Comments on: CC and data[bases]: huge in 2011, what you can do

By: Benjamin Jean

Benjamin Jean — Wed, 25 May 2011 10:25:15 +0000

Oops, sorry to reply so late …
A few points to substantiate my thoughts:
* OSM can not change its policy in « one shoot », but their will are, I guess, clearly to move to ODbL (and I think it’s a good, hard, choice) ;
* the Databases are not meant to be original, but rather to be functional: CC licenses would have therefore exceptionally effects (and more databases will be used, improved, etc.. more they will be harmonized / standardized – become non original)
* I do not understand how you can “simply” add a right (sui generis) with a so fundamental effect (not accessories as we have seen on other licenses – including GNU or CC) for people using your licenses.
* The rights of the owner of a database are very different from those of the author (on its work). So the way the license is written must also be different (in my opinion).
* Regarding the AGPL license: actually the example was effectivly irrelevant and a second version was drafted to make it compatible with the GNU AGPL only. Nevertheless, I was just arguing that promouving the use of Creative Commons licenses on the grounds that version 4 will intergrate these specificities is /so/ dangerous. Especially since I think the current licenses would be well advised to remain limited to the content – a new license can be write for databases (but OKF has already done a great job)
* I personally think that the speech should be that 1) the data is free (without any license) 2) the original data / content should be free (subject to a license CC) 3) databases should be free( ODbL type).
* Regarding ODbL (which combines IP assignment – Copyright and sui generis databases – and contract), I think its effects are quite well-controled (it also seems to me that it not extend to uses permitted by the sui generis database).
* Finaly, I think CreativeCommons’d rather work on the current database license (ODbL) to improve together their work (quite good, but I think some improvements could be done).

By: Katie Filbert

Katie Filbert — Sat, 19 Feb 2011 00:41:31 +0000

Re: “OSM has not yet changed their license to ODBL. They are in a multi-year process of doing so”. OSM is requiring new users to agree to contributor terms that give OSMF ability to relicense content as OBDL. On April 1, OSM will lock out existing contributors if they do not agree to the contributor terms by them. (http://lwn.net/Articles/422493/)

By: Mike Linksvayer

Mike Linksvayer — Tue, 08 Feb 2011 18:42:41 +0000

Hi Benjamin, thanks for your feedback.

* OSM has not yet changed their license to ODBL. They are in a multi-year process of doing so. We are indeed learning from this and talking to OSM.

* CC licenses extend to DB as whole to extent such falls under copyright, and with some caveats for some versions, apply only to copyrightable parts. It is certainly possible that 4.0 will extend to license sui generis consistently with copyright. This is all explained in the post, so I’m not sure what your additional point is.

* I’m also not sure what you mean by database rights being more like competition laws than copyright. Both can be characterized as incentivizing or inhibiting competition via automatic grant of rights. Feel free to explain.

* If a database is completely uncopyrightable, a CC license will do no harm. There’s a lot of uncertainty about what is copyrightable (if you look at various discussions, eg the one on LinuxWeeklyNews linked in the post) you see that much of discussion consists of conflicting assertions about what is and is not copyrightable. CC generally doesn’t take a stance on these points. Using a CC license ensures at a minimum some freedoms will be available even if copyright does apply.

* You’re wrong about AGPLv1 and GPLv3. AGPLv1 is not compatible with GPLv3. Former says “You may also choose to redistribute modified versions of this program under any version of the Free Software Foundation’s GNU General Public License version 3 or higher, so long as that version of the GNU GPL includes terms and conditions substantially equivalent to those of this license.” Which GPLv3 does not. Instead FSF published AGPLv3, and Affero published AGPLv2 as an upgrade path to AGPLv3. Nobody who released software under AGPLv1 had their software fall prey to ASP loophole as a result of FSF’s actions. I think FSF has set a good standard for license stewardship, one that CC aim to uphold, if not surpass. 🙂

* There are many data extraction and modification scenarios, some of which are covered by copyright, some by database rights (note except for EU 3.0 ports, no CC license currently even conditionally waives these), some (pure facts, whatever that might mean) by nothing at all, though as above, there’s plenty of uncertainty about what is what. The only anti-exploitation feature ODBL adds is a contractual layer, and it isn’t at all clear that it adds “protection” (because such very hard to enforce) and is probably bad policy. These are tradeoffs (and not remotely the only ones).

Look forward to your continued participation in the discussion, thanks!

By: Benjamin Jean

Benjamin Jean — Tue, 08 Feb 2011 14:43:24 +0000

Hi,

Thanks for the enlightenment, I always thought that CC should be as clear as possible about what the licenses actually permit and what they don’t. We need explanations to trust, and it’s impossible to rely on CC movement without transparency and honesty (and I’m personally disappointed in CC about this subject).

Nevertheless, I’m not convinced by your argument:
– why don’t you speak about the OdbL? This license was created because of the lack of confidence on actual CC licenses (when applied on data). OSM changed their license from CC to ODbL: it seems like a huge process and I hope you will take it in account during your work. I think CC need to learn and to bear.
– the “droit sui generis” about DB can’t be assimilated to copyright laws: the object of this right and the prerogatives are different (because the aim is different – it’s much more like competition laws than copyright laws), so I don’t think CC licenses extend to DB as a whole (they actually only apply one the copyrightable part of these DD)

Finaly, I can’t agree to your recommendation:
“(4) is a corollary of (2) and (3): use CC licenses for data and databases now, participate in the 4.0 process, and upgrade when the 4.0 suite is released, or at least do not foreclose the possibility of doing so.”
=> thus you ask us to use CC licenses on DB, even if they actually don’t protect no copyrightable DB, to wait and rely on your next version? This solution seems to me irresponsible:

1) what happen if you finally change your mind? Who will be responsible? Creative Commons ? Or people who choose to trust CC? I’d like to recall an event quite similar with the FSF: when the first Affero GPL was wrote, the FSF claimed that because the license was compatible with the next GNU GPL v3, people can use Affero GPL without fear about the ASP loophole. However, the GNU GPL v3 finally didn’t add these term (instead an other GNU Affero GPL license was created) and all works licensed under the first Affero GPL (fortunately not too much) were available without the ASP loophole.

2) more dangerous : what is the consequence if people choose to use the DB under the last license (2, 2.5 or 3 ?)? They can extract as much data as they wish, modify these data without give anything back. I suppose the new licenses won’t appear before one or two years, why is the reason to not use better license for data (I only know the ODbL, but maybe other license exist).

BTW, I’ll try to follow the cc-licenses ML, it’s an interesting topic.