[OSM-talk] Suggestion to add SA clause to CT section 3, describing "free and open license"

[Kai Krueger]

Frederik Ramm wrote:
> 
> The proposed license change makes two concessions to the PD advocates. 
> One is that you get a (symbolic) chance of officially declaring your 
> contribution PD. This does not have legal relevance, as you cannot 
> extract PD data from an ODbL protected database without triggering 
> ODbL's share-alike, but at least the PD faction can make their voices 
> heard. The other is that the contributor agreement does not completely 
> rule out moving to PD at a later time, if a large enough majority of OSM 
> contributors should favour that.
> 
> These two concessions are really minor and are a long way from actually 
> making anything in OSM PD.
> 

Unfortunately I don't think the CT one is minor in anyway. What it appears
to be doing is trade off a potential, move to PD, that currently is
completely uncertain if it will or will not ever gain acceptance, with a
definite large loss of data now. At least in Australia and the UK
(http://www.maps.webhop.net/osm_opendata/ gives a rough (over) estimate of
the scale of Data effected in the UK from OS OpenData that is CC-BY
compatibly licensed), but looking at the
http://wiki.openstreetmap.org/wiki/Import/Catalog many other countries are
(potentially) effected on a large scale too.

Also the combination of a  SA license ODbL and the potential PD element in
the CT appear to have constructed a situation, where OSM becomes
incompatible in one way or another with everything including itself. It is
not possible to mutually exchange data with PD sources, as ODbL prevents
giving back. However, it is also not possible to incorporate any datasources
such as e.g. OSM that are compatible with ODbL, as the CT prevent that as it
may move to PD.  Therefore to me it appears as if OSM is thus moving itself
into isolation and with respect to the point of sharing data has chosen the
"worst" of both PD and SA worlds.

So with respect to "concessions to the PD crowd", I think this is
unacceptable, as it destroys a large part of what OSM is today. Therefore I
would much rather see as a consession a strengthening of the first point,
i.e. the "I consider my data to PD" actually mean something and that one can
somehow extract "clean" PD data (however you end up technically and legally
defining clean) indeed as PD data. I also don't see a legal reason why it
couldn't mean something. Apart from the definition of "clean", you as a
contributor give the the data in an unencumbered way to OSMF, so it could
decide to also publish it as PD without wrapping it into the ODbL. This
would then allow for the potential of a much more organic PD fork off of
OSM. Now I really wouldn't wont to see a fork and we should all try and
prevent that, but as you say, there are many competing views on licensing
currently in the project and one has to compromise to achieve a community so
giving the PD crowd an "easy way" to split off may turn out to be the least
problematic option. However, I do think the onus of forking has to be on the
people wanting to change the nature of the license, not those who want to
keep it. 

However, I am not sure that the term in the CT was originally meant as a way
to switch to PD or change the nature of the license in any other way. At
least that was not the way it was originally "marketed" during the OSMF
vote. Instead, the argument was that despite all the diligence put into the
new ODbL and therefore all the likelihood of it being better than CC-BY-SA,
it is unfortunately a yet unproven license and so god forbid, it may turn
out to have fatal flaws that can't be fixed with the upgrade clause to an
ODbL 1.1. That's why we need an extra backup in the CT, as indeed a second
switch of license as we have now can be for all intents and purposes be said
to be impossible.

That line of argument is imho very reasonable and one therefore well worth
having, but somehow we also need to find a way to make it more compatible
with more free and open licenses such as the rather liberally licensed
Ordanance Survey data, or the Australian CC-BY.  This brings us back to the
originally question of this topic. Will (or can) adding a "Attribution
Clause" in the CT make the construct of ODbL and CT compatible with more
free and open licenses  such as CC-BY and thus allow us to retain more data?
With that, the original intent of a backup for should ODbL turn out to fail
would still be met, yet it would endanger the project currently less.

I am not sure a SA clause would help here, as moving to a different SA
license would still make it incompatible, so you would still not be able to
agree to the CT for a SA licensed source, but I would hope that situation
looks a little more promising for attribution only licenses. Now the LWG has
repeatedly said it is looking into the compatibility to CC-BY, so hopefully
some of these considerations will be taken into account and we will soon
have a clearer picture how they are intending to solve this issue. 


Frederik Ramm wrote:
> 
> If NearMap imagery is so important for OSM in Australia - and there are 
> countries which have been mapped very well without aerial imagery of 
> note - then let's make an exception for NearMap, let's include their 
> data without them signing the CT. This would mean that if at any later 
> time the license is changed, NearMap would have to be asked specifically 
> if they like that license. I assume that this is something we will have 
> to do for some other sources as well.
> 
> No reason to drop or modify the CT for everybody because of that.
> 

Now that is an interesting suggestion and for that matter one that has been
mentioned several times in the past, including by the LWG (Quote from the
wiki bulk import support page) "We hope organisations donating data will
also use them but _can make alterations in specific cases_ provided that the
general spirit it kept. Provisions allow the OpenStreetMap Foundation to
publish the data, provided that it is always and only made under a free and
open license". (emphasis my own) So I am hoping that that might be the way
out for satisfying both the existing data contributors and still having the
ability to react to unforeseeable licensing issues in the future. However, I
haven't seen any further documentations suggesting to what degree OSMF is
prepared to make individual exceptions, under what conditions and who ends
up deciding these "individual CTs". 

Hopefully we will soon get some clear clarifying statements on these issues
(data loss due to contributor terms and a potential vote for the cut over)
from the LWG and the OSMF board, in order to put some of the fears (and
discussions) to rest. However, it will no doubt take a while to get these,
as th LWG will presumable have to check the legal ramifications of any
changes with its lawyers and the OSMF board isn't even defined at the
moment, as the counting of votes for new board members hasn't be finished... 

So a bit more patience I guess is required....

Kai
-- 
View this message in context: http://gis.638310.n2.nabble.com/Suggestion-to-add-SA-clause-to-CT-section-3-describing-free-and-open-license-tp5311401p5318614.html
Sent from the General Discussion mailing list archive at Nabble.com.