[Osmf-talk] Results of OSMF Member Vote

[Matija Nalis]

On Mon, Dec 28, 2009 at 02:18:29PM +0100, Michael Kugelmann wrote:
> Frederik Ramm wrote:
> > If a significant amount of data is not relicensed, and if OSMF should go 
> > ahead nonetheless (which I doubt it will), then people will simply run 
> > the old data set on some server of their own[1].
> 
> looking at the current rate of information provided to OSM, we will get 
> the lost data back soon.

That rate of growth is assuming:

(1) deleting several years of one's work does not make one any less
    enthusiastic about contributing to the project (note that even if you
    yourself decide to accept the licence, if you edited over data of
    someone who can't or won't accept it or is simply not reachable, YOUR
    data will also be deleted)

(2) the whole issue does not result in one or more forks (which would then
    fight for contributors, duplicate data which will grow farther and
    farther apart from original shared dataset and thus making the chance of
    merging data smaller and smaller, and again loss of enthusiasm in many
    people). Enough forks could even kill them all as none might attract
    sufficient contributors to prosper even close as OSM did (or even stay
    alive).

If any of those does happen in noticeable amount, the rate of growth of OSM
could severely lessen (to say the least).

If it leads only to "oh we'll just have few months setback and then be
better then ever", than beautiful, I would have no big problem with that. 

However such outcome is IMHO pretty much on one edge of the probability
curve (the other, [equally improbable] end being whole project and all
effort being lost forever and people giving up on it and just going on with
Google Maps or whatever)

> IIRC while some discussions on SOTM or in mailing lists (could someone 
> please confirm or correct my statement), we would get an area back 
> within 1/2 or 1 year if all the data would be deleted (this was 
> mentioned within a discussion regarding vandalims and deleting of big 
> areas e.g. a country like Germany).
> So if we need to change someting on the license we should do it now: the 
> earlier the better (IMHO).

One of the questions is if the thing that should be done as soon as possible
is moving to the (more complicated) ODbL, or should CC0 (or some other
public domain alike license) be much wiser choice giving how much trouble
and grief are "intellectual property" (and soon to be other) laws causing ?

> Loosing of data is always a bad thing, I'm aware of that. But there's a 
> german sentence "Lieber ein Ende mit Schrecken als Schrecken ohne Ende" 
> (better to have an end with horror than horror w/o end).

Yeah, well... the question (in my mind) is weather current situation really
such horror ? Is OSM really currently so completely unusable and worthless
to everybody (because of that license issue), so it would be better that it
dies at once than to continue as it is ? If you think so, then yeah, the
German saying is applicable.

I happen to think it is not true. The project is *currently* very useful to
many people even if the license is problematic. The CC-BY-SA uncertainty
actually mostly affects just the companies and persons wishing to package
and redistribute or publically perform such data -- which are IMO a minority
of users.

So the questions are:

a) is fixing the CC-BY-SA because of that few users (who we want mostly for
   gaining popularity which we seem to be gaining fast enough anyway) worth
   the loss of data and risk of forks and alienating contributors ? and,

b) if it is worth the risk, would a CC0 be much better long-term solution
   than ODbL (or other similar law-encumbered license) ? 

   Note that ODbL solution has it's own sets of problems [1], so while it
   will help some folks (like TV stations wanting to show maps in evening
   news) it will make problems for other folks (like people wanting to
   import data released by goverments under CC-BY-SA or even ODbL itself).

   Which again provokes question (a) -- if we are going to alienate some
   people with both solutions, would it not be better not to stay with
   current solution (and at least avoid the grievance to be inflicted by
   the act of changing license itself) ? 
   
   Or if we are willing to suffer the loss inflicted by (any) license change
   in order to fix the license issue, at least change to CC0 then so then the
   problem could be fixed in maximum amount once and for all and the result
   could be used by biggest amount of users (with smallest amount of issues) ?

[1] http://wiki.openstreetmap.org/wiki/Open_Data_License/Why_You_Should_Vote_No

-- 
Opinions above are GNU-copylefted.