[Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[David Woodhouse]

I'm slightly loath to bring this up but I think it does need doing...

There are various people involved with GPL enforcement actions for the
kernel, and there seems to be a lot of misinformation about it all. 

Some parties are portrayed as acting excessively and inappropriately —
and with dubious motives — to enforce the GPL; others as being rabidly
*against* enforcement and effectively wanting a BSD-licensed free-for
-all because that supports their corporate interests.

None of these portrayals are entirely accurate, I'm sure, and the
histrionics don't really do anyone any good.

I'm sure we won't reach a universal consensus, but it would be useful
to get the interested parties together and have a coherent discussion
about it, so that everyone can have a proper understanding of the fact,
and also a reasonable idea of the 'feeling in the room' regarding if,
when and how we require compliance with the GPL.

-- 
dwmw2

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[David Howells]

Related to this is the issue of creeping GPL-only-isation of important
exported symbols.  This affects not only closed-source stuff, but also non-GPL
licensed open-source projects such as OpenAFS that can't relicense their code
under the GPL.

David

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[David Woodhouse]

On Sat, 2015-07-04 at 11:18 +0100, David Howells wrote:
> Related to this is the issue of creeping GPL-only-isation of important
> exported symbols. 

Perhaps, yes.

> This affects not only closed-source stuff, but also non-GPL
> licensed open-source projects such as OpenAFS that can't relicense 
> their code under the GPL.

Haven't you finished obsoleting the kernel parts of OpenAFS yet? :)

-- 
dwmw2

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Steven Rostedt]

On Sat, 04 Jul 2015 09:06:01 +0100
David Woodhouse <dwmw2@infradead.org> wrote:

> I'm slightly loath to bring this up but I think it does need doing...
> 
> There are various people involved with GPL enforcement actions for the
> kernel, and there seems to be a lot of misinformation about it all. 
> 
> Some parties are portrayed as acting excessively and inappropriately —
> and with dubious motives — to enforce the GPL; others as being rabidly
> *against* enforcement and effectively wanting a BSD-licensed free-for
> -all because that supports their corporate interests.
> 
> None of these portrayals are entirely accurate, I'm sure, and the
> histrionics don't really do anyone any good.
> 
> I'm sure we won't reach a universal consensus, but it would be useful
> to get the interested parties together and have a coherent discussion
> about it, so that everyone can have a proper understanding of the fact,
> and also a reasonable idea of the 'feeling in the room' regarding if,
> when and how we require compliance with the GPL.
> 

Would having a Lawyer be present also be required. That way we don't
have everyone saying BS about what the law actually implies. Having a
lawyer there as just someone to keep things real would be nice.

It also gets even more complex with what exactly the GPL can cover with
various laws of various countries.

-- Steve

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[David Woodhouse]

On Tue, 2015-07-07 at 14:34 -0400, Steven Rostedt wrote:
> Would having a Lawyer be present also be required. That way we don't
> have everyone saying BS about what the law actually implies. Having a
> lawyer there as just someone to keep things real would be nice.

Lawyers will present whatever opinion they're paid to present. There is
no "right" answer until/unless it's seen in court. And then the ruling
only applies to specific circumstances. And is only binding in certain
jurisdictions, as you suggested. And until it's appealed.

I don't think a discussion about 'the true meaning of the GPL', or its
interpretation under specific legal systems, would be particularly
productive.

I think it would be good, though, to have a better understanding of
what people *want* the GPL to mean and what they *think* it means —
perhaps better phrased as "what they would pay a lawyer to argue".

The point is that we aren't attempting to reach a simple resolution
with a 'right answer'. There are parties with different desires — from
demanding strict compliance with the *maximum* they can argue for the
GPL to mean, to basically wanting to act as if it's a BSD licence, in
order to avoid scaring commercial users away and feeding FUD stories
like this one¹.

It would be useful to have an idea of where the average core developer
falls within that spectrum — that was the first specific thing I was
hoping would come from the proposed session.

The other specific goal (and perhaps the more important one) was to
have a coherent report about the enforcement actions and behind-the
-scenes negotiations w.r.t compliance that there is so much
misinformation and politicking about.

To that end, we should probably invite Bradley Kuhn or Karen Sandler
from SF Conservancy to talk about their efforts. And someone from the
LF TAB will presumably also be there to discuss the compliance
viewpoint from the LF side.

-- 
dwmw2
¹ http://sdtimes.com/from-the-editors-when-did-open-source-software-get-so-scary/

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Greg KH]

On Tue, Jul 07, 2015 at 08:51:10PM +0100, David Woodhouse wrote:
> On Tue, 2015-07-07 at 14:34 -0400, Steven Rostedt wrote:
> > Would having a Lawyer be present also be required. That way we don't
> > have everyone saying BS about what the law actually implies. Having a
> > lawyer there as just someone to keep things real would be nice.
> 
> Lawyers will present whatever opinion they're paid to present. There is
> no "right" answer until/unless it's seen in court. And then the ruling
> only applies to specific circumstances. And is only binding in certain
> jurisdictions, as you suggested. And until it's appealed.
> 
> I don't think a discussion about 'the true meaning of the GPL', or its
> interpretation under specific legal systems, would be particularly
> productive.

I second this.

> I think it would be good, though, to have a better understanding of
> what people *want* the GPL to mean and what they *think* it means —
> perhaps better phrased as "what they would pay a lawyer to argue".
> 
> The point is that we aren't attempting to reach a simple resolution
> with a 'right answer'. There are parties with different desires — from
> demanding strict compliance with the *maximum* they can argue for the
> GPL to mean, to basically wanting to act as if it's a BSD licence, in
> order to avoid scaring commercial users away and feeding FUD stories
> like this one¹.
> 
> It would be useful to have an idea of where the average core developer
> falls within that spectrum — that was the first specific thing I was
> hoping would come from the proposed session.
> 
> The other specific goal (and perhaps the more important one) was to
> have a coherent report about the enforcement actions and behind-the
> -scenes negotiations w.r.t compliance that there is so much
> misinformation and politicking about.
> 
> To that end, we should probably invite Bradley Kuhn or Karen Sandler
> from SF Conservancy to talk about their efforts.

As those "efforts" are being driven by the developers of the kernel, I
think that the developers involved would be the best to present this,
not the SF Conservancy people.  As you point out, the SFLC is just the
lawyers being hired by the developers to do what they want them to do :)

thanks,

greg k-h

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[David Woodhouse]

On Tue, 2015-07-07 at 13:00 -0700, Greg KH wrote:
> As those "efforts" are being driven by the developers of the kernel, I
> think that the developers involved would be the best to present this,
> not the SF Conservancy people.  As you point out, the SFLC is just the
> lawyers being hired by the developers to do what they want them to do :)

A lot of those developers will be in the room too, of course.

But the "efforts" are being driven *by* Conservancy on *behalf* of
those developers who have asked Conservancy to do so. And I suspect
someone from Conservancy is better placed to give up-to-date
information about what's actually happening "on the ground".

The point is to get people in the room and have a direct discussion
without hearsay or misinformation — and doing that *without* someone
from Conservancy doesn't really make much sense.

-- 
dwmw2

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[John W. Linville]

On Tue, Jul 07, 2015 at 09:18:30PM +0100, David Woodhouse wrote:
> On Tue, 2015-07-07 at 13:00 -0700, Greg KH wrote:
> > As those "efforts" are being driven by the developers of the kernel, I
> > think that the developers involved would be the best to present this,
> > not the SF Conservancy people.  As you point out, the SFLC is just the
> > lawyers being hired by the developers to do what they want them to do :)
> 
> A lot of those developers will be in the room too, of course.
> 
> But the "efforts" are being driven *by* Conservancy on *behalf* of
> those developers who have asked Conservancy to do so. And I suspect
> someone from Conservancy is better placed to give up-to-date
> information about what's actually happening "on the ground".
> 
> The point is to get people in the room and have a direct discussion
> without hearsay or misinformation — and doing that *without* someone
> from Conservancy doesn't really make much sense.

I'm inclined to agree with David.  I'm not casting any aspersions
on Bradly or the Conservancy, but I definitely think they are heavy
influencers with regard to this topic.

John
-- 
John W. Linville		Someday the world will need a hero, and you
linville@tuxdriver.com			might be all we have.  Be ready.

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Tim Bird]

On 07/07/2015 01:18 PM, David Woodhouse wrote:
> On Tue, 2015-07-07 at 13:00 -0700, Greg KH wrote:
>> As those "efforts" are being driven by the developers of the kernel, I
>> think that the developers involved would be the best to present this,
>> not the SF Conservancy people.  As you point out, the SFLC is just the
>> lawyers being hired by the developers to do what they want them to do :)
> 
> A lot of those developers will be in the room too, of course.
> 
> But the "efforts" are being driven *by* Conservancy on *behalf* of
> those developers who have asked Conservancy to do so. And I suspect
> someone from Conservancy is better placed to give up-to-date
> information about what's actually happening "on the ground".
> 
> The point is to get people in the room and have a direct discussion
> without hearsay or misinformation — and doing that *without* someone
> from Conservancy doesn't really make much sense.

My experience with this is that you'll get one side (and perspective)
of the compliance enforcement story, only.  Companies involved in
compliance "improvement" exercises simply will not comment on them, so
it's not easy to tell if the remedies being requested are understood
by all parties or not, and whether I'd agree with them.

>From my standpoint it would be useful to hear the Conservancy's stance
on a few issues.
 -- Tim

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Greg KH]

On Wed, Jul 08, 2015 at 08:41:35AM -0700, Tim Bird wrote:
> 
> 
> On 07/07/2015 01:18 PM, David Woodhouse wrote:
> > On Tue, 2015-07-07 at 13:00 -0700, Greg KH wrote:
> >> As those "efforts" are being driven by the developers of the kernel, I
> >> think that the developers involved would be the best to present this,
> >> not the SF Conservancy people.  As you point out, the SFLC is just the
> >> lawyers being hired by the developers to do what they want them to do :)
> > 
> > A lot of those developers will be in the room too, of course.
> > 
> > But the "efforts" are being driven *by* Conservancy on *behalf* of
> > those developers who have asked Conservancy to do so. And I suspect
> > someone from Conservancy is better placed to give up-to-date
> > information about what's actually happening "on the ground".
> > 
> > The point is to get people in the room and have a direct discussion
> > without hearsay or misinformation — and doing that *without* someone
> > from Conservancy doesn't really make much sense.
> 
> My experience with this is that you'll get one side (and perspective)
> of the compliance enforcement story, only.  Companies involved in
> compliance "improvement" exercises simply will not comment on them, so
> it's not easy to tell if the remedies being requested are understood
> by all parties or not, and whether I'd agree with them.

s/will not/usually can not/g

I'm worried that since a lot of people can't talk about this, for
various reasons, that the usefulness of it might not be all that
helpful.

> From my standpoint it would be useful to hear the Conservancy's stance
> on a few issues.

What specific issues are you curious about that you think would be
good to discuss / hear about?

thanks,

greg k-h

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Tim Bird]

On 07/08/2015 02:25 PM, Greg KH wrote:
> On Wed, Jul 08, 2015 at 08:41:35AM -0700, Tim Bird wrote:
>> On 07/07/2015 01:18 PM, David Woodhouse wrote:
>>> On Tue, 2015-07-07 at 13:00 -0700, Greg KH wrote:
>>>> As those "efforts" are being driven by the developers of the kernel, I
>>>> think that the developers involved would be the best to present this,
>>>> not the SF Conservancy people.  As you point out, the SFLC is just the
>>>> lawyers being hired by the developers to do what they want them to do :)
>>>
>>> A lot of those developers will be in the room too, of course.
>>>
>>> But the "efforts" are being driven *by* Conservancy on *behalf* of
>>> those developers who have asked Conservancy to do so. And I suspect
>>> someone from Conservancy is better placed to give up-to-date
>>> information about what's actually happening "on the ground".
>>>
>>> The point is to get people in the room and have a direct discussion
>>> without hearsay or misinformation — and doing that *without* someone
>>> from Conservancy doesn't really make much sense.
>>
>> My experience with this is that you'll get one side (and perspective)
>> of the compliance enforcement story, only.  Companies involved in
>> compliance "improvement" exercises simply will not comment on them, so
>> it's not easy to tell if the remedies being requested are understood
>> by all parties or not, and whether I'd agree with them.
> 
> s/will not/usually can not/g
Agreed.

> 
> I'm worried that since a lot of people can't talk about this, for
> various reasons, that the usefulness of it might not be all that
> helpful.
> 
>> From my standpoint it would be useful to hear the Conservancy's stance
>> on a few issues.
> 
> What specific issues are you curious about that you think would be
> good to discuss / hear about?

I'm interested in the scope of remedies requested by the Conservancy.
Specifically, if the Conservancy finds a GPL violation for one piece of
software, or for one division or group in a company, whether they
then request to review and approve the compliance for other unrelated
pieces of software, or for software from other divisions, groups or
products.  That was the thing that caused the most concern within
companies I talked to a few years ago, when I got caught in a
firestorm for proposing support by the CE Workgroup for the Toybox
project.

It's unclear to me whether this is the type of subject David is
proposing, or whether there's much benefit to this discussion at
a kernel summit.  The benefit to me would be to gauge other developers'
reactions to these issues (though honestly I consider myself a 
very low priority for invitation to the summit.)
 -- Tim

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Chris Mason]

On Tue, Jul 07, 2015 at 08:51:10PM +0100, David Woodhouse wrote:
> On Tue, 2015-07-07 at 14:34 -0400, Steven Rostedt wrote:
> > Would having a Lawyer be present also be required. That way we don't
> > have everyone saying BS about what the law actually implies. Having a
> > lawyer there as just someone to keep things real would be nice.
> 
> Lawyers will present whatever opinion they're paid to present. There is
> no "right" answer until/unless it's seen in court. And then the ruling
> only applies to specific circumstances. And is only binding in certain
> jurisdictions, as you suggested. And until it's appealed.
> 
> I don't think a discussion about 'the true meaning of the GPL', or its
> interpretation under specific legal systems, would be particularly
> productive.
> 
> I think it would be good, though, to have a better understanding of
> what people *want* the GPL to mean and what they *think* it means —
> perhaps better phrased as "what they would pay a lawyer to argue".
> 
> The point is that we aren't attempting to reach a simple resolution
> with a 'right answer'. There are parties with different desires — from
> demanding strict compliance with the *maximum* they can argue for the
> GPL to mean, to basically wanting to act as if it's a BSD licence, in
> order to avoid scaring commercial users away and feeding FUD stories
> like this one¹.
> 
> It would be useful to have an idea of where the average core developer
> falls within that spectrum — that was the first specific thing I was
> hoping would come from the proposed session.

Maybe a survey would be better than open discussion?

I'll bet the average core developer wants to hack on the kernel, and
share in the work that everyone else is doing.  But I'll also bet the
conversation would be dominated by extremes, spiral down into
my-company-won't-let-me-say-xxyyzz, and then finally settle into
comparisons of US law with everyone else.

Eventually the real crazy talk will begin and Jens will have to admit
that he *still* hasn't open sourced his online shopping site for kernel
developer inspired action figures.

> 
> The other specific goal (and perhaps the more important one) was to
> have a coherent report about the enforcement actions and behind-the
> -scenes negotiations w.r.t compliance that there is so much
> misinformation and politicking about.
> 
> To that end, we should probably invite Bradley Kuhn or Karen Sandler
> from SF Conservancy to talk about their efforts. And someone from the
> LF TAB will presumably also be there to discuss the compliance
> viewpoint from the LF side.

I'm not sure how much can really be disclosed, but I'd much prefer to
see fact based sessions about current enforcement actions than open
discussions about the GPL.

-chris

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Luck, Tony]

> I'll bet the average core developer wants to hack on the kernel, and
> share in the work that everyone else is doing.  But I'll also bet the
> conversation would be dominated by extremes, spiral down into
> my-company-won't-let-me-say-xxyyzz, and then finally settle into
> comparisons of US law with everyone else.

A sizable fraction of commits come from people whose day job is
to work on Linux. All my contributions to the Linux kernel are owned
by my employer.  Do companies with large aggregated contributions
get a say in the GPL enforcement direction?

I'm certainly not speaking for my employer here - but you might think
that a company that makes a huge percentage of its profits from selling
silicon chips would want the minimum possible barriers to people buying
those chips to run s/w. That may, or may not, align with the opinions of
people working for the company.

-Tony

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Chris Mason]

On Tue, Jul 07, 2015 at 10:15:24PM +0000, Luck, Tony wrote:
> > I'll bet the average core developer wants to hack on the kernel, and
> > share in the work that everyone else is doing.  But I'll also bet the
> > conversation would be dominated by extremes, spiral down into
> > my-company-won't-let-me-say-xxyyzz, and then finally settle into
> > comparisons of US law with everyone else.
> 
> A sizable fraction of commits come from people whose day job is
> to work on Linux. All my contributions to the Linux kernel are owned
> by my employer.  Do companies with large aggregated contributions
> get a say in the GPL enforcement direction?
> 

I think they've had a say all along, and our stability as a community
has benefited from it.

> I'm certainly not speaking for my employer here - but you might think
> that a company that makes a huge percentage of its profits from selling
> silicon chips would want the minimum possible barriers to people buying
> those chips to run s/w. That may, or may not, align with the opinions of
> people working for the company.

But what level of forced sharing actually makes the lowest barrier over
time?  I'm happy to see us wander around with our wildly different
opinions about this.  It makes for great debates, but it's not something
we can answer at KS.

But going back to David's original topic, there are some important
current events here.  It's not a horrible idea to talk about them,
although Korea is pretty far to drag a lawyer for a single session.

-chris

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Guenter Roeck]

On 07/07/2015 12:51 PM, David Woodhouse wrote:
>
> The point is that we aren't attempting to reach a simple resolution
> with a 'right answer'. There are parties with different desires — from
> demanding strict compliance with the *maximum* they can argue for the
> GPL to mean, to basically wanting to act as if it's a BSD licence, in
> order to avoid scaring commercial users away and feeding FUD stories
> like this one¹.
>
I think that article is just a sign of the current Zeitgeist.
Everything has to be scary nowadays, or it doesn't count.

> It would be useful to have an idea of where the average core developer
> falls within that spectrum — that was the first specific thing I was
> hoping would come from the proposed session.
>
> The other specific goal (and perhaps the more important one) was to
> have a coherent report about the enforcement actions and behind-the
> -scenes negotiations w.r.t compliance that there is so much
> misinformation and politicking about.
>
> To that end, we should probably invite Bradley Kuhn or Karen Sandler
> from SF Conservancy to talk about their efforts. And someone from the
> LF TAB will presumably also be there to discuss the compliance
> viewpoint from the LF side.
>
... and then have the LF attorneys brief LF member companies about the
outcome of the meeting. I always find those legal briefings very useful.

Guenter

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[Christoph Lameter]

On Tue, 7 Jul 2015, Guenter Roeck wrote:

> ... and then have the LF attorneys brief LF member companies about the
> outcome of the meeting. I always find those legal briefings very useful.

One subject that I would like to have some clarification on is relicensing
of in-kernel-code. There were a couple of license changes recently from
GPL to "pick onee GPL or BSD" in the infiniband stack. I was a bit
surprised by that being possible.

Presumably the copyright holders have to agree to this. What exactly
qualifies to be a copyright holder? If one is mentioned in the file
header? If one modified a line of code in the file?

Re: [Ksummit-discuss] [CORE TOPIC] GPL enforcement actions

[James Bottomley]

On Fri, 2015-07-31 at 13:45 -0500, Christoph Lameter wrote:
> On Tue, 7 Jul 2015, Guenter Roeck wrote:
> 
> > ... and then have the LF attorneys brief LF member companies about the
> > outcome of the meeting. I always find those legal briefings very useful.
> 
> One subject that I would like to have some clarification on is relicensing
> of in-kernel-code. There were a couple of license changes recently from
> GPL to "pick onee GPL or BSD" in the infiniband stack. I was a bit
> surprised by that being possible.
> 
> Presumably the copyright holders have to agree to this. What exactly
> qualifies to be a copyright holder? If one is mentioned in the file
> header? If one modified a line of code in the file?

It's the copyright law definition: for a work (or a modification of a
work) to be eligible for copyright protection it (or the modification)
must rise to the level of being original.  There's a lot of case law
over what this means for a book, but very little over what this means
for a computer program.  In one of the book decisions, the US Supreme
Court did say that originality requires "a modicum of creativity" (that
was a case about copyrighting the phone book) which likely applies to
programs as well.  There's also case law that implies things that are
common patterns in programming (like correcting a use after free) may
not be original enough but beyond that, it's anyone's guess.

James