[OSM-talk] Wanted feature : VALID_PER tag ?

[Matt Amos]

2009/6/8 ce-test, qualified testing bv - Gert Gremmen <g.gremmen at cetest.nl>:
> The essential thing of my proposal (however it be implemented)
> is that NOTHING changes with the current software or apps.
>
> A standard request will just deliver the current state of the map.
> The API will filter out all “deleted” (just as now) and additionally all
> object that have a valid_per tag that evaluates to false (new).
> If you think about it, deleted just means:  valid_per= false per today

so the API isn't current software?

> JOSM might request “CURRENT + FUTURE” to let people edit future stuff.

wouldn't this require a change to JOSM?

> The  valid_per data needs not be implemented in OSM
> but can be retrieved from a separate “bus routing” database,
> where the OSM data just tell the user where to find “valid_per”
> data.

that sounds like an awesome idea. even better if the external database
says where to get the OSM data, so the OSM data doesn't need to be
changed.

> I , just like Peter, do not see how problems in mixed data
> (the roman way) could get in the way, as nothing is deleted,
> just becomes invalid for standard application.

thanks, but we already have a few problems with ways referring to
"invalid" nodes. i don't really want more (and more complex,
time-dependant) problems.

cheers,

matt