Friday 18 September 2015 09:47:12, Frederik Ramm:
> When someone adds copious "is_in" tags to things, or when they create
> relations like "all cycleways in Cambridge", we tell them: Don't bother
> doing that, we use a spatial database, and we can compute these things
> from the data.
>
> I'd say the same applies to houses. Whether something is one half of a
> double house, or semi-detached, or terraced, or free-standing - isn't
> that something that I can automatically determine by looking at the
> nearby mapped buildings?
>
> Or is the redundancy desirable so that we can then write new debugging
> tools ("house at end of row but not tagged as semi-detached", etc?)I thought that at first about building=detached as well. If I remember correctly, someone said that figuring it out is a database query that takes long. The original import tagging page [1] says semidetached_house is used for Doppelhäuser which is not exactly the same as just two houses that share a wall, right? For example: I wouldn't consider these houses[2] semi-detached because they have a very different architecture and no shared roof, while I would have to think twice about these ones[3]. Their architecture and colour is the same, but they have a flat, not-really-shared roof IIRC. According to Wiktionary, it is called a duplex in American English, in British, Australian and Canadian English a semi and in all Englishes a semi-detached. That would make the "house" part even unnecessary, but useful to understand the meaning of the tag (unlike building=detached). [1] https://wiki.openstreetmap.org/wiki/Rostock:Geb%C3%A4ude_und_Strukturen [2] http://osm.org/go/0EhfysLT0?m= [3] http://osm.org/go/0EhfynMd2?m=
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
