Sounds like a good Fermi problem to me: http://en.wikipedia.org/wiki/Fermi_problem (e.g. how many piano tuners are there in Los Angeles, or how many atoms of Jesus did you eat today?)
I think the measurable variables are: - graphics device rendering efficiency (this is probably why it's a power of two, e.g. for opengl textures) - display size (dictating number of tiles on screen at any moment) - bandwidth (how fast can you download tiles) - http download overhead (what's the cost per tile, apart from the data itself) - number of simultaneous downloads (e.g. what does randomly picking a subdomain do to help) - file format (e.g. satellite images are usually jpeg, road maps are usually png) Then there are subjective qualities: - difference between deep zoom (which loads the large scale tiles first and enlarges them, then fades in the current zoom level) and google maps which loads the current zoom level only - how quickly data is on screen when interacting - I suspect this is why some mobile tile providers use 64px squares Data complexity will also affect the result, since the compression of the images will vary from place to place. The answer to your question is a complex combination of the above factors, and others I've no doubt forgotten or am not aware of. But a quick back of the envelope with some good guesses should give you roughly the right answer. Tom. 2009/6/10 Pablo Rodríguez Madroño <[email protected]>: > Hi, folks > > I've been following you for several weeks, and this is my first post. > So, hello everyone! > > I was wondering if you have any experience with tile sizing for web > mapping. I.e. how can you calculate the size in pixels of each tile? > Why is it 256x256 in Google Maps, for instance? > > Any pointer will be greatly appreciated... and thanks in advance. > > -- > Pablo Rodríguez Madroño > [email protected] > http://www.pablorodriguez.info > http://www.publictransports.info > > _______________________________________________ > Geowanking mailing list > [email protected] > http://geowanking.org/mailman/listinfo/geowanking_geowanking.org > _______________________________________________ Geowanking mailing list [email protected] http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
