Having an automatic algorithm that figures all this out is a nice dream, and we should work towards it, but I don't think it will ever be perfect. A map-maker uses a lot of information to decide which places to show on a given map. Some of it is available to a renderer, and some isn't. A "relative importance" tag that can have +2, +1, -1 etc to modify the otherwise automatically calculated values for the object may turn out to be required. A renderer can ignore the tag if the programmer is sure they have figured out a way to do it that does not require the information.
Also - for those trying to figure out an automatic method - In Australia, we have a different problem to Europe and the US. We want places that wouldn't normally show at higher zooms to do so. Otherwise we end up with huge areas of blank map, that actually have locations in them that are smaller than would normally show in that zoom, but are important to the local areas. Any rule that looks at just the population or other size factors will fail in this case. The further you are from any other displaying object, the more important smaller objects become. If you are a hamlet with a petrol station, a general store and 5 houses, but there is nothing else on the road for 200km (or even 100) in any direction, then you should show on much higher zooms than a similar place in the middle of a group of towns. It should be possible for some sort of distance algorithm to be used to bump the importance of such places, but I suspect this would slow rendering a lot. Stephen _______________________________________________ talk mailing list talk@openstreetmap.org http://lists.openstreetmap.org/listinfo/talk
