[Talk-GB] addr:place cleanup process

[Colin Smale]

Another solution is to align with the existing practice outside of OSM, i.e. adopt the PAF data model. "Street" is abstracted to "thoroughfare", and "substreet" becomes "dependent thoroughfare". Post town is post town. In between you have "locality" and "dependent locality" (and even "double-dependent locality" if required). Note the use of neutral terms, which indicate the hierarchy clearly while avoiding discussions about "is it a street" or "shall we call it a hamlet or a suburb". We are agreed (I believe) that the addr:* tags are for the "Postal Address", i.e. not intended to indicate or convey (exact) location or identify a building. We have also agreed that we are talking about a UK-specific model, and not a universal canonical world-wide model for any kind of addressing. So why the urge to reuse existing tags like addr:suburb? That's not reuse, that's overloading (meaning can vary according to the context) and just creates confusion.

What we need is a tagging scheme that is easily explained, and minimises the risk of misinterpretation (either by mappers or consumers). Misappropriating tags in this way is asking for trouble.

Aligning with the PAF data model means:
* the model is guaranteed to work
* it's already documented
* there exists a source of absolute truth for any arbitration required


> On 01/15/2022 1:25 PM Mark Goodge <mark at good-stuff.co.uk> wrote:
> 
>  
> On 14/01/2022 19:07, Rob Nickerson wrote:
> > 
> > If you have an address which has some settlement name before the postal 
> > town then there are only ever two tags that you need: addr:suburb and 
> > addr:city. The case we haven't yet worked out how to handle is when the 
> > address includes 3 settlements / settlement sub areas ("locality 
> > elements" in RM language). At that point we use addr:suburb for the 
> > smallest, addr:city for the largest and an as yet undefined tag for the 
> > middle.
> > 
> > So "suburb" in addr:suburb should also not be interpreted based on our 
> > view of what a suburb is or isn't.
> 
> The problem with this is that it goes against the normal meaning of the 
> word "suburb" in British English. Not every OSM editor is reading this 
> list, and not every OSM editor is going to read the wiki. So not every 
> editor is going to realise that "suburb" means something different in 
> OSM to what it does in everyday use. Which means that people will, 
> whether you want them to or not, use addr:hamlet, addr:village, 
> addr:town or whatever seems most appropriate, because they simply don't 
> know that OSM addr:* tags don't follow normal English usage, and, 
> without reading either the wiki or this list, have no way of knowing.
> 
> It seems to me there are really only three possible solutions to this:
> 
> 1. Find some way to prevent people adding or editing addr:* tags until 
> they have shown they have read and understood the wiki.
> 
> 2. Have an ongoing project of repeatedly correcting the wrong use of 
> addr:* tags.
> 
> 3. Stop caring about it, and accept that, in the UK, addr:* tags will 
> be, and can be, duck-tagged.
> 
> The first may seem attractive (and would result in a much cleaner 
> dataset), but I really don't see how it could be enforced. The second is 
> a lot of work, and is likely to result in edit wars where people think 
> that their edits are being wrongly changed by people who don't 
> understand the situation on the ground ("No, Kidsgrove isn't a suburb of 
> Stoke-on-Trent, it's a town in its own right"). So it seems to me that 
> only the last is a practical choice. It may upset the purists, but if we 
> genuinely want to encourage more OSM users to become OSM editors then we 
> need to accept that commonly used tags have to follow normal English 
> usage rather than requiring people to adopt non-standard and 
> non-intuitive terminology.
> 
> Mark
> 
> _______________________________________________
> Talk-GB mailing list
> Talk-GB at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-gb