[OSM-talk] Locks

[David Earl]

Sorry to nag on this, but I now realise also that representing lock gate
direction by means of the direction of the (conventionally) two segments
attached to the node representing the gate doesn't work for the case of
locks at each end of canal summits. In that case one would have to add yet
another convention that the direction of flow changes at a node other than
the lock gate somewhere on the summit(*). In some cases that node would have
to be introduced specifically for that purpose.

Also, something else which doesn't work with this representation of lock
gates is that it can't distinguish between a staircase and locks separated
by pounds. Consider this:

   ---->--->--->--->---

Is this a staircase of three locks or two locks separated by a pound? Unless
I label the lock rather than the gates, I can't say.

I shall put a proposal for waterway=lock in proposed_features summarising
the arguments, then we can all decide.

David


* One might argue that the direction of flow should change at the point
where the feeder channel from the reservoir enters the canal, but that
doesn't really have anything to do with the lock gate, and in any case there
might be more than one feed on a long summit.