On 31 January 2018 at 11:13, Will Phillips <wp4...@gmail.com> wrote: > I favour using addr:parentstreet rather than addr:substreet for the > following reasons:
+1 > 1. The majority of OSM data users and tools/services using OSM data don't > know that either addr:substreet or addr:parentstreet exist. They will > recognise addr:housenumber and addr:street. Therefore I think the most > important part of the address should use these tags, which is usually the > dependent street and number. If someone is searching for an address, I think > this is usually the part they will enter. > > 2. When mapping addresses I consider the addr:housenumber and addr:street to > go together. Otherwise it is ambiguous, because there is no way of knowing > whether the number relates to the dependent or parent street. This can get > very confusing because the same numbers will often be used for addresses > along the main part of the street and again on any subsidiary parts. If we > wanted to use addr:substreet without any ambiguity, we would need something > like addr:substreetnumber as well. > > 3. Using addr:substreet is just more confusing. New mappers aren't going to > understand it. If a street has its own name and is separately numbered, it's > intuitive to put this into the widely recognised addr:housenumber and > addr:street tags. I know from surveying a lot of addresses that is difficult > to guess whether Royal Mail considers a street to be dependent or not: such > streets often don't look any different from other nearby streets that > aren't. It's not something mappers should have to worry about in the > majority of cases. > > Just to add, I don't think addr:parentstreet it always the best approach. > Where a 'dependent street' is say a block of flats without a physically > separate street, using addr:housename and addr:flats is often simpler. -- Robert Whittaker _______________________________________________ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb