<br><br><div class="gmail_quote">On Fri, Apr 24, 2009 at 5:31 PM, Peter Childs <span dir="ltr"><<a href="mailto:pchilds@bcs.org">pchilds@bcs.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="border-left: 1px solid rgb(204, 204, 204); margin: 0pt 0pt 0pt 0.8ex; padding-left: 1ex;">
They don't look large, but they do looked stretched, ie the wrong<br>
shape. 100 Pixels East West is NOT the same distance as 100 pixels<br>
North South. It also makes the scale pointless as you need one for<br>
North-South and a different for East-West.<br>
<br>
I think we ought to be able to do a projection where we turn the globe<br>
into a symmetrical regular polyhedra. In effect at zoom level zoom<br>
level 1 the world is a tetrahedron, as you go up you add more<br>
equilateral triangles. while keeping your angles the same.<br>
<br>
At each higher zoom level each triangular faces is split into 3<br>
triangles. creating what gets closer and closer to a globe. Same way<br>
as some footballs are made up of hexagons......<br>
<br>
Just a thought.<br>
<font color="#888888"><br>
Peter.</font></blockquote><div><br><br>The spherical Mercator projection (no matter how badly it distorts surface area in lower zoom levels) is the de facto standard among online Slippymaps (Google Maps, Yahoo! Maps, Microsoft Live Maps, etc.). It's very easy to deploy since it facilitates the use of quadtiles, and its rectangular projection makes it simple to convert graphical coordinates to spherical coordinates. In addition, the Mercator projection has the nice quality that it preserves angles and shapes in higher zoom levels. So I wouldn't advise advocating a different projection for the OSM main map.<br>
</div></div>