[OSM-talk] Deprecation/move of incorrect tags

Alex Mauer hawke at hawkesnest.net
Sat Jul 14 02:11:33 BST 2007


Stephen Gower wrote:
>   This came offlist - was that your intention?

No, I just accidentally hit reply instead of reply to all.  Thanks for
catching that, I've added the list again.

> On Fri, Jul 13, 2007 at 02:40:06PM -0500, Alex Mauer wrote:
>   Ahh, I misunderstood.  When you wrote "My thought exactly. In my
>   opinion, the highway= tag should (and mostly does) imply automotive
>   traffic." I *assumed* that it was your objection, and it's one I,
>   as an English English speaker, disagree with (carriageway is the
>   term that implies automotive traffic).  Now reading between
>   the lines, I *think* we have yet another difference in what we
>   think the term highway means!
>   
>   Let me know if I've got this wrong, but you seem to have "Highway"
>   as synonymous to "Right Of Way", yes?
>   
>   For me, highway means more than just a right to travel along it. 
>   By being a highway, it means it will be built and maintained to a
>   certain standard - in the case of a Cycle Track, it will be of a
>   certain width, not have blind corners and other obstacles, and the
>   surface will not be too bumpy.
>   
>   Analogous to this is why we have (for example) Primary and
>   Secondary classification of roads - extending your proposals,
>   aren't these both just covered by "route=path; car=yes"?

This is a lot to respond to, but:

I understand highway to be synonymous with right of way, yes. If
something is a "highway" that says nothing about what standard it is
built and maintained to: how and to what extent that highway is built
and maintained is a function of what *kind* of highway it is. So yes,
primary and secondary would each be maintained to a certain standard.

Everyone I've talked to from the UK (other OSMers), as well as the
wikipedia page I linked to suggested that there is no such thing as a
"cycleway" classification of highway.  There are routes built for bikes,
there are lanes for bikes and they may be built to a standard; but it
would require little in the way of official procedures to reroute or
disallow cycle traffic on them.  Perhaps they were wrong and a "cycle
track" has the same legal status as a bridleway (only substituting
bicycle for horse

When I said that "highway" implies automotive traffic, I meant that in
the context of OSM: the tag is very much oriented towards motor vehicle
traffic.  My thought and/or intent is to make that even more the case; I
see a strong need to separate out "routes intended to be used by cars"
from other routes.  "route=path" would provide a general tag for any
non-automotive routes.  (so route=path; car=yes would be very unusual,
if not invalid)  Bicycle paths, ski trails, bridle ways, hiking trails,
snowmobile trails, etc. have much more in common with each other than
they have with roads designed and built for automotive traffic.  Someone
driving a car is likely to have little interest in a bridleway going the
way he needs to go.  Someone on a bicycle, horse, or skis is likely to
want to avoid heavily-trafficked automobile highways. (assuming it's not
actually illegal for them to use them)

>> Anyway, this need is *not* met in the existing tagging guidelines, in
>> that there is no way to define something that is not a highway, but is
>> an equally-shared use route for bicycles and foot (and horses as well)
> 
>   Don't get me wrong, I'm not objecting to a new tag for something
>   that is not one thing or the other, I'm objecting to the
>   depreciation of useful, working, existing tags.

Understood.  The reason I'm trying to deprecate those tags is because
their use is very unclear to me, and I do not see a benefit in having so
many ways to tag a route that mean so nearly the same thing.  That is,
as far as I can see, the difference between highway=footway, horse=yes;
and highway=bridleway horse=yes (and route=path, horse=yes,foot=yes) is,
as Ian Sergeant put it, "some esoteric aspects of English law"

Now, perhaps there is a need to show "this route is built to the
standards of the english legal term 'bridleway'" (or cycle track or
footway)... but I don't see that need, and I certainly don't have that
need (being in the US). At the very least, it doesn't seem to fit very
well with the otherwise international spirit of OSM, and
highway=bridleway (etc) shouldn't be used outside of the UK.

And if they shouldn't be used outside of the UK, that should be
clarified in the documentation (wiki) and ideally those tags should be
moved to something like uk:highway=bridleway to make it even more clear
that it's UK specific.

However, I think it's better to at least make the basic set of tags
described in map features be as country-neutral as possible (and thus
deprecate highway=bridleway/cycleway/footway since they have no meaning
 outside of the UK.)

>   route=path certainly has some merit, and after appropriate
>   discussion, I'll probably support it. Personally, I'd prefer
>   highway=path, rather than creating Yet Another Key, even though
>   this isn't a highway - it's only a word after all, and as long as
>   the description said "A path which is neither an offical cycle
>   track or an official footpath" or similar, I think people would
>   mostly do the right thing.

Please note that "route" is not a new key; "path" is a new value for the
existing key.

>   
>   I do echo what another poster said - thanks for all the work you're
>   doing to tidy up the tagging guidelines, it's much appreciated.

Thanks, I'm glad it is.


-Alex Mauer "hawke"

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 252 bytes
Desc: OpenPGP digital signature
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20070713/47bd19d2/attachment.pgp>


More information about the talk mailing list