El Domingo, 19 de Abril de 2009, Greg Troxel escribió: > There has been some discussion about what might be a similar problem on > talk-us, where different states have different signs on Interstate > highways. [...] > In the US, the essence of the issue is that we have Interstate highways, > which have a standard sign, but some states, especially California, have > variants, and people want maps to show the local variants so they match > what's on the ground.
Here in Spain (where this proposal comes from) we are in a very similar situation. The national scheme is quite clear, but lately, some regions* are starting to colour the road sings their own way just to "stand out". * Technically, "autonomous communities" > The proposal is to tag the road as being an interstate in California so the > renderer can find the interstate/california sign variant. The main problem I see with this approach is the complexity of the rendering rules. There will be soooo many exceptions to the general rules that modifying the renderers will prove to be a daunting task. > Using ref:color means that has to match ref. So there's the question of > when a physical stretch of road has two route numbers. (This happens > often in the US and I would think it must be common in Europe as it's > hard to avoid without having people not to be able to follow the lesser > road.) That's hard to find in europe AFAIK. There are the E-* signs, of course, but we have int_ref for those. Here in Spain, any big road sign will tell you the ref of the road you're in (top-center) *and* the refs of the roads it connects to (under the ref of the current road). > It seems for that one needs to move to relations with no ref > tags on ways, but only on the relation, and then the way can be in two > route relations. +1. > I guess my biggest question is if the colors are arbitrary or they are > encoding some property of the road. If they are telling people > something about the road, then perhaps those rules are what should be > encoded and then renderers should have access to the > (jurisdiction,property)=>color table. The color encodes jurisdiction and administrative level. The problems I see are: - Sometimes, there are more real administrative levels than OSM administrative levels (i.e. yellow vs. purple tertiaries in Soria). - That table can be a *real* mess. Cheers, -- ---------------------------------- Iván Sánchez Ortega <i...@sanchezortega.es> Proudly running Debian Linux with 2.6.29-1-amd64 kernel, KDE 3.5.10, and PHP 5.2.9-1 generating this signature. Uptime: 16:33:40 up 2 days, 18:16, 4 users, load average: 0.57, 0.60, 0.54
signature.asc
Description: This is a digitally signed message part.
_______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
