[od-discuss] a clause to achieve compatibility

[Kent Mewhort]

I agree that dual-licensing (or multi-licensing) would be a step further 
and would certainly be ideal. However, as least for the Government of 
Canada, I don't think it's a possibility this round; hopefully in the 
future.

The problem I see with relying on implicit compatibility ("whittling 
down the problematic terms", as Mike says) is that it leaves a large 
degree of legal uncertainty.  This is especially a concern where a 
licencee combines multiple licences -- and these can quite easily number 
10 or more different licences when it comes to data.  Faced with this 
scenario, many users might simply combine the works and take the word of 
the supplier lawyers and others that the licences are indeed 
compatible.  This means that the user is taking on quite a bit of legal 
risk.  Then, for institutional and business users, the word of others 
isn't going to cut it.  The in-house lawyers will have to go through 
each licence and assess it at great expense.  When it comes to more than 
two licences, the end conclusion is always likely to be that there are 
some ambiguities and therefore legal uncertainly and legal risk.

I do think an explicit right to sublicence under CC-BY is much more 
certain than a general right to sublicence.  The explicit right makes it 
clear that the user is authorized to licence the work specifically under 
CC-BY.  As for a general right to sublicence, this seems like new ground 
and very ambiguous: I can't think of a single common open licence that 
grants one, so it seems far from clear what it implies (if anyone knows 
of existing open licences that make this grant, please let me know).  In 
private commercial licences, the sublicense clause is usually 
accompanied by specifics that set out details such as whether any 
royalties still must be paid for downstreaming licencing, whether the 
sublicence includes a further right of sublicence, etc.

Granted, permissive licences such as BSD have done fine without explicit 
sublicensing clauses.  However, BSD is also only 2-3 three clauses long 
and I don't see governments getting their terms whittled down this much 
at the present time.  Moreover, I would argue that a explicit right to 
sublicence is even broader and easier to deal with. For example, let's 
say I combine a BSD-licenced work with a CC-BY work, then make some 
modifications.  We get:
   [CC-BY portion] -> [Still under CC-BY licence]
   [BSD-licensed portion] -> [Still under BSD licence]
   [My changes] -> [Under a compatible Licence C of my choosing]

Thus, the derivative work is actually under three licences and 
downstream users must comply with the terms of ALL three licences (there 
just aren't many terms wrt to BSD, so it's easy enough to do and you can 
safely ignore the fact that it's even still under BSD).  On the other 
hand, let's say the "OGL-C Hypothetical" includes a right to sublicence 
under CC-BY and I combine the two. We get:
   [CC-BY portion] -> [Still under CC-BY]
   [OGL-C H portion] -> [*Now licensed under CC-BY to downstream users*]
   [My changes] -> [Under a compatible Licence C of my choosing]

If I choose CC-BY for my changes, the whole derivative work is now under 
CC-BY, rather than under two or three different licences in the 
non-sublicencing case.  This is easier, IMO, especially when we're 
talking about five, six or ten licences rather than two.

Now, of course, if a non-specific sublicencing clause also allows me to 
do the same, so much the better.  I'm just a bit uneasy without some 
case law, or at least a existing widely-used open licence, that backs-up 
the ability to do so.  I'd loved to be pointed out to such as thing and 
proved naive though!

Kent


On 13-02-05 05:46 AM, Mike Linksvayer wrote:
> On Mon, Feb 4, 2013 at 3:25 PM, Tomoaki Watanabe
> <tomoaki.watanabe at gmail.com> wrote:
>> I think it is good if many of the national open data/
>> open govt licenses had a clause saying something like this:
>>
>> - if you use this Work in combination of any other work
>> that is under a Creative Commons Attribution License,
>> this Work is licensed the same Creative Commons
>> Attribution License for that use.
> Making works released under this OGL-hypothetical effectively
> multi-licensed under CC-BY and whatever else (with caveat of with
> respect to use with works under these licenses), right?
>
> I guess this would be a step further than Kent Mewhort suggested in
> http://www.creativecommons.ca/en/oglc-consultation --
>
> "Although the OGL-C draft takes the step of including a right to
> sublicense, its broad stroke does not set out any details and the
> legal implications remain uncertain. For example, the law is clear
> that a sublicense clause allows a user to re-license her or his rights
> under a different licence; however, does the user need to re-propagate
> the letter of each obligation? To clear up this confusion, we suggest
> that the final licence should grant an explicit right to sublicense
> under a Creative Commons Attribution licence."
>
> Which I was intending to comment on eventually, but it is related, so
> I will now.
>
> I'm 110% for more compatibility among licenses, but this strikes me as
> a little odd (major caveat: IANAL). Unless the grant goes further, and
> says that only the CC-BY license needs to be complied with (maybe as
> Tomo suggests), I'm not sure how explicit right to sublicense under
> CC-BY clears up described confusion more than just explicit right
> sublicense.
>
> Let's say a hypothetical OGL-C, same as the proposal but with a "You
> are free to...sub-Licence under CC-BY" clause in addition to "You are
> free to...sub-Licence"; is it really then more clear that a downstream
> user has only to comply with CC-BY, and can ignore any problematic
> clauses of OGL-C?
>
> But the real reason it strikes me as odd is that
> non-copyleft/sharealike licenses have never before had explicit
> compatibility clauses with named other licenses, have they? The reason
> the OGL-UK statement about "aligned to be interoperable" with CC-BY
> and ODC-BY is odd is not because it isn't an explicit compatibility
> statement, but because it doesn't really seem to be well aligned.
>
> Multi-licensing would be more unambiguous, I guess, but a concern I'd
> have for either is setting expectation for permissive licenses that
> explicit compatibility of some sort is necessary, otherwise you might
> not want to combine works under different permissive licenses.
> Currently that kind of expectation is limited to copyleft/sharealike
> licenses, where it is obviously necessary. But it is also the #1
> drawback of copyleft. I'd hate to have a desire for better
> interoperability among permissive licenses to lead to more fear and
> uncertainty about the same. That would only be good for lawyers. :)
>
> My instinct rather to push governments and other license creators and
> stewards to whittle down the problematic conditions of their
> permissive licenses and allow sublicensing, such that interoperability
> is a non-issue.
>
> But my instinct/concern/analysis may be woefully wrong. Please tell me.
>
> Mike
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/od-discuss/attachments/20130205/fadb9d97/attachment-0001.html>