[OSM-talk] A new highway tagging scheme - thinking about

[Peter Miller]

Thanks Andy, sounds good; The general message seems to be that lcn is used
as the tag or tag prefix for information to do with the local cycle network.
Similarly with rcn and ncn. I fully agree that route=ncn should be
deprecated. Using a combination of lcn_ref and lcn it will be possible to
construct a model that can be used for rendering and also potentially for
routing.

I agree that in addition to the lcn=yes tag there is the potential to have
other values in the future to indicate the quality of the route, 'no',
'poor' and 'excellent' might be useful additions for routing. However, for
lanes and tracks we already have tags 'cycleway=track' and 'cycleway=lane'
tags for physical infrastructure so I don't think that information belongs
in lcn. These tags do get a little fussy where there are cycle lanes on both
sides of the road 'cycleway=lane;opposite_lane' but it is fundamentally
fine. I have tagged some cycle lanes/tracks this way in Ipswich and it would
be good to know if you will be rendering from it, I suggest you do this by
changing the colour of the appropriate road edge (blue is conventional) :)
If you do render from it then I will complete the tagging.

To be clear though, just because a road has a lane or track does not make it
'recommended for cyclists' and part of the local cycle network. It might
just make it a little less dangerous so it is really important to keep tags
about lanes/tracks separate from tags about the local cycle network just as
the same is true for signed routes which again don't necessarily imply a
good route.



Regards,



Peter



> -----Original Message-----
> From: Andy Allan [mailto:gravitystorm at gmail.com]
> Sent: 28 August 2007 09:49
> To: Peter Miller
> Cc: talk at openstreetmap.org
> Subject: Re: [OSM-talk] A new highway tagging scheme - thinking about
> 
> On 8/28/07, Peter Miller <peter.miller at itoworld.com> wrote:
> 
> > We also have tags for signed routes (lcn_ref, rcn_ref, ncn_ref) which
> are
> > fine for official signed routes. Andy from GravityStorm suggests using
> > 'route=lcn' for non-signed cycle routes which is a good concept and
> might be
> > sufficient but currently muddles the cycle network in with 'routes' for
> > buses and pub crawls.
> > http://www.gravitystorm.co.uk/shine/cycle-info/
> 
> I've actually been considering trying to deprecate route=lcn etc,
> simply because it's a muddle and hard to deal with. I was thinking
> about using another tag scheme, namely lcn=* (with corresponding rcn
> and ncn, of course)
> 
> lcn = yes - a local cycle network
> lcn = proposed - a proposed route. Gives me a mechanism to sort out
> http://www.gravitystorm.co.uk/osm/?lat=6803040.44631&lon=-
> 221228.46028&zoom=12&layers=B00
> lcn = *some other thing* - gives us flexibility to tag other things -
> perhaps lcn=onpavement, lcn=onroad, lcn=buslane or who knows what. I
> would assume any value to indicate yes (c.f. flexibility), unless
> known otherwise (e.g. 'proposed' or 'no')
> 
> lcn_ref would still be the numerical reference for a given cycle route.
> 
> Basically it makes it much easier to tag, since there's no
> semicolon-delinated-'route'-tag stuff going on. I'm interested in
> considered comments regarding this.
> 
> Cheers,
> Andy