(Multiple answers) First, I'm aware that full classification of every populated places is not possible world wide with one tag only, the proposition states that clearly. administrative, interest, tourism, local concerns need to be recorded with other tags. The problem I see with actual place usage is that it is not standardazided world wide and serves merly for writing a label on a map.
It makes it quite hard for newcomers to guess in wich case they should use wich place, and that proposal tries to help have a first easy step to record at least a population estimation. On jeudi 27 mai 2010, Simone Saviolo wrote: > -1, if it's exclusively population-based. The risk is that the US have > tenths of cities and smaller countries - say, dunno, Uganda - get > none. Well, this is the truth ;-) based on such a scale. I don't see any problem in uganda having no megacity, because it's what is. On jeudi 27 mai 2010, Roy Wallace wrote: > I like your motivation. But maybe it's not necessary. Using > population=* achieves the same goal. Yes it does, and it does much more precisely, this is the utimate solution. Unfortunetly, having access to this information is much harder when you are driving your car thru, than a rough estimate that gives you the "approximate size of a hamlet" (I have to admit that the upper part of the scale is kind of useless as population data is much easier to get in those cases) On vendredi 28 mai 2010, John Smith wrote: > +1 I doubt you could ever standardise cities, especially not based on > population alone. That proposition is not about standarisation of cities around the world, it's about standardisation of the place tag usage in the osm database. I don't see any problem if a place=city is actually a town in the local language usage, (let's add a common_denomination=town) just like in france we dont care that our "villes" are tagged place=town. I'm concerned about database storage and meaning of the data, and key values could just have been numbers, but it's harder to remember. On vendredi 28 mai 2010, Simone Saviolo wrote: > consequence, while it is good to see the map of France showing five > cities, Italy's map should instead contain a few tenths. Of course, > this cannot be achieved if we only look at population - especially if > we want to use a world-wide population criterion. > > I hope the idea is clear. Unfortunetly, I think I get it, you have invented/searched perfect Italian thresholds in order to make the map...@osm.org map look what you are used to see. Maybe you endeded to 47000 people max for a city In order that city X of 46999 isn't shown, because it's administrative role is not enough to apear at zoom 10 ? Wich, in other words is a way to say you addapted the tagging of places to the actual mapnik style sheet at osm.org. Is osmarender rendering the way you want ? maybe no... Will it survive the next map...@osm.org stylesheet update ? maybe no... I don't blame that need, of course, map...@osm.org is the perfect portal of the projet, and it needs to bring a decent map that every country is happy with, but that's tagging for the/a renderer. What I'm looking for is go thru a first step where maps won't be as perfect has we want them to look (population based representation of population) before we consider the need of other tags in order to render correctly. -- sly Sylvain Letuffe sylv...@letuffe.org qui suis-je : http://slyserv.dyndns.org _______________________________________________ Tagging mailing list Tagging@openstreetmap.org http://lists.openstreetmap.org/listinfo/tagging
