[OSM-talk] Critical Mass for license change-over

[Michael Collinson]

This is a report from the License Working Group and a request for 
feedback. If anyone can do translations or summaries for other language 
mailing lists, I would be very grateful.

Our moderators have agreed that this is a general topic of concern to 
the whole OSM community. If you are a continuing mapper, please feel 
free to respond and give your opinion. Only strictly "legal" questions 
will be pointed at legal-talk list.

As the license change process evolved, concern was expressed that an 
unacceptable amount of data might be lost from the current version of 
the OSM database and consensus was reached that phase 5 [1] - the actual 
license cut-over - should only happen when a "critical mass" was achieved.

The question I ask you is, do you agree that we have reached critical mass?

Here is our report.

I and the License Working Group think we clearly have reached critical 
mass and that the situation will only improve over the next few weeks. 
An intense effort is being made to reach still undecided mappers. We 
have already asked your help in the UK, Philippines, Canada and USA. We 
will go global soon. A number of decliners have also kindly allowed us 
to continue using their contributions after making sure that their 
concerns were known. A few more may still do so. The OSM Foundation 
board has asked us to target April 1st for the change-over.

First, the good numbers.

Several hundred thousand mappers are now actively mapping under the new 
contributor terms. Only 420 older contributors have currently explicitly 
declined. At least 97.1% of nodes [2] and  96.6% of highways [2] in the 
current database were created by continuing mappers. However, some of 
those may have been edited later. From up-to-date figures, [3], it looks 
as though 3.2M out of 120M ways are problematic in some way.  That is 
2.68%. It is declining. So, if we can use just one figure, I suggest we 
could be at 97.32% readiness ... feel free to challenge!

But what about negative factors?

- There are subjective criteria.  The removal of 100 hospital nodes may 
be far worse than than the removal of several million import points. ... 
Or the loss of a repeatable import may be bad because folks have editted 
over the top. It is difficult to judge whether this has a positive or 
negative bias overall.

- There are regional and country [2, 4] variations. You might be in an 
area where there are bigger problems than than implied by the figures I 
have given you.  The easiest way to see this is with OSMI License View 
tool [5] .

- We still have not been able to get responses from about 35,000 older 
contributors who have mapped at least one node. Sorry, this is an 
approximate figure at the moment. One impact of this is that there are a 
lot of folks who have mapped a small town, stopped mapping and have not 
responded.

- On a national level, there are still specific issues we are working on 
in Poland and the Czech Republic.  In Australia, Montenegro, Kosovo, 
Albania, Macedonia and, on a regional basis, in Germany there large 
concentrations of data by folks who have specifically declined. In 
Liberia and Cyprus, there are key large contributors who have so far not 
responded. In Japan, there is also one very large contributor who has 
declined, but we understand this is a POI import that will be dropped.

- http://odbl.de/ [4] gives a more pessimistic view than the numbers I 
have given you. This is probably due to bot edits and changes which are 
harmless, but should be taken into consideration.


And, lastly, you can see what the "new" map will look like if we changed 
over today at http://cleanmap.poole.ch/.  This is running on a small 
machine, so please be patient and try again later if lot's of folks are 
hitting it.


Mike
License Working Group

[1] 
http://wiki.openstreetmap.org/wiki/Open_Database_License/Implementation_Plan#PHASE_5_aka_Done_-_License_Cut-over_from_CC-BY-SA_to_ODbL_.28date_to_be_decided.2C_depends_on_the_technical_work.29

[2] http://odbl.poole.ch/ (based on early December data)

[3] http://tools.geofabrik.de/osmi/munin.html

[4] http://odbl.de/

[5] http://tools.geofabrik.de/osmi/?view=wtfe