[OSM-legal-talk] License Cut-over and critical mass

[Kai Krueger]

Andy Allan wrote:
> 
> However, I'd be interested in hearing what you think. Could you put
> some numbers on what would make you feel comfortable? I've tried such
> an exercise myself (and came to the same conclusions as the LWG in the
> end) but that doesn't stop you from having an answer, and it might
> help with constructive discussion here.
> 

No question, coming up with such limits is insanely difficult, as it is
rather difficult to comprehend what these limits would mean. They are as you
mentioned also no doubt multi-dimensional which doesn't make it any easier.
So perhaps any limits would be more about a trust building exercise than
actually meaningful limits (which is why I would advocate them as minimum
requirements, rather than a definitive limit). 

My gut feeling towards these limits (which would like change in an
argumentation) would probably be something like: 95% of data (both nodes and
ways) has to remain in some form or another, 90% of data has to remain
unchanged (i.e. the complete history agrees to the change) and 80 - 90% of
active contributors (as defined in e.g. the CT) have to agree. At 10Gb
planet size, 5% data loss would still be 500Mb loss, which is still be about
twice the size of the UK, or roughly half of Germany. But given the strong
growth and keeping the vast majority of the community, that would probably
be not too bad to catch up again and should give confidence in the process.

Such limits clearly don't address the spatially unevenness, and I too really
don't know how to define spatial limits sensibly either, thus can only hope
that data loss would be distributed reasonably evenly.


But as has emerged in the talk thread, perhaps John Smith's idea of the
final vote achieves similar effects and is much easier to do.

Kai
-- 
View this message in context: http://gis.638310.n2.nabble.com/License-Cut-over-and-critical-mass-tp5279719p5292193.html
Sent from the Legal Talk mailing list archive at Nabble.com.