2009/6/4 Stephan Plepelits <sk...@xover.htu.tuwien.ac.at>: > Hi Folks! > > I'm the one who originally created the proposal. > > On Wed, Jun 03, 2009 at 10:52:33PM +0800, Eugene Alvin Villar wrote: >> Hmmm... I guess the main problem that people have is that a tag like >> "importance" (or its synonyms) is inherently subjective and a subjective >> tag is > Yeah, I see that point, but many things in the OSM are quite subjective (as > you say yourself, the level of a road, or is it a halt or a station? is > this a touristic spot or not?) > > At the moment renderers have the problem, that the information in OSM is > very huge, and maps look cluttered. The importance-key (which is more or > less derived from features on the ground) could help. Maybe we should give > more examples what you can use the key for: > > Place of Worships: > Cathedral amenity=place_of_worship importance=regional/national > Church amenity=place_of_worship importance=urban > Chapel amenity=place_of_worship importance=suburban >
You could have done church_type=cathedral, church_type=church, and church_type=chapel (arbitrary tag name choice... probably not a good one) and let the renderer figure out that for itself. Then my local chapel of huge international historic architectural significance (but zero religious significance) won't get promoted over the cathedral down the road which everyone else happens to be more interested in but is only of national religious significance. (scenario made up of course :-)) It's not so much about the subjectivity but about the lack of definition of what importance actually is about, even for objects of the same type. >> For what it's worth, I viewed the purpose of the importance tag in the >> context >> of a general-purpose map which OSM is by default (though it can and does >> support specialist maps like the Cycle Map). I guess such importance/ >> popularity/prominence data can conceivably be compiled as a separate database >> to be maintained by those who are interested in it instead of being included >> into OSM. > Brrr ... seperate databases ... I don't see a reason why we would need a > seperate database for this (I think this would be really hard to maintain > and nobody would use it). > Yeah, no need for separate databases, just need to be way more precise with the tags in the first place. A universal scale of importance just doesn't exist. Dave _______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk