[OSM-legal-talk] What happens on April 1?

Frederik Ramm frederik at remote.org
Wed Mar 7 08:16:00 GMT 2012


On 03/07/12 04:06, Steve Bennett wrote:
>    Could someone explain exactly what will be happening on April 1?

I had initially assumed that we would take the database offline, drop 
all decliners' data, and then come back online. But it now seems that 
this might not even be required, and that it might be possible to use a 
bot to make the license change preparations in the live system.

"License change preparations" means that every object would be modified 
into an ODbL compatible state (worst case: deleted); after this bot has 
completed its work, the database would still be CC-BY-SA, but from that 
point on, OSMF would, at any time, be able to decree that "as of now" 
the database was ODbL.

> Will we really be purging all data from decliners? And if so, is this
> not terrible timing, given the recent, high-profile signups of
> companies like foursquare?

There are many aspects to this.

1. Any timing is terrible, so why not do it now.

2. We have no obligations to Foursquare; they have made a business 
decision in the full knowledge about the upcoming license change.

3. If they, or their tile provider, MapBox, don't like what they see 
after the license change, they may choose to remain with the last 
CC-BY-SA data set for however long they want.

> Given that many people are now actively remapping, is there any
> prospect of pushing back the cutover deadline?

If there really are people actively remapping and our rushing through 
the license change would sabotage their work and alienate them then yes, 
we should postpone for a month or two. Sadly, here in Germany many 
people are of the opposite opinion and they say "let's wait until after 
the license change, and then see what's missing and fix it". I would 
much prefer people to remap now but it seems that remapping is not for 

The current graphs - http://tools.geofabrik.de/osmi/munin.html - point 
steadily downwards but if you extrapolate you'll see that they are 
unlikely to reach zero before autumn.

> Is there any reason not
> to?

I think that a number of people on the OSMF board - Steve and Mikel at 
least because I've spoken to them in a management team conference call 
about a month ago, but likely others too - are of the opinion that OSMF 
must be seen by the world to be reliable and be in charge; they fear 
that if OSMF should now renege on the "1st April" promise they've made, 
then people might come to the conclusion that OSMF cannot be trusted. 
However they see a "trustworthy OSMF" as a necessary basis for dealing 
with the business community, and acquiring funding, data, or other 
support from them.

In the aforementioned management team telephone conference I said, "You 
can't tell me that April 1st is success, and April 2nd is failure" and 
was told that "the board thinks different". (This is from memory.)

(In my eyes, it is a very bad idea for OSMF board to "commit" themselves 
to something which is not under their control; and we must definitely 
avoid this kind of ambitious goal-setting in the future. OSMF can set 
goals for OSMF, but OSMF must not set goals for OSM. But that's a 
discussion we can, and should, have after the license change is through.)

This doesn't mean that a postponement cannot happen; certainly board 
won't simply shut down OSM on April 1st until the bot run is complete 
just to be able to say that they met their target. But it does mean that 
a postponement would need really solid reasons which would allow those 
on the board who "committed" themselves to the 1st April "deadline" to 
save face.

"If we wait another month then 5% more data can be remapped" is not a 
solid reason, and neither is "I'm sure Foursquare would be unhappy to 
lose a few roads in the US". These reasons are especially bad because 
they an be repeated month after month and thus could make the process 
drag on endlessly.


Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"

More information about the legal-talk mailing list