<!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="#000066">
I came across a problem with Velomap, which I find is a good
autorouting version of OSM for my purposes for cycling and still
works fine for walking too.<br>
<br>
Velomap has made an assessment of the various tags to guess how best
to route for bicycles on roads, tracks and paths. Being an
international model there is a point where the definitions don't
seem to work for the needs of the map. Most countries have a
definition of trunk which can be paraphrased as "nearly motorway"
whereas the UK has adopted a stance of "if it gets a green sign, it
is a trunk road'. Velomap determines that highway=trunk means not
suitable for cycling.<br>
<br>
I came across the A511 which goes through a little village and the
detour is some 9 miles and there are clearly many roads like the
A441 around Redditch where they are not true national routes, though
they do link major(ish) centres which are acceptable cycling routes.
It does of course mean that OSM matches commercial maps in
displaying these routes and the decision to go with this tagging is
well embedded in OSM.<br>
<br>
This can be resolved with a bicycle=yes tag.<br>
<br>
So my question is: should we be encouraging the tagging of bicycle
accessible highway=trunk roads explicitly? My instinct is that all
roads are accessible to bikes aside from M roads and those sections
that are explicitly signposted as such. So the other workaround
would be to produce a script to tag all UK trunk roads as
bicycle=yes where not already tagged and then explicitly set the tag
to bicycle=no where it is restricted.<br>
<br>
I recognise that Velomap is just one guy's interpretation, however,
it is a reasonable interpretation from an international tagging
perspective.<br>
<br>
Spenny<br>
<br>
<br>
</body>
</html>