<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta content="text/html; charset=ISO-8859-1"
http-equiv="Content-Type">
</head>
<body bgcolor="#ffffff" text="#000000">
On 09/03/2011 07:30 AM, sergio sevillano wrote:<br>
<blockquote
cite="mid:2A357BB1-B1D5-408D-8624-6931825808FD@gmail.com"
type="cite">
<pre wrap="">are we mapping reality or "for the router" ?
</pre>
</blockquote>
The question is at what resolution are we mapping?<br>
<br>
First pass, the roads connect.<br>
Second level of detail, the intersection is stoplight controlled.<br>
Next level of detail: it has two left turn lanes.<br>
Next level of detail: it has four stoplight poles.<br>
Next level: the 2nd and 3rd pole have three heads each, two of which
have left arrow balls and a bicycle phase ball.<br>
After that: a relation to the next signal up for signal
synchronization with a design speed of 35mph between them.<br>
What's next then? This light is green now with 4 seconds left on the
red left arrow phase?<br>
<br>
<br>
Personally I choose to concentrate on blank areas of the map, and
features that are more stable (the configuration of poles may change
over time, but the intersection will likely remain stop light
controlled indefinitely). osm does not restrict anyone from mapping
to any level of granularity.<br>
<br>
-----<br>
We'd should indeed map to the needs of routers.<br>
But that router-aware mapping can reflect reality at a variety of
levels of physical and temporal detail.<br>
<br>
I'd argue that if you're going to place stoplights at their physical
location, the<i> node should still be tagged as a stoplight.</i><br>
<br>
</body>
</html>