[OSM-talk] Potlatch :-)

[Richard Fairhurst]

Gerald A wrote:

> Does that mean that the current work is useless? Not at all. But we
> have to take such complaints with a grain of salt. OSM has advanced
> leaps and bounds, it is still an evolving project. That means that
> errors are going to happen, and we could even have an API issue where
> it deletes or mangles something. Everyone wants true API based undo,
> and it will get there, but it will take time.

Sorry, I've not explained myself properly (wouldn't be the first time!).

Potlatch's waiting-to-go-live undo is true undo implemented through  
the API: in other words, you can undelete ways that have been deleted  
by another user previously (with any editor), or revert a way changed  
by someone else to an earlier version. It's non-destructive, so you  
don't lose the later revisions.

Because Potlatch's API is subtly different to the API used by  
everything else, this won't be available to other clients until the  
main API is changed. But I hope it'll set out the stand for the main  
API to do it in the future (a bit like Potlatch was effectively  
segmentless before the rest of OSM was).

Thanks for the encouraging comments as ever. :)

cheers
Richard