11:36 a.m.
This is a HBR cure.
First, for those of you not following the official draft closely, as
of a recent commit, I have audaciously and iconoclastically added this
sentence to the (now single) section on object code disribution:
Each recipient of the Covered Work from You is an intended
third-party beneficiary of this License solely as to this section []
with the right to enforce its terms.
If any of the lawyers on this list have any comments of a strictly
insider-professional-knowledge *legal* nature on this sentence,
including arguments for its deletion or modification, feel free to
speak up (in a HBR-compliant manner of course).
If anyone (non-lawyer or otherwise) thinks it would be bad *policy* to
live in a world where a GPL-like license could be enforced through the
legal system by the obvious immediate beneficiaries of the source code
requirement, feel free to speak up (in a HBR-compliant manner of
course).
That said, on to the HBR cure: I had a conversation with Bradley Kuhn
in which we very briefly discussed this issue, as Bradley was long
under the impression that the GPL could not *possibly* support an
argument that mere licensees have some standing to enforce the source
code requirement (despite the fact, of which Bradley might not be
aware, that over the years some legal academics have at last taken the
question seriously, if skeptically, and some private practice lawyers
[admittedly seeking to profit from the erstwhile open source boom, the
SCO crisis, etc.] purported to take this qite seriously too).
I noted the obvious point that copyleft-next and GPLv2/GPLv3 are not
worded identically, so his comment was partially irrelevant.
Bradley mentioned some interesting work that Australian lawyer Brendan
Scott had done attempting, AIUI, to use Australian consumer protection
law to give a basis for GPL licensees to sue upstream distributors for
failure to provide source code. It is not clear whether such efforts
have been successful.
- RF
12:21 p.m.
On Mon, Feb 18, 2013 at 12:36 PM, Richard Fontana
<fontana(a)sharpeleven.org> wrote:
This is a HBR cure.
First, for those of you not following the official draft closely, as
of a recent commit, I have audaciously and iconoclastically added this
sentence to the (now single) section on object code disribution:
Each recipient of the Covered Work from You is an intended
third-party beneficiary of this License solely as to this section []
with the right to enforce its terms.
If any of the lawyers on this list have any comments of a strictly
insider-professional-knowledge *legal* nature on this sentence,
including arguments for its deletion or modification, feel free to
speak up (in a HBR-compliant manner of course).
I believe it is effective for the task. Problems with trying to claim
the benefit of a contract when one is a third party are generally
because the contract doesn't expressly state that there is a
third-party beneficiary and it is very difficult to win the argument
in the absence of the express provision. So my inclination is that
this works fine although I didn't look anything up so I could be
completely wrong.
If anyone (non-lawyer or otherwise) thinks it would be bad *policy* to
live in a world where a GPL-like license could be enforced through the
legal system by the obvious immediate beneficiaries of the source code
requirement, feel free to speak up (in a HBR-compliant manner of
course).
This is, in theory, a significant poison pill to the use of the
agreement, so it's a policy decision. That said, I suspect that anyone
who isn't sophisticated about copyleft (which is most everyone)
believes that this is already a feature of the GPL licenses. In other
words, the great fear of GPL is that it will "infect" your proprietary
code and you will be forced to release the source code. (Correct me
if I've misunderstood; I think this provision will have that effect.)
So I'm not sure that adding the actual mechanism for it to work will
make the license less palatable to a commercial user of copylefted
code than the GPL already is, although it should.
Pam
1:21 p.m.
On Mon, Feb 18, 2013 at 01:21:16PM -0500, Pamela Chestek wrote:
> If anyone (non-lawyer or otherwise) thinks it would be bad
*policy* to
> live in a world where a GPL-like license could be enforced through the
> legal system by the obvious immediate beneficiaries of the source code
> requirement, feel free to speak up (in a HBR-compliant manner of
> course).
>
This is, in theory, a significant poison pill to the use of the
agreement, so it's a policy decision.
I am aware of this, but I wonder whether there are compensatory
aspects of copyleft-next (or more could be added) to nullify the
increase in poison. Already, let us say up till this change, I
consider copyleft-next to have the appearance of greater commercial
palatability than GPLv2/GPLv3 (this is kind of obvious for GPLv3 since
the anti-lockdown provisions were deleted as were most of the
patent-related provisions). That's not a major goal, but it's a minor
goal which I acknowledged at FOSDEM.
As a related aside, whether commercial vendor (or user, especially in
the case of AGPL-like licenses) sentiment towards FLOSS licenses has
any material impact in their uptake is an unanswered question. A key
assumption made by the FSF in the way it handled its quasi-public
drafting process for GPLv3 (2006-2007) was that in some way this was
so (even though the official view of the FSF remains that 'popularity'
of their licenses is a non-goal). I see no post-2007 evidence to
justify this view one way or another. I could elaborate on that though
it might get a bit off-topic.
That said, I suspect that anyone
who isn't sophisticated about copyleft (which is most everyone)
believes that this is already a feature of the GPL licenses. In other
words, the great fear of GPL is that it will "infect" your proprietary
code and you will be forced to release the source code. (Correct me
if I've misunderstood; I think this provision will have that effect.)
It is not intended to have that effect, because the license in general
is not intended to have that effect. (If anything I would like to
somehow make copyleft-next strong copyleft while making clearer that
this does not somehow equate to mass 'infection' of a larger
proprietary code base, though I am not quite sure how to do this.) It
is, however, intended to increase the (perceived) likelihood of
enforcement.
I am not sure if you know the FSF party line since approximately the
late 1990s, which has been quite influential (among developer
communities at least). It is that the GPL, at least, is "not a
contract" (this was even the initial title of GPLv3's acceptance
clause!) and only supports enforcement through an action for copyright
infringement by the copyright holder. Without getting into the
substance of this I note that it has shaped the nature of GPL
compliance activity at least on the community-upstream-licensor and
consumer-type end-user sides.
So I'm not sure that adding the actual mechanism for it to work
will
make the license less palatable to a commercial user of copylefted
code than the GPL already is, although it should.
Right, that makes sense. In a certain sense, you could say that adding
this mechanism to, say, GPLv2 would have made anti-GPL paranoia among
some corporate lawyers (or, say, opportunistic encouragement of
paranoia on the part of *some* outside counsel during the frothier
years of the 2000s) slightly more rational.
- RF
6:48 p.m.
On Mon, Feb 18, 2013 at 2:21 PM, Richard Fontana
<fontana(a)sharpeleven.org> wrote:
> On Mon, Feb 18, 2013 at 01:21:16PM -0500, Pamela Chestek wrote:
As a related aside, whether commercial vendor (or user, especially
in
the case of AGPL-like licenses) sentiment towards FLOSS licenses has
any material impact in their uptake is an unanswered question. A key
assumption made by the FSF in the way it handled its quasi-public
drafting process for GPLv3 (2006-2007) was that in some way this was
so (even though the official view of the FSF remains that 'popularity'
of their licenses is a non-goal). I see no post-2007 evidence to
justify this view one way or another. I could elaborate on that though
it might get a bit off-topic.
An interesting question and I will continue to pursue it, as off-topic
as it might be. I always thought of Red Hat an accidental advocate of
GPL. That is, it built a business around this hot new thing Linux, and
Linux happened to be GPL, so it had no choice but to cope with it. My
decidedly anecdotal observation is that it's generally really hard for
a company (and people, for that matter) to truly cede control if they
can find any way to avoid it. Who are the successful companies who use
a true open source, copyleft model (not open core, not dual-licensed)
and also own most of the code? So your observation sounds right to me.
> That said, I suspect that anyone
> who isn't sophisticated about copyleft (which is most everyone)
> believes that this is already a feature of the GPL licenses. In other
> words, the great fear of GPL is that it will "infect" your proprietary
> code and you will be forced to release the source code. (Correct me
> if I've misunderstood; I think this provision will have that effect.)
It is not intended to have that effect, because the license in general
is not intended to have that effect. (If anything I would like to
somehow make copyleft-next strong copyleft while making clearer that
this does not somehow equate to mass 'infection' of a larger
proprietary code base, though I am not quite sure how to do this.) It
is, however, intended to increase the (perceived) likelihood of
enforcement.
I misspoke a bit - the perception is about what the penalty is if
there has been an inadvertent "infection," not that the license itself
causes the infection per se. But from my experience in the proprietary
world (which was Progress Software, which may be part of the reason
too), it's the accidental infection they're worried about, not the
deliberate creation of a derivative work. And the reason it is
accidental is because it's so bloody unclear when you've created a
derivative work.
So I guess what this means is, if you want to avoid the "infection"
concern, but want strong copyleft, then you need to be very crisp and
perhaps less expansive about what a "Derived Work" is. But then we're
back to trying to appease an audience that probably isn't going to
adopt copyleft in the first place anyway.
Pam
9:34 p.m.
On Mon, Feb 18, 2013 at 07:48:16PM -0500, Pamela Chestek wrote:
I misspoke a bit - the perception is about what the penalty is if
there has been an inadvertent "infection," not that the license itself
causes the infection per se. But from my experience in the proprietary
world (which was Progress Software, which may be part of the reason
too), it's the accidental infection they're worried about, not the
deliberate creation of a derivative work. And the reason it is
accidental is because it's so bloody unclear when you've created a
derivative work.
Sure, that's consistent with some of the GPL-phobia I've encountered
in some companies.
So I guess what this means is, if you want to avoid the
"infection"
concern, but want strong copyleft, then you need to be very crisp and
perhaps less expansive about what a "Derived Work" is.
That seems related to some of the comments Ted Ts'o has sometimes made
on this list.
I have to admit I've thought about this -- not so much the 'less
expansive' part but the 'very crisp' part. I don't know how to do that
(yet), beyond what's already been done. Or without adding a *very*
long paragraph.
But then we're
back to trying to appease an audience that probably isn't going to
adopt copyleft in the first place anyway.
Yes, especially today, where 'adopt' means 'for our new stuff that we
might open source'. Though even by early 2008 this should have been
clear, peculiar institution dual-licensors aside. If copyleft-next is
used at all (by someone other than the great Kuno Woudt [may he be
praised!]) it won't be corporations that lead the way. Nor indeed
would I want that to be so.
- RF
5:18 p.m.
On 2/19/13, Richard Fontana <fontana(a)sharpeleven.org> wrote:
> So I guess what this means is, if you want to avoid the
"infection"
> concern, but want strong copyleft, then you need to be very crisp and
> perhaps less expansive about what a "Derived Work" is.
That seems related to some of the comments Ted Ts'o has sometimes made
on this list.
I have to admit I've thought about this -- not so much the 'less
expansive' part but the 'very crisp' part. I don't know how to do that
(yet), beyond what's already been done. Or without adding a *very*
long paragraph.
Perhaps this is an item for the FAQ or documentation. If it becomes explicit
and understandable enough as FAQ item, its introduction in the license may
be easier (unless it will be useless, but proving that for sure is also a small
gain?).
8:41 p.m.
On Wed, Feb 20, 2013 at 6:18 PM, Engel Nyst <engel.nyst(a)gmail.com> wrote:
On 2/19/13, Richard Fontana <fontana(a)sharpeleven.org> wrote:
>> So I guess what this means is, if you want to avoid the "infection"
>> concern, but want strong copyleft, then you need to be very crisp and
>> perhaps less expansive about what a "Derived Work" is.
>
> That seems related to some of the comments Ted Ts'o has sometimes made
> on this list.
>
> I have to admit I've thought about this -- not so much the 'less
> expansive' part but the 'very crisp' part. I don't know how to do
that
> (yet), beyond what's already been done. Or without adding a *very*
> long paragraph.
Perhaps this is an item for the FAQ or documentation. If it becomes
explicit
and understandable enough as FAQ item, its introduction in the license may
be easier (unless it will be useless, but proving that for sure is also a
small
gain?).
I like the idea of explaining it in an FAQ. While FAQs wouldn't be binding
in
an express contract way, but they would show what the practice in the
industry is, which is probably good enough.
Pam
10:38 p.m.
On Wed, Feb 20, 2013 at 09:41:49PM -0500, Pamela Chestek wrote:
On Wed, Feb 20, 2013 at 6:18 PM, Engel Nyst <engel.nyst(a)gmail.com> wrote:
On 2/19/13, Richard Fontana <fontana(a)sharpeleven.org> wrote:
>> So I guess what this means is, if you want to avoid the
"infection"
>> concern, but want strong copyleft, then you need to be very crisp and
>> perhaps less expansive about what a "Derived Work" is.
>
> That seems related to some of the comments Ted Ts'o has sometimes made
> on this list.
>
> I have to admit I've thought about this -- not so much the 'less
> expansive' part but the 'very crisp' part. I don't know how to
do that
> (yet), beyond what's already been done. Or without adding a *very*
> long paragraph.
Perhaps this is an item for the FAQ or documentation.
That's basically what I'd been assuming all along, I suppose. The FSF
did the same thing with the GPL, but I'm envisioning something better. :)
I like the idea of explaining it in an FAQ. While FAQs wouldn't
be binding in
an express contract way, but they would show what the practice in the industry
is, which is probably good enough.
With the GPL, for better or worse, the FSF's FAQ has been remarkably
influential in affecting industry assumptions about GPL interpretation.
- RF
8:26 a.m.
On Wed, Feb 20, 2013 at 11:38 PM, Richard Fontana
<fontana(a)sharpeleven.org>wrote:
On Wed, Feb 20, 2013 at 09:41:49PM -0500, Pamela Chestek wrote:
> I like the idea of explaining it in an FAQ. While FAQs wouldn't be
binding in
> an express contract way, but they would show what the practice in the
industry
> is, which is probably good enough.
With the GPL, for better or worse, the FSF's FAQ has been remarkably
influential in affecting industry assumptions about GPL interpretation.
When I was on the proprietary side of the house I just assumed (without
agreeing) that the FAQs were the proper interpretation. It was like my
neighbor who thought the property line was in a different place than what
we thought; if I just observed his boundary line he wouldn't come out on
his porch when he was drunk and yell at me for crossing it, and I would
tolerate him coming around the fence and mowing the skinny little strip of
grass between the fence and our driveway. And I suppose ultimately someday
he would adversely possess that little plot of land, which is the point of
the FAQs....
Pam
2:50 p.m.
On Thu, Feb 21, 2013 at 09:26:01AM -0500, Pamela Chestek wrote:
When I was on the proprietary side of the house I just assumed
(without
agreeing) that the FAQs were the proper interpretation. It was like my neighbor
who thought the property line was in a different place than what we thought; if
I just observed his boundary line he wouldn't come out on his porch when he was
drunk and yell at me for crossing it, and I would tolerate him coming around
the fence and mowing the skinny little strip of grass between the fence and our
driveway. And I suppose ultimately someday he would adversely possess that
little plot of land, which is the point of the FAQs....
The complication for the GPL is that by some point that I assume took
place before 1995, FSF-copyrighted software came to represent a
minority of all the GPL-licensed software in the universe by any
reasonable metric, and the FSF's copyright share of the GPL pie has
only decreased further over time. (The FSF regarded this as a sign of
great success, it should be noted.)
Assume for sake of argument that a FAQ comes into existence. What do
you think of the license explicitly referencing it in a way that
promotes its authoritativeness?
- RF
4:48 p.m.
On Sat, Feb 23, 2013 at 3:50 PM, Richard Fontana <fontana(a)sharpeleven.org>wrote:
On Thu, Feb 21, 2013 at 09:26:01AM -0500, Pamela Chestek wrote:
> When I was on the proprietary side of the house I just assumed (without
> agreeing) that the FAQs were the proper interpretation. It was like my
neighbor
> who thought the property line was in a different place than what we
thought; if
> I just observed his boundary line he wouldn't come out on his porch when
he was
> drunk and yell at me for crossing it, and I would tolerate him coming
around
> the fence and mowing the skinny little strip of grass between the fence
and our
> driveway. And I suppose ultimately someday he would adversely possess
that
> little plot of land, which is the point of the FAQs....
The complication for the GPL is that by some point that I assume took
place before 1995, FSF-copyrighted software came to represent a
minority of all the GPL-licensed software in the universe by any
reasonable metric, and the FSF's copyright share of the GPL pie has
only decreased further over time. (The FSF regarded this as a sign of
great success, it should be noted.)
I recognized this, but it also seemed to me that the FSF interpretation was
the most restrictive, so if I abided by its interpretation I was probably
on safe ground.
Assume for sake of argument that a FAQ comes into existence. What do
you think of the license explicitly referencing it in a way that
promotes its authoritativeness?
If you're trying to make the FAQs authoritative, then I consider it
somewhat underhanded. Although it's commonplace, I don't like incorporation
by reference of documents that can be changed and I believe that courts are
interpreting them as what existed at the time you manifested assent to the
primary contract (which means there is probably a proof problem). If it's
that important, put it in the original document. And I consider it
cheating if your goal is low word count. ;)
I think of the collection of materials in the same way I think of statutory
interpretation. I was taught that the words control, but if there is
ambiguity one looks to the legislative history, then the policy, as
additional tools for interpretation, particularly about intent. Contract
interpretation is similar; if the words are clear you stay in the four
corners, otherwise you might look at extrinsic evidence including custom
and practice in the industry. So I think of the role of the FAQs as the
legislative history that is showing the intent of the drafters, or a
description of what the custom and practice in the industry is (with
probably no small amount of actually creating that custom and practice in
the first place, too). All of this should work in your favor; if there is
ambiguity it can mean that a loophole you didn't see in the drafting is
closed based on your intent, not a crabbed or overly literal reading.
For this case, then, I think it would be useful to point out that the FAQs
are meant to be relied upon for this interpretative function, something
along the lines of "to interpret this document, look at the FAQs which
explain what we were trying to accomplish." This documents that they indeed
have weight that should be considered in the interpretation. But you only
get there if the language itself isn't clear and you're allowed to consider
extrinsic evidence, in which case the FAQs would probably be considered
fairly illuminating whether you had mentioned them or not. So is it worth
the word count?
Pam
5:10 p.m.
On Sat, Feb 23, 2013 at 05:48:50PM -0500, Pamela Chestek wrote:
The complication for the GPL is that by some point that I assume
took
place before 1995, FSF-copyrighted software came to represent a
minority of all the GPL-licensed software in the universe by any
reasonable metric, and the FSF's copyright share of the GPL pie has
only decreased further over time. (The FSF regarded this as a sign of
great success, it should be noted.)
I recognized this, but it also seemed to me that the FSF interpretation was the
most restrictive, so if I abided by its interpretation I was probably on safe
ground.
Ah, yes, I suppose I've said something similar to this. I've talked
about how, back when the FSF (and/or RMS) was more actively involved
in GPL interpretation, the FSF/RMS view was valuable in establishing a
reasonable-enough (some of course would disagree with that)
interpretive outer bound.
[...]
For this case, then, I think it would be useful to point out that the
FAQs are
meant to be relied upon for this interpretative function, something along the
lines of "to interpret this document, look at the FAQs which explain what we
were trying to accomplish." This documents that they indeed have weight that
should be considered in the interpretation. But you only get there if the
language itself isn't clear and you're allowed to consider extrinsic evidence,
in which case the FAQs would probably be considered fairly illuminating whether
you had mentioned them or not. So is it worth the word count?
Probably not.
- RF
10:20 p.m.
On Mon, Feb 18, 2013 at 07:48:16PM -0500, Pamela Chestek wrote:
An interesting question and I will continue to pursue it, as
off-topic
as it might be. I always thought of Red Hat an accidental advocate of
GPL. That is, it built a business around this hot new thing Linux, and
Linux happened to be GPL, so it had no choice but to cope with it. My
decidedly anecdotal observation is that it's generally really hard for
a company (and people, for that matter) to truly cede control if they
can find any way to avoid it. Who are the successful companies who use
a true open source, copyleft model (not open core, not dual-licensed)
and also own most of the code? So your observation sounds right to me.
Companies will be willing to cede control if it is in their interest
to do so. There have been a number examples of this under a
BSD/MIT/Apache-style license, where a company has taken previously
proprietary code, released under an Open Source license, and then
continued to develop it.
It's probably harder to find examples where this has been done under a
copyleft license, but companies can and have ceded control under a
more general Open Source set of licenses. The reasons why in many of
these cases companies like IBM, Google, Intel, among others, have
chosen an Apache or BSD or MIT style license instead of a copyleft
style license are certainly things we could discuss if there is
interest, but I don't think "companies are scared of ceding control"
is a sufficient explanation.
Regards,
- Ted
12:31 p.m.
Hello,
On 02/18/2013 06:36 PM, Richard Fontana wrote:
That said, on to the HBR cure: I had a conversation with Bradley
Kuhn
in which we very briefly discussed this issue, as Bradley was long
under the impression that the GPL could not *possibly* support an
argument that mere licensees have some standing to enforce the source
code requirement (despite the fact, of which Bradley might not be
aware, that over the years some legal academics have at last taken the
question seriously, if skeptically, and some private practice lawyers
[admittedly seeking to profit from the erstwhile open source boom, the
SCO crisis, etc.] purported to take this qite seriously too).
I noted the obvious point that copyleft-next and GPLv2/GPLv3 are not
worded identically, so his comment was partially irrelevant.
Bradley mentioned some interesting work that Australian lawyer Brendan
Scott had done attempting, AIUI, to use Australian consumer protection
law to give a basis for GPL licensees to sue upstream distributors for
failure to provide source code. It is not clear whether such efforts
have been successful.
Under dutch law a mere licensee should be able to enforce the GPL.
Dutch IT lawyer Arnoud Engelfriet blogged [1] about this last year.
I will give a quick English summary of the relevant bits of that blog post:
Dutch contract law has a "third-party beneficiary" concept. An example
of where this can be used is when you purchase something from an online
retailer. The retailer has a contract with a courier service to deliver
the package, as a buyer you do not have a contract with the courier
service. However you are a beneficiary of the contract between the
retailer and the courier, you can sue the courier and demand that the
courier honors the contract it has with the retailer (on penalty of a
"dwangsom" -- a conditional fine).
This should also work for copyleft licenses. If you purchase a device
with e.g. the Linux kernel in it, and the manufacturer does not give you
the source code when you ask for it, you can sue the manufacturer of the
device and demand that it honors the contract (GPLv2) the manufacturer
has with the Linux kernel developers.
-- kuno / warp.
[1]
http://blog.iusmentis.com/2012/03/16/kun-je-als-koper-van-een-tomtec-tabl...
6:06 p.m.
On 02/18/2013 12:36 PM, Richard Fontana wrote:
This is a HBR cure.
First, for those of you not following the official draft closely, as
of a recent commit, I have audaciously and iconoclastically added this
sentence to the (now single) section on object code disribution:
Each recipient of the Covered Work from You is an intended
third-party beneficiary of this License solely as to this section []
with the right to enforce its terms.
If any of the lawyers on this list have any comments of a strictly
insider-professional-knowledge *legal* nature on this sentence,
including arguments for its deletion or modification, feel free to
speak up (in a HBR-compliant manner of course).
If anyone (non-lawyer or otherwise) thinks it would be bad *policy* to
live in a world where a GPL-like license could be enforced through the
legal system by the obvious immediate beneficiaries of the source code
requirement, feel free to speak up (in a HBR-compliant manner of
course).
That said, on to the HBR cure: I had a conversation with Bradley Kuhn
in which we very briefly discussed this issue, as Bradley was long
under the impression that the GPL could not *possibly* support an
argument that mere licensees have some standing to enforce the source
code requirement (despite the fact, of which Bradley might not be
aware, that over the years some legal academics have at last taken the
question seriously, if skeptically, and some private practice lawyers
[admittedly seeking to profit from the erstwhile open source boom, the
SCO crisis, etc.] purported to take this qite seriously too).
I noted the obvious point that copyleft-next and GPLv2/GPLv3 are not
worded identically, so his comment was partially irrelevant.
Bradley mentioned some interesting work that Australian lawyer Brendan
Scott had done attempting, AIUI, to use Australian consumer protection
law to give a basis for GPL licensees to sue upstream distributors for
failure to provide source code. It is not clear whether such efforts
have been successful.
- RF
_______________________________________________
copyleft-next mailing list
copyleft-next(a)lists.fedorahosted.org
https://lists.fedorahosted.org/mailman/listinfo/copyleft-next
I am not a lawyer. I will say that developing a license and
promulgating it becomes an academic exercise if the ways to enforce it
are minimal to non-existent. There needs to be teeth to back up
something to make it work regardless of how strange our current
copyright legislation may be.
Stephen Michael Kellat
4087
days inactive
4092
days old
copyleft-next@lists.fedorahosted.org
14 comments
6 participants
participants (6)
-
Engel Nyst -
Kuno Woudt -
Pamela Chestek -
Richard Fontana -
Stephen Michael Kellat -
Theodore Ts'o