[OSM-talk] waterway - "routable network" and reservoirs/lakes

Colin Smale colin.smale at xs4all.nl
Tue Jul 28 08:05:49 UTC 2015


If we can separate the flow direction discussion from the routing, the latter becomes a more generic "routing through areas"  problem which has been discussed before in the context of pedestrian routing. The idea being that it should be possible to construct a routing engine to take you from any point on an edge of a polygon to any other such point, while remaining within the polygon boundary.
An alternative is a full mesh network, where every possible entry node is directly connected to every possible exit node.
--colin

On 28 July 2015 08:01:51 CEST, Maarten Deen <mdeen at xs4all.nl> wrote:
>On 2015-07-27 23:39, Lester Caine wrote:
>> On 27/07/15 20:55, Mike Thompson wrote:
>>> I assumed that when the wiki spoke about "routable" it was referring
>
>>> to
>>> the water flow rather than boat/ship/barge traffic.   In any event,
>a
>>> routing engine for boats could use the presence of a dam or weir
>>> (combined with the absence of a lock) to deduce that ship navigation
>
>>> was
>>> not possible.
>> 
>> 'This way used should point in the direction of water flow' is only
>> applicable to non-tidal flows, and reservoirs may well control water
>> flow in a way that makes a 'water flow map' somewhat difficult to 
>> deduce.
>
>Only if they are entirely artificial. A dam in a river or stream makes 
>the direction of water very clear: high to low. Only when there is an 
>artificial reservoir with no natural tributary it is not clear.
>
>> The use of 'routable network' is rather ambiguous, but this is little
>> different to the problem of routing through other land based open
>areas
>> where several waterway features link into an area of open water. The
>> jury is still out on putting in all the paths through the area, but
>if
>> there is a navigable route designated through a water body it should
>be
>> drawn, but an imaginary link just showing water flow should not be
>> necessary? Any routing process should be able to deduce the relation,
>> there is no need to draw it.
>
>Causality. Does a water area need a way indicating the direction of 
>water? Of is it that when you draw a way through the water area it 
>should point in the direction of the water flow.
>
>Maarten
>
>
>
>
>_______________________________________________
>talk mailing list
>talk at openstreetmap.org
>https://lists.openstreetmap.org/listinfo/talk
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20150728/2c730e96/attachment.html>


More information about the talk mailing list