Re: [eigen] legal question

[Benoit Jacob <jacob.benoit.1@xxxxxxxxx>]

Here's a answer from Luis Villa of Mozilla, that is relevant to this discussion:

http://groups.google.com/group/mozilla.governance.mpl-update/msg/12e90f946369ad71

Basically:
A. we can indeed decide that very short code snippets are not
copyrighted, in which case there's no problem at all (and I would then
recommend CC-BY i.e. no copyleft on the docs).
B. what Mozilla is doing is that both code AND documentation are
MPL-licensed. The MPL alllows that i.e. it is a license that applies
well to both code and docs.

Unfortunately, the docs licensing problem is something we need to fix
NOW, we can't wait till the MPL v2 is out and everybody agrees to use
it (which might never happen).

So I think that our best option for now is A. We do have very short
code snippets at many places in the docs, e.g. the typedefs, the
template declarations etc. And in a template metaprogramming context
like ours, the border between this and "actual code" is just too
fuzzy.

Is everyone OK to license our documentation as CC-BY, including the
snippets it contains? That means that we need to refrain from
including significant pieces of Eigen code into the docs (snippets
should be trivial), which could have been useful in more advanced,
developer documentation. To work around this problem, we should
instead link to some online copy of those Eigen source files with
anchor to the relevant line number.

ok?

Benoit

2010/7/25 Benoit Jacob <jacob.benoit.1@xxxxxxxxx>:
> On a related note, I just stumbled across this LWN article which
> should now be in open access:
>
> http://lwn.net/Articles/394219/
>
> It is about GCC developers struggling to be able to include snippets
> from their GPLv3-licensed code into their GFDL-licensed documentation.
>
> It does note, as I just did, that "the real problem is that we as a
> community lack a copyleft license that works well for both code and
> text."
>
> I will make sure to bring this to the attention of Mozilla's lawyers
> currently revising the MPL...
>
> Benoit
>
> 2010/7/25 Benoit Jacob <jacob.benoit.1@xxxxxxxxx>:
>> 2010/7/24 Keir Mierle <mierle@xxxxxxxxx>:
>>> On Sat, Jul 24, 2010 at 3:31 PM, Benoit Jacob <jacob.benoit.1@xxxxxxxxx>
>>> wrote:
>>>>
>>>> 2010/7/24 Benoit Jacob <jacob.benoit.1@xxxxxxxxx>:
>>>> > 2010/7/23 Keir Mierle <mierle@xxxxxxxxx>:
>>>> >> +1 to CC-attribution
>>>> >> -1 to CC-share alike. That makes the license viral
>>>> >
>>>> > aaaaaaaaargh
>>>> >
>>>> > Everytime someone uses the word 'viral' to refer to copyleft, the
>>>> > aztec gods kill some kittens and Bill Gates has a burst of evil
>>>> > laughter.
>>>> >
>>>> >> like the GPL. Is in
>>>> >> really necessary?
>>>> >
>>>> > I just wanted to emulate the kind of weak copyleft as offered by the
>>>> > LGPL and MPL licenses.
>>>> >
>>>> > It could well be, indeed, that CC-BY-SA is actually closer to GPL: I
>>>> > just don't know!
>>>>
>>>> This is indeed the case, unfortunately: CC-BY-SA is a stronger
>>>> copyleft than what we want.
>>>>
>>>> http://creativecommons.org/licenses/by-sa/3.0/us/legalcode
>>>>
>>>> Helmut's use case would actually be allowed since that would be a case
>>>> of "Collective Work", I guess, but slightly more intricate cases would
>>>> be unclear.
>>>>
>>>> For example, I want to allow people to incorporate short excerpts from
>>>> our documentation, possibly in altered form, and have our copyleft
>>>> only cover the _part_ of their work that is a direct adaptation of
>>>> ours.
>>>>
>>>> The 2 questions below remain. The lack of simple weak-copyleft
>>>> licenses, here too, is a real pity, and might indeed force us to
>>>> release our documentation without any copyleft if we really can't find
>>>> a suitable copyleft license.
>>>
>>> What's wrong with CC-attribution? I see no reason for copyleft here.
>>
>> Indeed, I've been thinking about this, and CC-BY should be good enough
>> for us. The material in our documentation is mostly Eigen-specific,
>> so, the more it spreads, the better for Eigen.
>>
>> Others, what are your opinions?
>>
>> Benoit
>>
>>
>>> Keir
>>>
>>>>
>>>> Benoit
>>>>
>>>> >
>>>> > So we have 2 questions to answer:
>>>> >  - do we indeed need some sort of weak-copyleft for the documentation
>>>> > or are we OK to release documentation under a slap-me-in-the-face
>>>> > license?
>>>> >  - if we do want some sort of weak-copyleft for documentation, what
>>>> > license would offer that? Is the SA in CC-BY-SA a too strong form of
>>>> > copyleft?
>>>> >
>>>> > Benoit
>>>> >
>>>> >> Keir
>>>> >>
>>>> >> On Fri, Jul 23, 2010 at 3:43 PM, Benoit Jacob
>>>> >> <jacob.benoit.1@xxxxxxxxx>
>>>> >> wrote:
>>>> >>>
>>>> >>> Hi,
>>>> >>>
>>>> >>> You do well to ask: indeed, we forgot to pick a free license for our
>>>> >>> documentation. We need to do that for a variety of reasons. Debian
>>>> >>> among others is distributing our documentation:
>>>> >>> http://packages.debian.org/unstable/doc/libeigen2-doc
>>>> >>>
>>>> >>> There are 2 mainstream licenses that would make sense for us:
>>>> >>>  - GNU FDL
>>>> >>>  - CC BY-SA
>>>> >>>
>>>> >>> This page is a good read on this topic:
>>>> >>> http://en.wikipedia.org/wiki/GNU_Free_Documentation_License
>>>> >>>
>>>> >>> My feeling is that we're a little better off with the slightly simpler
>>>> >>> CC license. It seems simpler because (AFAIK) it just doesn't have
>>>> >>> provisions of Cover Text / Invariant Sections. It's also nice that CC
>>>> >>> licenses like CC BY-SA make it clear in their very name what they are
>>>> >>> doing. No strong opinion though.
>>>> >>>
>>>> >>> Our wiki has defaulted to the FDL 1.2 but since there are few
>>>> >>> copyright holders, it should be easy enough to relicense. We must have
>>>> >>> compatible licenses for the wiki and docs, to allow moving content
>>>> >>> between them.
>>>> >>>
>>>> >>> OK for CC BY-SA ?
>>>> >>>
>>>> >>> Benoit
>>>> >>>
>>>> >>> 2010/7/22 Helmut Jarausch <jarausch@xxxxxxxxxxxxxxxxxxx>:
>>>> >>> > Hi,
>>>> >>> >
>>>> >>> > I'm selling my lecture notes in C++ to my students below cost.
>>>> >>> > I'd like to attach Eigen's QuickRefPage (including the URL) as an
>>>> >>> > appendix.
>>>> >>> >
>>>> >>> > Are there any legal problems about this?
>>>> >>> >
>>>> >>> > Helmut.
>>>> >>> >
>>>> >>> > --
>>>> >>> > Helmut Jarausch
>>>> >>> > Lehrstuhl fuer Numerische Mathematik
>>>> >>> > RWTH - Aachen University
>>>> >>> > D 52056 Aachen, Germany
>>>> >>> >
>>>> >>> >
>>>> >>> >
>>>> >>> >
>>>> >>>
>>>> >>>
>>>> >>
>>>> >>
>>>> >
>>>>
>>>>
>>>
>>>
>>
>