On Thu, Jan 30, 2020 at 10:09 AM Kevin Kenny <kevin.b.ke...@gmail.com> wrote: >> (1) The `name=*` field gets the road's actual name. If the road's only >> name is 'County Route 12' (New York consistently uses 'Route' rather >> than 'Road' for these), to the extent that `addr:street=*` will show >> that for the name, then `name=*` gets that name. (Yes I know that >> there are mappers who would prefer `noname=yes` in that situation, but >> address validation has an easier time with the way I do it.)
On Thu, Jan 30, 2020 at 11:20 AM Paul Johnson <ba...@ursamundi.org> wrote: > Please stop. This gets very annoying for data consumers (especially your > average joe just using a satnav). Fix your validation process instead. I knew you were going to say that. The sentiment seems to run about equally between 'fix the navigation software not to read the ref twice,' and 'fix software that recognizes street addresses to deal with the fact that an address of '2367 County Route 34' might need to be translated to a ref=*'. Either one needs a modicum of natural-language processing to recognize that the street 'name' is actually a reference number - it's a question of whether data consumers that read 'addr:street' or ones that read 'name' have to do it. With the current state of the art, it seems safer to have redundant information - the worst case is the annoyance of "Turn right on County Route 34, County Route 34", rather than "I can't find the street address , 2367 County Route 34". In my mind, for this case, pragmatism trumps Platonism. I know that "a reference number is not a name," but don't have a better answer for "how do I deal with street addresses that use a reference number as a surrogate for a name on a road that's otherwise nameless?" We're agreed that if the street has a name, then the reference number should not be 'name_1' or 'alt_name' or anything like that - and I routinely delete those whenever I encounter them. `name="Balltown Road" ref="NY 146"` is correct for that case. (A possible exception is if E911 or the Postal Service insists on the ref as a name - but that's rare on a road that actually is named.) -- 73 de ke9tv/2, Kevin _______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
