[OSM-dev] Data corruption :) II

[Frederik Ramm]

Hi,

Stefan de Konink wrote:
> Frederik Ramm wrote:
>> It would also send out the (wrong) message that you can painlessly 
>> continue not caring for 0.6...
> 
> What is the problem with the last thing? If the community wants 0.6 they 
> talk 0.6... now you just break every editor that isn't 0.6 capable.

That's the point, yes. An editor that is not 0.6 capable will only be 
able to participate in a limited way because it will be unable to track 
version numbers on upload. So even if one would initially employ the 
timestamp hack I suggested, if you change the same object again in your 
editor you will definitely not be able to upload that change. A non-0.6 
editor will also not ask the user for a changeset description which we 
consider an essential part of our new changeset concept; the 
compatibility API would have to make something up, thus greatly reducing 
the usability of the data.

All this is acceptable for a one-off "I have this file lying around from 
the day before the big change" situation, but not acceptable for a 
prolonged changeover time.

Bye
Frederik

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