I know a couple of buildings where the ground floor is populated with commercial content: stores/shops, supermarkets, bakeries, post offices… But everything above the ground floor consists of residential apartments. By the way, there are no offices.
I googled a bit, and there seems to be a similar question and answer at help.openstreetmap.org. One user suggest that this building should be tagged as a landuse=retail, the other user suggests: landuse=commercial. But I do not understand the difference between these two tags.
I would be grateful for any kind of reply.
Thank you in advance.
buidling=yes and then provide details on the individual building parts.
Whether such a building is predominantly residential or retail will depend on the circumstances, e.g I would consider an older UK high street with flats above the shops as predominantly retail, but a new multi-storey block where the council insisted on retail space being retained on the ground floor might be predominantly residential. As such details can depend on motivations, I think it safer not to make the choice, and use the generic “yes”.
I think you are, again, seeking a level of precision of definition that doesn’t and can’t exist in a crowd sourced map.
(One could, of course, use “buidling=residential;retail”, but I doubt that will add much value as far as consumers are concerned.)
landuse=retail is for areas where shops are opens to the general public.
landuse=commercial is for areas with other kind of bussines, financial services, import/export of goods, etc.
For your case, I would map the building outline with building=residential, and just put nodes for the shop/services in the ground floor (completing the addr:* keys and also addr:floor=0), and also map the entrance=* of the building.
I would avoid mapping each shop with a node. At least not at the moment. This would require to get back to that city, walk to that area and write down all shops. I also do not possess a GPS device nor a mobile phone with GPS tracking, to record the locations of the shops.
At the moment, I just want to tag that this particular building (and some others in its vicinity) have shops on the ground floor and apartments above.
Still just to mention, that as a beginner I somehow find difficult the option of not being able to extract the building use information from the very building outline. And having to combine it with information from node tags inside that building outline.
For example: let’s say a person would like to use overpass-turbo.eu to select all buildings of a certain type (for example: “retail” buildings). This specific building would not be selected, as the outline will be tagged as: “building”= “apartments”/“residential”.
I’ve read on gis.stack-exchange that some GIS applications have a possibility to check for point inclusion in closed ways. So I guess with this possibility, a user who is looking for “retail” buildings would be able to select this building outline also, as there are “retail” tagged nodes (stores/shops) inside it.
But this seems like an advanced use of Openstreetmap data, and not all of us are able to do it. I certainly am not.
I googled a bit, and there seems to be one more solution OSM mailing list from 2015. It suggests the following tagging:
I quite like this solution as it enables extracting the building use information directly from the building outline, rather than performing complex point in polygon analysis in some GIS software. And it does not require semi-colon separate values.
I am just talking out loud. I am a beginner after all.
I already contributed, but not with bernard1995 account. I hope I am not violating any rules by having two of them.
Thank you escada.
I wasn’t aware of this building typology/use differences.
So now that you explained it to me, I guess there is no other way (if shop nodes are not used) than using the semi-colon separated values (building=apartment, building:use=residential**;retail;**commercial).
Thank you vvoovv. I hope the issue with two accounts is not a problem.