After OSMF was originally wanting to start the license change at
the beginning of April, due to technical problems with the
redaction process of non relicensed data, it has been delayed till
The OSMF has now announced a new schedule . It will begin
removing data for which it does not have permission to relicense
from CC-BY-SA to ODbL 1.0 this coming Wednesday 11th.
It will happen as a slow process while normal editing of OSM data
continues. From a data consumers perspective the removal will look
like any normal deletion of data and will be mixed in the minutely
/ daily diffs with the normal changes. The duration of this
process is not yet clear due to the uncertainty caused by the load
of the normal edits, but very rough estimates supposedly range
from a week to several months, but hopefully it will be completed
within a month.
Luckily, worldwide only a little more than 1% of the total current
data has to be removed, but regionally this can unfortunately
still be a lot more despite all the efforts to minimize the
The question now is how should the wikipedia osm-tile-servers
handle this situation?
1) continue as normal, importing the minutely diffs. This has the
advantage of staying up to date with all the ongoing mapping
efforts, but the disadvantage of degrading the data in areas where
the redaction-bot deletes things.
2) Suspend the diff imports during the redaction phase: This
allows mappers time to fix some of the damage done during the
process, but removes the up-to-dateness of the map.
This affects at least the tileserver run on toolserver.org which
serves tiles for the osm gadget for various wikipedias and the
WIWOSM project. But depending on when the new tileserver run by
the foundation for its mobile project goes into production, it
would also need to make this decision. I don't currently know how
the mapquest open tiles will handle the situation.
During the redaction period, all data continues to be CC-BY-SA.
When the actual cut-over to ODbL will occur is not yet clear, as
the timing of the redaction bot is still uncertain but presumably
very shortly after it has finished.
On 03/22/2012 02:15 AM, Tim Alder wrote: