[OSM-talk] Timezones (was: Deleting data)
dieterdreist at gmail.com
Mon Oct 21 13:30:35 UTC 2013
2013/10/21 Pieren <pieren3 at gmail.com>
> It was only a consensus in the group of contributors thinking that
> (which is then easy to reach a consensus).
> This remembers me similars discussions about:
> - hi-res aerial imagery coverage by huge polygons (Yahoo!)
agree that this is not really a datum suitable (from a puristic view point)
for osm, but on the other hand there are mappers in certain areas with
worse coverage than your average European Country, where they misuse the
osm db for this kind of information (and it makes it easy for them to
maintain this data together) and then improve the map with stuff we all
want to have (based on these boundaries their workflow is much easier).
Yes, you could say that there could be a different workflow suitable
without putting these polygons into our db, but maybe they don't have to
capacities, and surely the overhead of these few and simple polygons is
very low, so I tend to suggest to let them do it.
> - parcels
IMHO there is really no good reason not to put parcel data, as many of our
other data actually depends on them (landuse and addresses), besides that
we're probably not prepared currently because of the sheer quantity.
- underground facilities (sewer, parkings, phone cables)
are a little bit difficult (verificability unless there are public
datasets) probably also maintenance (as this is something that very few
people map and hence verify / correct).
> - geologic stuff (mountain strings, stratifications)
geologic stuff (stratifications) doesn't belong to OSM IMHO (data not
surveyable/verificable by the crowd, resolution not adequate to the rest
of our data).
By "mountain string", are you refering to mountain ranges? IMHO ridges and
arêtes are suitable (simple lines, well defined, surveyable), making a
collection of these ridges to get a mountain range might be doable.
> All such features have the problem that they are often not verifiable
> (underground) or creates high density maps with many different layers
> where none of the OSM editors can handle easily and correctly
> different layers
Some of these are NOT different layers (except the geologic stratifications
and the aerial imagery coverage). As they depend one of eachother (e.g.
landuses depend on parcels, addresses depend (at least in some countries)
on parcels, even underground lines depend often on what is above ground and
vice versa) they should not be disjunct.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the talk