[OSM-dev] how many api calls/time for an app allowed?

Gary68 gary at gary68.de
Fri Jul 3 14:45:48 BST 2009


ok,

then maybe we have to try :-) but of course i could implement some speed
bumps and maybe it would even be enough to do the checks once a day. and
maybe then we can find a time slot. let's see.

the checks i would run would include

- touch checks (nearly touching nodes/ways, improperly not connected)
- cross checks (crossings of highways on same layer, without bridge,
tunnel etc.) 
- roundabout check, direction
- maybe more...

the idea is to provide good data, not outdated. currently i run those
and other checks on planet excerpts once a week. but of course in doing
so i calculate large untouched areas and data might be old (in someones'
opinion) ... 

when there will be a bug clearinghouse i will provide the data there.
for germany that is. i don't have a supercomputer and that's where the
number of ~500 is estimated for.

cheers

gerhard


Am Freitag, den 03.07.2009, 14:27 +0100 schrieb Tom Hughes:
> On 03/07/09 14:01, Gary68 wrote:
> 
> > i am playing with the idea of providing "real time" error checking on
> > osm data. usually one would use the XAPI. but since it is unreliable and
> > obviously misses data and it seems no one really cares (no ansers to
> > questions about that matters on the mailing list - at least not
> > satisfying ones) one would use the API instead.
> >
> > so let's say i would need ~500 calls a day for an application, each
> > requesting a "tile" of 0.01 by 0.01 deg. would i or my application be
> > stopped doing so? it is only read access!
> 
> There is no hard and fast limit - the limit is the point at which your 
> activities cause me to have to start looking at the server to try and 
> work out why performance has been impacted. If at that point you appear 
> to be the cause of the performance impact then you get yourself banned.
> 
> I'm not quite sure what sort of error checking you can provide with such 
> a small number of calls for such a small area though... Presumably 
> whatever the checking is only being done for a small part of the planet?
> 
> BTW saying "only read access" doesn't help - read access is typically 
> more expensive that write access for us.
> 
> Tom
> 





More information about the dev mailing list