[OSM-talk] Cycle route won't render

Steve Bennett stevagewp at gmail.com
Sun Jan 17 10:32:10 GMT 2010


(Just some musings, not sure I have much to contribute to this thread)

On Sun, Jan 17, 2010 at 2:31 AM, Felix Hartmann
<extremecarver at googlemail.com> wrote:
> Well I think route=mtb is more or less a consensus. Differentiation to
> route=bicycle is in 99% of all routes clear. It's more on the

Good!

> route=bicycle side where "they" have to think about differentiations on
> trekking routes, commuter routes (routes for tourists vs commuters have
> completely different objectives),

Do they? I can think of some rail trails in my area that are used by
bike tourists, recreational day rides, and commuters. But anyway,
that's not really the topic.

> What we have to think about are inofficial routes. OSM would really
> profit if nice and popular Transalp routes are included into the
> database (we are free mtbikers, we don't need no stinking signs to tell
> us that we are using a mtb route). However I do think we could should
> have some differentiation between mtb routes that are "flagged" out (be
> it only in official tourist brochures, be it on signs,...) and mtb
> routes that are not made by official authorities but by users (e.g. my
> friday afternoon workout route, my favourite transalp route).

Yeeees...to a point. OSM has primarily focused on mapping stuff that
everyone agrees is there: signposted routes, physical roads etc. But
whereas Google Maps has a nicely integrated "My Maps" feature, there
isn't a single, central way to incorporate user data, or information
that is subjective or only interesting to a small community. If you
were going to incorporate routes that were invented by a club or one
person, but that weren't signposted, you'd want to do it in a way that
wouldn't pollute the main database. So people could download just the
normal database, or database+mtb routes.

Personally, I would love the millions of "upload your GPS trace" sites
to coalesce into a smaller number of richer resources that tied in
well with OSM. It's very frustrating that traces that people have
uploaded are so dispersed. It's also annoying that there's very little
network effect: most of the sites seem to treat every trace as totally
independent, and never think about ways to link them up.



> Then we need a differentiation between oneday roundtrip routes (local)
> from A to B to A, and routes spanning multiple days (usually regional)
> from A to B to C.

Do we? For starters, that seems pretty arbitrary: there are routes
that some people would do in one day, that others would do in several.
And can't you tell from the distance anyway?

> Just to answer how many regional routes I know, I do
> know a lot of regional and even official multiple days routes in the
> European Alps.

Cool. Now that I think about it, I researched and nearly did such a
route starting from Innsbruck, but I had too much gear in the end and
it wasn't feasible.

> And there is no problem of all if several routes use the same way, I
> think by now it has become clear to most that routes will have to be
> using relations, because otherwise we get into trouble with multiple
> values for unique keys.
> I think the network key would be good to be used for differentiation of
> the different routes. We simply have to think about unified tagging so
> that renderers know what kind of route they are analyzing.

Nod.

Steve




More information about the talk mailing list