Opened 14 months ago
Last modified 13 months ago
#7550 new defect
admin_centre should handle for multipolygons
| Reported by: | iav | Owned by: | team |
|---|---|---|---|
| Priority: | normal | Component: | Core validator |
| Version: | latest | Keywords: | |
| Cc: | Larry0ua |
Description (last modified by iav)
admin_centre should work for places multipolygons as for boundary as in card #6792.
role admin_centre should be point, optional and not more than one.
In case of name property is present for relation and point, it should be equal for relation and admin_centre point.
Attachments (1)
Change History (11)
comment:1 Changed 14 months ago by Don-vip
comment:2 Changed 14 months ago by iav
- Description modified (diff)
comment:3 Changed 14 months ago by iav
- Description modified (diff)
comment:4 Changed 14 months ago by skyper
+1
I think admin_centre is not a offical role for multipolygons only for boundaries but I can not find a reason for it.
Right now I move all tags to the relation and use the node in the outline to keep the history but that is not the best solution.
comment:5 follow-up: ↓ 6 Changed 14 months ago by iav
At least for places objects - it helps to place a place name text to more suitable place on map
comment:6 in reply to: ↑ 5 ; follow-up: ↓ 7 Changed 14 months ago by skyper
Replying to iav:
At least for places objects - it helps to place a place name text to more suitable place on map
and mapnik renders the name twice ! (In my area often even more as the multipolygons contain several areas and there exists separate multipolygon relations for boundaries.)
Do not get me wrong. I am in favour of "admin_centre" but we need to make this official and have mapnik use it !
comment:7 in reply to: ↑ 6 ; follow-up: ↓ 9 Changed 13 months ago by simon04
Replying to skyper:
Do not get me wrong. I am in favour of "admin_centre" but we need to make this official and have mapnik use it !
The part that JOSM can fulfil is to get rid of "Role admin_centre unknown - Role verification problem" – did you mean that?
comment:8 Changed 13 months ago by stoecker
Although Frederik introduced it, the type=multipolygon for boundaries is still somewhat broken. I'm not in favour of improving the support for this.
Mainly it would mean to define types base on relation type and additional key (in this case administrative), which is not supported yet.
The problem is, that type=multipolygon administrative=... is a boundary, but type=multipolygon no "administrative=..." is something else (an area on the earth). I don't like the idea of introducing data types of "this is an area except it is ...". As soon as somebody introduces something new, it will break all software.
All the other multipolygons are some way of land use, but boundaries are an additional layer overlaying the land uses.
comment:9 in reply to: ↑ 7 Changed 13 months ago by skyper
I map on both sides of the Rhine. Once I changed some multipolygon to boundaries on the right hand side of the river and received a few days later a not very kindly email asking me to revert my changes and stick to local custom. Seems to me that the wall in many minds presists concerning (national) borders.
No border, no nation ...
Replying to simon04:
Replying to skyper:
Do not get me wrong. I am in favour of "admin_centre" but we need to make this official and have mapnik use it !
The part that JOSM can fulfil is to get rid of "Role admin_centre unknown - Role verification problem" – did you mean that?
Or setup a proposal and create 1000 relations to call it official.
While this would be in my favour I am not sure about it as the warning seems to be correct for now.
comment:10 Changed 13 months ago by iav
Ok, would you, please, to check a usage count of type=multipolygon place=*?



Can you attach a sample .osm file ? Thanks.