Is it possible to use addr:locality for both towns and villages? That could simplify things quite a bit and I have yet to see an address that needs a post town and two levels of localities below.
Having said that, I still don't understand the objections to addr:town and addr:village. Can anyone come up with an example of an address where they wouldn't work? I normally don't care about names but locality sounds almost offensive. Business parks and other campuses are not localities - their names are written before street names, not after them. They're IMO what RM calls "dependent thoroughfares". For these I would simply use addr:place, which can already be combined with addr:housename and addr:housenumber. Alternatively we could make a new tag like addr:campus. Best regards, Andrzej On 28 January 2019 20:36:24 GMT+08:00, Colin Smale <colin.sm...@xs4all.nl> wrote: >Hi Will, > >On 2019-01-28 13:19, Will Phillips wrote: > >> Hi, >> >> I agree we need another tag below addr:city for localities. For this >I have usually used addr:suburb when mapping in urban areas and >addr:locality elsewhere. Ideally I think it would be best to have just >one recommended tag, perhaps addr:locality, because having addr:town >addr:village and addr:suburb seems too complicated. Eventually it would >be good if editing software, in particular iD, could provide an extra >field to enter the locality, and it would perhaps be easier for that to >happen if there was only one tag. New mappers often seem to have >difficulty entering addresses to the form that they wish and I think >the lack of a locality field is part of the reason. >> >> For what Royal Mail calls 'Double Dependent Localities' using >addr:sublocality is a possibility, although I wonder whether just >sticking with addr:village for this less common situation would be >easier. It depends a bit on whether this tag is only likely to be used >for villages and hamlets, or whether it might be useful in other cases. >For example, sometimes names of industrial estates appear in addresses >in a similar way to sublocalities. > >I don't see any advantage in "addr:village" and "addr:suburb" just >because they sound familiar or are existing tags. What we are >discussing >here is a UK-specific solution. The (Double) Dependent Localities may >or >may not correspond to what people perceive as a "village" or "suburb". >In the quoted example, "Cambridge Science Park" is IMHO neither. > >> I only use addr:city for post towns, although I recognise not all >mappers agree with this, and I appreciate there are arguments both >ways. I was thinking about this recently when adding addresses in Lees >near Derby. The post town is Ashbourne, but this seems slightly >incongruous because the village is much nearer to Derby. I chose not to >include addr:city and only used addr:locality for the village name. > >> I feel the main argument in favour of using post towns for addr:city >is that it helps to keep the data consistent because what to use often >becomes confusing otherwise. To use the example of Lees I mentioned >above, it would be easy to end up with a situation where addr:city >contained perhaps four values if the data was entered by different >people without any guide as to what to use (the most likely >possibilities being Lees, Dalby Lees, Derby or Ashbourne). > >> In cases where local residents consider Royal Mail's choice of post >town to be contentious, usually because it is miles from where they >live, it might be sensible to recognise addr:posttown as an >alternative. > >The accepted paradigm is that the address should represent the postal >address, and not any administrative relationships. As you will know RM >have their own particular ideas of the geography of the UK, all done >for >their own convenience. It would certainly avoid some confusion if we >used addr:posttown instead of addr:city. > >Regards, >Colin
_______________________________________________ Talk-GB mailing list Talk-GB@openstreetmap.org https://lists.openstreetmap.org/listinfo/talk-gb
