[Talk-transit] catching up

Peter Miller peter.miller at itoworld.com
Sat Jan 10 20:55:12 GMT 2009

Was this meant to be on the list? I think it probably should be and  
have copied my reply to the list:

On 10 Jan 2009, at 20:37, OJ W wrote:

> I mean, 'unambiguous' might only be (automatically) decided by
> something capable of doing bus-routing.  any program which only knows
> the route-points and the OSM data might incorrectly use footpaths or
> power lines or roads with maxwidth<2.5 or psv != yes to connect the
> points?
> And any program without the OSM data or without A* capability might be
> completely unable to figure-out where the bus would go.

It might well be sensible for OSM to do routes that include 'way'  in  
order for the reason you give.

As an alternative we could use route-point/waypoint model  but include  
so may 'route points' that it is very simple to see what infrastrcture  
is used, indeed Google Transit does this using their 'tracks' which  
are almost GPS trails through the networrk.



More information about the Talk-transit mailing list