[OSM-talk] Historical Data in OSM database

[Pierre-Alain Dorange]

M?rtin Koppenhoefer <dieterdreist at gmail.com> wrote:

> > Perhaps what is actually wrong IS differentiating between version history
> > and 'changes to reality'.
> 
> 
> I think this has to be done, and it can be done. We could invent a way
> to flag stuff that we remove because it ceased to exist as such, and
> could thereby get some actual history information in. But most of the
> stuff we currently delete gets deleted because it was wrong --- be it
> newbies, import/software bugs or vandalism.

It's a good idea.
For the long future of OSM it would nice to differentiate kind of object
delete. As you said actual deletion are just deletion. 
But for historical reason (long term) we perhaps think of a way to
specify (probably require an API evolution) that the object has been
delete because it has beeen demolished of has change its destination
(ie. living building become something else).
There is no simple solution to do this but it could be a mid-term
reflexion for a future API evolution.


-- 
Pierre-Alain Dorange
OSM experiences : <http://www.leretourdelautruche.com/map/>