[Talk-GB] Missing place=city nodes: Manchester, Leeds

Richard Fairhurst richard at systemed.net
Sun Apr 28 11:52:32 UTC 2013

Philip Barnes wrote:
> Leeds, however it is in the middle of a pedestrianised area, so 
> makes routing results at best unhelpful as it directs you to a 
> dead end side street. Would be better on a main road from 
> which the central car parks are accessible.
> http://osrm.at/33S

Agreed that routing is a really important use for place nodes.

Many settlements have an obvious 'centre point'. In Oxford it's Carfax, in
Cambridge it's the Market Place / Great St Mary's, in Gloucester it's the
main crossroads, in Leicester it's the Clock Tower, in Peterborough it's the
square outside the Cathedral, here in little old Charlbury it's the
crossroads by the Rose & Crown, and so on. London's routing centre is
famously Charing Cross. A place node is a good way of mapping this.

But when a place node is arbitrarily situated, or isn't present, routing
instructions tend to start/finish at an obscure alley somewhere. I've
recently moved several place nodes to these centre points for exactly this

I'd slightly modify Philip's advice, though. I don't think accessibility of
car parks should be a consideration. OSM data is used for all modes of
transport and we shouldn't make value judgments that favour motor traffic.
Better to choose the established/historic centre point - i.e. a factual
approach rather than simply "tagging for the router". :)


View this message in context: http://gis.19327.n5.nabble.com/Missing-place-city-nodes-Manchester-Leeds-tp5758812p5758845.html
Sent from the Great Britain mailing list archive at Nabble.com.

More information about the Talk-GB mailing list