+1
>
> -DeWitt
>
>
>
>
> >
>
I'd be perfectly happy if we achieved my goal with some other model
than one based on ASF, but doing so seems unlikely (and undesirable)
right now.
-Gabe
On Thu, Sep 25, 2008 at 9:28 AM, DeWitt Clinton <dew...@google.com> wrote:
--
Gabe Wachob / gwa...@wachob.com \ http://blog.wachob.com
That Apache exists and has been rather successful at what it does
means that we can devote our energy to the hard, new tasks -- which
are primarily social and socializing our work and refining our purpose
and what we offer -- whereas developing or innovating in the
organizational structure might otherwise be a misapplication of our
collective (and very limited) energies.
Chris
--
Might it be worthwhile to memorialize the specific regards in which
this effort is intended to be different from the Apache model? For
the benefit of newcomers and also, as noted, for future reference?
I'm thinking it might not be obvious to everyone who reads this why
the OWF effort isn't duplicative. (And no, I'm not implying that it
is, I'm just offering a suggestion.)
I think you're mischaracterizing my statement a bit. What I tried to
say was that the reuse of the ASF framework is a means to an end for
me, not an end.
And for what its worth, I had the same initial feeling - "We can't
possibly use ASF - specs are so different from code".
But if you think about the IPR policy itself as a piece of code, then
actually the same community effort around such output is very similar
to what comes out of an open source software effort at ASF. Here at
OWF, we're just one level of meta away from code. To wit, we're
producing the code (the IPR policy, process) for systems (spec
contributors) to interoperate (e.g. not sue each other).
Its a bit of a stretch, but in this way, the cultural fit of the
ASF-style structure actually seems appropriate.
-Gabe
This is not something we actually have agreement on and has proved
> Rather, the OWF is designed to create an
> IP policy that explicitly protects the right to fork and create derivative
> works of specifications -- even to the extent that those derivative works
> may deviate from interoperability.
itself to be the most contested item.
My motivation for this work hasn't been the lack of existing licenses.
There are plenty of licenses that with minor modifications would suit
my needs and will make me happy to create specifications. In the trade-
off between a wider participation to unlimited ability to create
derivative work, I vote for participation.
I think the easiest way to express this (another generalization) is that
copyright is about the tangible expression and patents are about the
application -- a mixture of what an implementation does, how it does it, and
where (on what behalf) it does it.
Continuing this over-simplification, I suggest that is why patent claims
usually do not apply to specifications but their implementations and what is
done with those implementations. Notice that the existing non-assertion
statements and covenants that have been cited on this list are all worded in
terms of implementation of such-and-such specification.
- Dennis
PS: Important subtlety: It is of course, valuable to make sure that there is
no patent claim that can't be escaped in making any practical implementation
whatsoever, if one wants to ensure that royalty- and license-free
implementations are possible. Unfortunately, an IPR agreement by
participants/contributors to a specification id not by itself sufficient to
avoid that situation, if the necessary claim subsists in a patent held by a
non-participant.
PPS: I don't even play a lawyer on television, only in my head.
Dennis E. Hamilton
------------------
NuovoDoc: Design for Document System Interoperability
mailto:Dennis....@acm.org | gsm:+1-206.779.9430
http://NuovoDoc.com http://ODMA.info/dev/ http://nfoWorks.org
-----Original Message-----
From: DeWitt Clinton
http://groups.google.com/group/open-web-discuss/msg/c680e358c4325b21?hl=en
Sent: Friday, September 26, 2008 10:55
To: open-web...@googlegroups.com
Subject: Re: Goals of the Open Web Foundation (strawman)
[ ... ]
Rather, the OWF is designed to create an IP policy that explicitly protects
the right to fork and create derivative works of specifications -- even to
the extent that those derivative works may deviate from interoperability.
[ ... ]
-----Another Message-----
From: Eric Hammer-Lahav
http://groups.google.com/group/open-web-discuss/msg/a80c0b46295ceb0b?hl=en
[ ... ]
The technical aspects of forking or creating derivative work from source
code is completely different from specifications.
[ ... ]
So, I guess it depends what you mean by a "right to fork". Setting
ourselves up so that you can't reuse language from a spec would be
bad. But I agree with Eran that you can't expect IPR holders to grant
a licence to implementations of a spec and all possible derivatives,
since that would include all possible specs. It seems to me quite
reasonable that if you want to get IPR assigned to your new spec, then
you have to ask (or be prepared to show you were already covered
somehow).
-----Original Message-----
From: Ben Laurie
http://groups.google.com/group/open-web-discuss/msg/dc79304bb89101d2?hl=en
Sent: Friday, September 26, 2008 12:05
To: open-web...@googlegroups.com
Subject: Re: Goals of the Open Web Foundation (strawman)
[ ... ]
So, I guess it depends what you mean by a "right to fork". Setting
I have no idea if this can be achieved in legalese, but I think this
sounds like a perfectly reasonable aspiration. Probably you would have
to call out what A, B and C were (i.e. define their boundaries) in the
original spec, or, as I suggested earlier, argue after the fact that A
and B are clearly covered by the original licence.
Simply assuming that because A is used in the original spec, A can
therefore be used in _any_ spec gets you back to the original problem:
if you require a blanket licence, you're going to hurt participation.
--‾--‾---------‾--‾----‾------------‾-------‾--‾----‾
You received this message because you are subscribed to the Google Groups "Open Web Foundation Discussion" group.
To post to this group, send email to open-web...@googlegroups.com
To unsubscribe from this group, send email to open-web-discu...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/open-web-discuss?hl=en
-‾----------‾----‾----‾----‾------‾----‾------‾--‾---
I think this is a good simplification. However, it misses the derivative situation I think David was touching on (and which I am most concerned about).
What happens when derivative specification β breaks functionality of specification α? That is, you have forked to an incompatible specification?
As long as β's use of α is entirely compliant, then β can be defined as α + ε, and the re-used parts of α are covered under the IPR regime of α and the ε components need new IPR. But if β breaks α, it can no longer be expressed in terms of α + ε, and therefore breaks the interoperability of α in a way that should, IMO, require re-licensing from any IP holders who agreed to the IPR for α.
The quid-pro-quo for permissive licensing of IP is that the open specification enables interoperability. If IP holders simply wanted the pure benefits of open source licensing (of which there are many), they could have already adopted any one of many open source licensing terms. If OWF actually wants to bring commercial vendors in who have significant IP, it seems to me we should encourage fair value (interoperability) in exchange for IP.
I get the benefits of N.E.A.-style open systems, but I worry that the commitment to permissiveness by some of the folks here comes at the expense of interoperability between implementations. As one who sees dataportability and service portability as the critical next step in the open web, I argue that this exuberance for permissive, incompatible derivatives is misplaced.
There's a challenging dance between open innovation and interoperability. If OWF ends up adopting the extremes of either, I think we will have failed. However, if we can find an effective way to enable NEA systems that ensure both interoperability and a forward path for innovation... that would be work well worth all of our effort.
-j
--
Joe Andrieu
SwitchBook
http://www.switchbook.com
j...@switchbook.com
+1 (805) 705-8651
From:
open-web...@googlegroups.com [mailto:open-web...@googlegroups.com] On
Behalf Of DeWitt Clinton
Sent: Friday, September 26, 2008 12:27 PM
To: open-web...@googlegroups.com
Subject: Re: Goals of the Open Web Foundation (strawman)
I've always had this crazy idea that we could create a license that grants a) the right to create an implementation of a specification, b) the right to create derivative works of that specification, and c) the right to create an implementation of portions of the specification that appear in derivative works.