On Fri, 17 Apr 2020 at 23:38, Martin Koppenhoefer <dieterdre...@gmail.com> wrote:
because it leads to key bloat. It makes evaluation harder or more > complicated if you have to cater for a lot of different keys which all > basically are about the same thing: the ref that an operator has assigned > to it > But the other way leads to namespace collisions which cause problems when searching for keys. Like it or not, ref:*=* is used for a lot of things, not just heritage. Too late to change. So do we want a place where searches for ref:foo are going to turn up false hits unless you can think of other appropriate search conditions to narrow it down or do we want heritage:ref:foo so we can avoid collisions? There are pros and cons both ways. heritage sites are already within the scheme, this was introduced with the > protected area proposal and class 22 already covers heritage protection > Some, but by no means all, POIs with heritage tags are protected areas. All of the heritage POIs I've mapped have been listed buildings or scheduled monuments: buildings, bridges, castles, megaliths, mottes, that sort of thing. Over 100 listed buildings in my town alone: https://britishlistedbuildings.co.uk/wales/cardigan-ceredigion (I think I've mapped them all). Over 250 scheduled monuments in my county: https://ancientmonuments.uk/wales/ceredigion (I've not mapped anywhere near all of them). Not a single one of them is a suitable candidate for boundary=protected_area. Yes, there are also protected areas in Wales. But not as many as there are listed buildings in my town. The fraction of heritage POIs which are protected areas is less than 1%. -- Paul
_______________________________________________ Tagging mailing list Tagging@openstreetmap.org https://lists.openstreetmap.org/listinfo/tagging
