Open aoles opened 2 years ago
this seems related to #925 and the corresponding PR #927
Thanks @koebi ! The snapshot date of the OSM file is a different thing than the date on which it is being imported and parsed by ORS and when the graphs are being created. The former one for example is useful to determine whether a given change in OSM is expected to be reflected in ORS, the latter one makes difference when it comes to e.g. seasonal road closures. Only recently questions relating to these tow separate issues came up the forum: https://ask.openrouteservice.org/t/how-can-i-apply-my-paths-to-the-directions-service-with-the-openstreetmap-web-editor/3970 and https://ask.openrouteservice.org/t/disable-winter-restriction/224/14, respectively. Therefore I would argue that we should include both the OSM snapshot date and ORS import/build date.
Apparently, the planet.pbf
has a timestamp. This would have to be checked, or the date of the latest change set could be used if it's not in the metadata.
Either way, this seems to be a bit more involved and not something that will happen soon.
It would be helpful if the
metadata
section of the geoJSON API response would include, apart from the ORS graph build date ingraph_data
, also information on the OSM data version on which a given build is actually based on.