling at SFLC
Richard Fontana.

http://www.softwarefreedom.org/team.html

I knew that gplv3 process was destined to deliver first class 
circus... and it turns out to be just stunning. ;-)

regards,
alexander.

------
Comment 641: Incompatibility and attorney's fees

Regarding the text: your terms may add limited kinds of additional
requirements on your added parts
In section: gpl3.licensecompat.p1.s1
Submitted by: lrosen on 2006-01-23 at 21:08 EST

Comment noted by lrosen on 2006-01-23 at 21:08 EST:

    Eben reminded me to ask about this: Does the fact that another
license has an attorney's fees provision automatically make it
incompatible with GPLv3? I would much rather that GPLv3 were merely
incompatible with "licenses that contradict the terms of this GPLv3
license for this work," and leave out all the other list of "limited
kinds of additional requirements" below. Otherwise, we'll be arguing
about various kinds of license compatibility provisions forever.

Claimed for Committee A by lrosen on 2006-01-23 at 21:09 EST
Identified as an Issue by lrosen on 2006-01-23 at 21:09 EST
Issue downgraded, moved back to Inbox by fontana on 2006-01-25 at 17:24
EST
Remove from docket for Committee A by fontana on 2006-01-25 at 17:24 EST

Comment 639: Contemplated by whom?

Regarding the text: reasonably contemplated use of the covered work
In section: gpl3.licensingpatents.p0.s2
Submitted by: lrosen on 2006-01-23 at 20:13 EST

Comment noted by lrosen on 2006-01-23 at 20:13 EST:

    What is the reach of the "reasonably contemplated" provision?
Contemplated by whom? As of what date? Must the contemplation be written
somewhere? Most patent licenses are limited to claims that are
"necessarily infringed by" or "embodied in" the software as delivered.
Anything more is risky for holders of large and diverse patent
portfolios.

Claimed for Committee A by lrosen on 2006-01-23 at 20:17 EST
Identified as an Issue by lrosen on 2006-01-23 at 20:17 EST
Issue downgraded, moved back to Inbox by fontana on 2006-01-25 at 17:24
EST
Remove from docket for Committee A by fontana on 2006-01-25 at 17:24 EST

Comment 638: True, but so what?

Regarding the text: Not a Contract.
In section: notacontract.0.0
Submitted by: lrosen on 2006-01-23 at 18:58 EST

Comment noted by lrosen on 2006-01-23 at 18:58 EST:

    NONE of the free and open source licenses are contracts without the
external formalities of offer, acceptance and consideration, and the
GPLv3 IS a contract when those external formalities are
undertaken--despite what the GPLv3 license says. The fact that some
licenses memorialize the contract formation externalities doesn't
actually make them contracts; it is the formalities themselves that do.
Each licensor decides for him/herself whether to require formalities;
the license author cannot decide that on anyone else's behalf. There are
many advantages to forming a contract, including the opportunity for the
licensor to seek contract remedies such as specific performance. There
are no disadvantes to forming a contract with the GPLv3, because the
license terms are still enforceable under either contract law or
copyright law. If a licensor seeks to enforce the GPLv3 under copyright
law rather than contract law, the license then is merely a potential
defense to a claim of infringement. Contract law is the same, except
more alternative remedies exist for licensors besides statutory or
actual damages, or injunction. Why isn't that something good we should
encourage, rather than discourage them with factually inaccurate phrases
such as "Not a Contract"?

Claimed for Committee A by lrosen on 2006-01-23 at 20:16 EST
Identified as an Issue by lrosen on 2006-01-23 at 20:16 EST
Issue downgraded, moved back to Inbox by fontana on 2006-01-25 at 17:24
EST
Remove from docket for Committee A by fontana on 2006-01-25 at 17:24 EST

Child comment of 638:
Comment 259: Should this be removed?

Regarding the text: Therefore, by modifying or propagating the Program
(or any covered work), you indicate your acceptance of this License to
do so, and all its terms and conditions.
In section: gpl3.notacontract.p0.s5
Submitted by: joe on 2006-01-17 at 02:25 EST

Comment noted by joe on 2006-01-17 at 02:25 EST:

    Rational: (1) It is probably unnecessary. (2) It could be
interpreted by some as turning the GNU GPL into a (unilateral?) contract
(yes, even though it is under "NOT A CONTRACT" in big bold letters).

    If this sentence (and others) is merely supposed to explain
copyright law or what is a conclusion of the rest of the license, maybe
such sentences should be highlighted in a way that makes it clear that
they are non-normative (like the preamble) on the basis that it is
helpful to have explanations for the layman without the danger that they
be misinterpreted as part of the licensing terms or add to the
complexity of those terms for readers or users of the GNU GPL.

Comment noted by verdyph on 2006-01-17 at 15:32 EST:

    unilateral contracts are illegal in many countries. The obligations
and rights from one party MUST be reasonnably be balanced by obligations
and rightsbythe other party. As the GPL gives obligations and rights to
the licencee, the full exemption of obligations in the GPL is illegal.

    In my opinion, this is the WEAKEST aspect of the GPL, meaning that
it insufficiently lawfully protects the rights of authors and users face
to other threats like patent claims. Please enforce the GPL so that it
becomes a legal contract with the complete protection of laws and
international treaties (notably at WIPO, WTO, and the related
international treaties like Conventions of Berne, Paris and Washington),
which can be easily defended face to a court (at least in all OECD
countries and in the European Union, including the Internation Court of
Human Rights, and also used as a valid licence to cover the texts and
software applications used in governmental and international
institutions for the public interest).

Made a sub-comment of ticket #638 by lrosen on 2006-01-23 at 20:46 EST
Claimed for Committee A by lrosen on 2006-01-23 at 20:46 EST

Child comment of 638:
Issue 394: Internationalization: copyright licenses are contracts in
some jurisdictions

Regarding the text: Not a Contract
In section: notacontract.0.0
Submitted by: oliva on 2006-01-17 at 16:37 EST

Comment noted by oliva on 2006-01-17 at 16:37 EST:

    A Brazilian lawyer once told me that a copyright license is indeed a
contract in Brazil. It might be a specific kind of unilateral contract
whose name escapes me now, but it is a contract nevertheless.

Comment noted by neroden on 2006-01-18 at 18:20 EST:

    Change the name to "Acceptance by Actions"

Made a sub-comment of ticket #638 by lrosen on 2006-01-23 at 20:44 EST
Claimed for Committee A by lrosen on 2006-01-23 at 20:44 EST
Identified as an Issue by lrosen on 2006-01-23 at 20:44 EST

Child comment of 638:
Comment 463: Not a contract in the U.S. - but others?

Regarding the text: Not a Contract.
In section: notacontract.0
Submitted by: dwheeler1 on 2006-01-17 at 23:12 EST

Comment noted by dwheeler1 on 2006-01-17 at 23:12 EST:

    My understanding is that the GPL is _NOT_ a contract in the U.S. ---
it's a license. But in other countries, I understand that copyright law
is considered a subset of contract law. Is there a way to keep this
concept, yet internationalize it? If nothing else, rename this to "This
is a license, not a contract"; otherwise, in countries where copyrights
ARE contracts, "not a contract" seems self-contradictory. 

Comment noted by pde on 2006-01-23 at 14:42 EST:

    As I understand it, the GPL is not a contract, but an *offer* of a
contract. Once accepted, in the U.S., it is a contract. The thing that
makes the GPLv3 draft unusual amongst contracts is that litigation
following an irreconcilable dispute would most likely be for copyright
infringement, not breach of contract. The case would be fought as breach
of contract only if both parties preferred it that way or if unusual
circumstances forced at least one party onto that ground. GPLv2, with
automatic termination, was even more strongly biased in favour of
dispute settlement under copyright law.

Made a sub-comment of ticket #638 by lrosen on 2006-01-23 at 20:45 EST
Claimed for Committee A by lrosen on 2006-01-23 at 20:45 EST

Child comment of 638:
Comment 528: This heading does not describe the content of the section

Regarding the text: Not a Contract.
In section: notacontract.0.0
Submitted by: stone on 2006-01-18 at 23:04 EST

Comment noted by stone on 2006-01-18 at 23:04 EST:

    Unless the text of the section expressly states that this License is
not a contract, I don't see the point of implicitly making this claim in
the section header. Perhaps `Acceptance as a Condition of Modification
or Propagation' would be better.

Comment noted by terjebr on 2006-01-19 at 03:34 EST:

    I think "Acceptance of Linsence" is better.

    Anyway, I agree that "Not a Contract" is a bad heading. Either it
is, or it is not, and that cannot then be changed by a heading. Even in
some countries copyright law is so that it must be regarded as a
contract, or it is not valid. Then in the US it is best if it is not
regarded as a contract. So it is best for all parties if the License
itself do not try to address the issue.

Comment noted by pde on 2006-01-23 at 14:23 EST:

    Spelling aside, I think terjebr has a point :).

Made a sub-comment of ticket #638 by lrosen on 2006-01-23 at 20:45 EST
Claimed for Committee A by lrosen on 2006-01-23 at 20:45 EST

Child comment of 638:
Comment 628: Consideration

Regarding the text: Not a Contract
In section: notacontract.0.0
Submitted by: weel on 2006-01-22 at 21:09 EST

Comment noted by weel on 2006-01-22 at 21:09 EST:

    Seeing as the GPL is not a contract, wouldn't it, at least in common
law jurisdictions, be revokable? If so, are there any known workarounds?
I understand that there was at some point the idea that you might sell a
license to the FSF for some small amount of consideration, thereby
giving the FSF the rights stated in the GPL, but irrevokably so. The FSF
presumably can be trusted not to revoke licenses. Is such a scheme made
harder by the fact that the license explicitly states that it is not a
contract? [Note: I submitted something like this earlier, but somehow it
doesn't show up.]

Made a sub-comment of ticket #638 by lrosen on 2006-01-23 at 20:55 EST
Claimed for Committee A by lrosen on 2006-01-23 at 20:55 EST

Comment 637: Derivative works

Regarding the text: that is to say, a work containing the Program or a
portion of it, either modified or unmodified.
In section: gpl3.definitions.p0.s2
Submitted by: lrosen on 2006-01-23 at 18:39 EST

Comment noted by lrosen on 2006-01-23 at 18:39 EST:

    The statement beginning "that is to say..." is not an accurate
description of "derivative works" under US copyright law. If you want
the copyleft provisions of GPLv3 to apply to "collective works" then you
should say so explicitly, rather than use language reminiscent of the
definition of collective works when trying to describe derivative works.
The current draft, in this respect, is both ambiguous and potentially
very misleading. See 17 USC 101. 

Claimed for Committee A by lrosen on 2006-01-23 at 20:17 EST
Identified as an Issue by lrosen on 2006-01-23 at 20:17 EST
Issue downgraded, moved back to Inbox by fontana on 2006-01-25 at 17:24
EST
Remove from docket for Committee A by fontana on 2006-01-25 at 17:24 EST

Child comment of 637:
Issue 372: still unclear

Regarding the text: that is to say, a work containing the Program
In section: gpl3.definitions.p0.s2
Submitted by: manuell on 2006-01-17 at 12:43 EST

Comment noted by manuell on 2006-01-17 at 12:43 EST:

    In short, a linux distro contains a GPL'ed kernel but is not a
'derivative' of that kernel. Why keep the 'that is to say...' text ?
(confusion beetween 'derivative' and 'compilation')

Comment noted by neroden on 2006-01-18 at 17:54 EST:

    Yes. Drop this, it confuses people.

Comment noted by mhoyles on 2006-01-19 at 11:33 EST:

    Count me among the confused. I don't know if this is intended to
broaden the meaning of derivative work (to include collections of
software) or if it is mearly a (misleading?) explaination of the legal
term derivative work. It might be better to seperate explainations,
expressions of intent, etc. from the definitions to avoid ambiguity.

Made a sub-comment of ticket #637 by lrosen on 2006-01-23 at 20:39 EST
Claimed for Committee A by lrosen on 2006-01-23 at 20:39 EST
Identified as an Issue by lrosen on 2006-01-23 at 20:39 EST

Child comment of 637:
Issue 466: Can we define 'derivative work' more clearly?

Regarding the text: derivative or collective works based on the Program.
In section: gpl3.distribmod.p5.s1
Submitted by: tungsten on 2006-01-17 at 23:27 EST

Comment noted by tungsten on 2006-01-17 at 23:27 EST:

    I know the term 'derivative work' is a somewhat gray area of the law
in the US at least, but I think this license should define "derivative
work" in a precise way, so that a programmer reading this license can
decide whether his code counts as a derivative work. Right now he has to
ask a lawyer for an opinion (which of course is not definitive), or go
to court to settle the matter.

Comment noted by swillden on 2006-01-18 at 15:18 EST:

    IANAL, but I don't think it's useful for the GPL to clarify
"derivative work". The legal definition is fuzzy, so the GPL can't
clarify it without providing a somewhat different definition... one that
is either weaker or stronger in a given case. I don't think the GPL
wants to weaken the meaning of "derivative work". If anything, it would
better serve the purpose of Free Software to strengthen it a little,
closing some common escape hatches (dynamic linking, etc.). But the GPL
can't actually strengthen the definition of derivative work, because if
a programmer's work is not derivative per the law, then the programmer
doesn't need the GPL and any definitions it provides are moot. So if
there's no legal force behind any clarification that strengthens the
definition, and if we don't want a clarification that weakens the
definition, it seems better to omit the clarification.

Made a sub-comment of ticket #637 by lrosen on 2006-01-23 at 20:41 EST
Claimed for Committee A by lrosen on 2006-01-23 at 20:41 EST
Identified as an Issue by lrosen on 2006-01-23 at 20:41 EST
------


reply via email to

[Prev in Thread] Current Thread [Next in Thread]