[OSM-talk] Proliferation of path vs. footway

[Lauri Kytömaa]

Shaun McDonald wrote:
>> As fine as it as a guideline, verifiability as a topic and was
> Even so the on the ground rule and verifiability have not been on the wiki 
> for long. They have been the unwritten norms of the community since the

I'm all for referring to that verifiability where it comes to legal and 
physical attributes (e.g. access=yes/designated/no or building:levels or 
width), yet trying to squeeze by force the old tags to comply - to the 
letter - to that norm seems counter productive. As with car access on 
very rough tracks: there could be tens of tags to describe the ground 
clearance, wheel size and suspension travel etc. required to get through, 
but instisting people start measuring them is too much work that anyone 
else would start doing so - users require something simplified from that - 
even if there's no widely accepted solution yet.

The description of a way for other users than cars varies on multiple axes 
and fitting all that into one tag seems impossible; yet it's most of the 
time reasonable and simple to divide the decision space into two sets: 
cycleway and footway and use additional tags from there on. Some ways then 
are borderline cases or sufficiently outside of those two sets that they 
necessitate some other solution.

Most of the time the intended use is unambiguous, either signedposted or 
evident from the location or structure. Where it's not, I trust people can 
classify things on a closed scale, even if with some personal judgement 
And to make those cases easier, there is a need for something in addition 
to the footway/cycleway pair.

(Where does a coniferous forest turn into a mixed forest? One birch? One 
birch for every ten pinetrees? 25:75 distribution?)

Much of the discussion would have been avoided if the documentation of 
footway and cycleway had been more exact already in the fall 2007 - it 
took me then quite a lot of reading to get to the logic and implications 
behind them, and many don't read that much of the scattered documentation 
which has lead to some of the pages having been changed around and 
misconceptions.

The big question is just how can that be fixed?

Alv