Open dieterdreist opened 3 years ago
lonvia
commented
3 years ago I see that Rome has areas for administrative boundaries as well as place areas. Could you quickly outline what the expected hierarchy would be here between this areas? And is this special for Rome or do you have the same scheme for all of Italy?
dieterdreist
commented
3 years ago sent from a phone
On 17. Dec 2020, at 10:30, Sarah Hoffmann notifications@github.com wrote:
I see that Rome has areas for administrative boundaries as well as place areas. Could you quickly outline what the expected hierarchy would be here between this areas?
I would show the complete administrative hierarchy down to the smallest rank (regione, provincia, comune, municipio where there is) and maybe after this the 1-2 smallest places (if any, namely quarter and neighborhood/hamlet, maybe suburb/village/town), but would generally give priority to admin objects. They both exist, and are both officially defined (toponymy / toponomastica is not complete in OpenStreetMap or at least wasn’t when I last checked).
And is this special for Rome or do you have the same scheme for all of Italy?
not sure, I would not expect them everywhere but would not be surprised if there were some other places in Italy with a similar, “dual” classification.
I have seen it in Tuscan towns where I guess the smallest admin subdivision was comune (level 8), while the town was divided into several “contrada”
lonvia
commented
3 years ago I haven't expressed myself well. I was not askingwhat you would like to see in the geocoding results. I was asking about the current mapping practises. What admin boundary and place area types are used in Rome and how do they relate to each other?
dieterdreist
commented
3 years ago I haven't expressed myself well. I was not askingwhat you would like to see in the geocoding results. I was asking about the current mapping practises. What admin boundary and place area types are used in Rome and how do they relate to each other?
the current practise is to map at least all admin entities which have their own political representation (president and council, somehow similar to a mayor in smaller cities), i.e. admin_level=10 which are called "Municipio". These are completely mapped. For admin_level=11 there is no general agreement as far as I am aware of, but some are mapped (representing the "zona urbanistica", e.g. https://www.openstreetmap.org/relation/10319358 ).
There are also (AFAIK with the same border) place objects, whose type changes their name (and nature) according to the position (their history), namely the "Rione" (medieval quarters in the historic center), "Quartiere" and "suburbio" (a belt of more recent quarters) and "zone dell'Agro romano" (at least in past times these were outside of the limits of the contiguous city). Both, the aforementioned kind of place objects (called "... toponomastico") and the zone urbanistiche, have different names and refs. I am not sure about completeness, but I just fixed a wikidata link which was added by a supposedly remote editor with some automatic matching tool from the wikipedia tag, because there is only one wikipedia article for both kind of objects (indeed in wikidata there are 2 objects for this, but for unknown reasons only 1 wikidata object can be linked to a wikipedia article and you cannot link the same wikipedia article from more than one wikidata item).
https://it.wikipedia.org/wiki/it:Ostiense?uselang=en https://www.wikidata.org/wiki/Q1114050 https://www.wikidata.org/wiki/Q101687164
Place on the other hand is also used for various different kind of other "things", place=village / hamlet may be used as nodes on historic settlement nuclei, local area names might have their own tags like quarter, neighbourhood, etc.
While for admin entities it is guaranteed that they are strictly hierarchically structured and covering the whole territory on all admin levels without partial overlap (borders of higher and lower entities are at the same positions, you do not find a smaller entity not fitting completely into the next higher entitiy), this is not necessarily so for place objects (might partially overlap, might have uncertain hierarchy details).
There is an uncomplete work in progress overview page about the subdivisions of Rome here: https://it.wikipedia.org/wiki/Suddivisioni_di_Roma
What did you search for?
reverse geocode points in Rome, Italy
What result did you get?
arbitrary results, sometimes with the place objects, sometimes admin boundaries. E.g. https://www.openstreetmap.org/search?whereami=1&query=41.8661%2C12.4036#map=14/41.8661/12.4036 restitutes Casetta Mattei, Suburbio VIII Gianicolense, Rome, Roma Capitale, Lazio, 00164, Italy
the very similar https://www.openstreetmap.org/search?whereami=1&query=41.8663%2C12.4087#map=14/41.8663/12.4087 gives me: Municipio Roma XII, Rome, Roma Capitale, Lazio, Italy
What result did you expect?
I would generally expect the Municipio to be given. I guess Casetta Mattei in the first result comes from here: https://www.openstreetmap.org/node/262115845 it might also come from here, maybe? (there is maybe a data problem here, I will check this independently, but the problem that place is preferred was also observed in other Municipi) https://www.openstreetmap.org/node/5697223742
For example https://www.openstreetmap.org/search?whereami=1&query=41.8646%2C12.4805#map=16/41.8646/12.4805 gives: Via Giuseppe Libetta, Garbatella, Ostiense, Quartiere X Ostiense, Rome, Roma Capitale, Lazio, 00154, Italy
It is (almost) all correct, but I would expect to also see the Municipio: https://www.openstreetmap.org/relation/1459589
On a side note, actually I would not think this point is in Garbatella, because Garbatella stops at the railway. But it is "Ostiense" (in both senses, as being the area of the road via Ostiense and as being inside the toponomastic "Quartiere" X Ostiense ("Quartiere" is an official toponomastic area, within the center they are called "Rione" and further away "Suburbio"). I am aware that the Garbatella problem cannot easily be solved with the current data (node), this issue is about the Municipio information not being displayed.
Is the result in the right place and just named wrongly?
Please tell us the display name you expected.
Is the result missing completely?
Make sure that the data you are looking for is in OpenStreetMap. Provide a link to the OpenStreetMap object or if you cannot get it, a link to the map on https://openstreetmap.org where you expect the result to be.
To get the link to the OSM object, you can try the following:
Further details
Anything else we should know about the search. Particularities with addresses in the area etc.