[Tagging] Feature Proposal - RFC - value 'basic_network' for keys 'network:type', 'lcn' and 'lwn'

[Sebastian Gürtler]

Am 18.11.21 um 23:11 schrieb stevea:
> A component of this that feels "missing" to me is: what do the
> governments who put up the signs (or maybe they are not governments,
> but NGOs instead) call these things? Do they define a "network" by
> this signage? Do they define "a route" or "routes" or "a node-oriented
> network of routes?" Perhaps even something different from those which
> you characterize as "basic_network?" Is part of my / others / our
> problem that the creators of this "bicycle thing expressed with signs"
> is itself vaguely defined?

I started to translate my user page in the wiki where I made excerpts of
the guidelines for North Rhine-Westphalia (NRW).
(https://wiki.openstreetmap.org/wiki/User:Segubi/Elemente_NRW-Radwegenetz)
and had the plan to summarize the differences between the states
(https://wiki.openstreetmap.org/wiki/User:Segubi/%C3%96ffentliches_Radverkehrsnetz_in_Deutschland)
but the latter was too much work so far.

Shortly: There had been an initial proposal for bicycle signposting in
1998 by the German "society for research on streets and transport" (I
don't have that original). The implementation of this is in the
responsibility of the single states. I just know that in the preface of
the 2017 edition of the guidelines for signposting bicycle traffic in
NRW they state that it had been implemented now in all states of Germany.

Now to your questions: How do the call these things (I made a glossary
in German on my user page "Elemente NRW-Radwegenetz")? They call it
"Radverkehrsnetz" = bicycle network. They speak of a combined system of
"route orientated" and "destination orientated" guideposting. The speak
of "nodes" where routes can branch and there are destination signs
mandatory and route inserts below the destination signs possible. The
define certain rules for the consistency of the destinations (an
introduced destination has to be existent on all following destination
signs until you reach the destination and so on). They call the series
of ways between these "nodes" a route. They add the option for
characterizing the destinations (before a destination) with "destination
symbols" (train station, tourist information, museum...) and
characterizing the routes (after a destination) with "route symbols"
(leisure route = low grade way ;-), fast way, strong incline, former
train embankment).

They prescribe unique signs with only arrows and a bicycle logo without
any additional information for the guidance between nodes ("interim
signs"= route_marker).

They prescribed some rules for "theme routes" (osm: named routes), the
get a "route insert" below the destination signs. The route insert
mustn't show any information about the destinations (to be reusable in
both directions, to make sure to have exactly one unique route insert
for a given route).

The made up rules for the implementation of numbered nodes to make up
numbered node networks. There are "numbered node hats" for the posts and
numbers as route inserts.

They define the temporary possibility to keep old styled routes that can
share parts of the network but have a special sign for a "route branch"
without destination sign that "may be used if the routes don't fullfill
the rules yet".

I think that's it with the original terms. (And you see the problem:
there is no "basic_network", in principle you would have the whole
network with routes from node to node and these routes would possibly
make up theme routes or routes in a numbered node network... This
wouldn't be compatible with any of the previous schemes including the
actual scheme for the numbered node network. So the "basic_network" is a
compromise to include the routes that aren't covered otherwise.

Important: it is a compromise between the existent mapping of the
network, the feasibility, and the aim to have a possibility to gather
the complete information about the extent of the network and open it for
a simple retagging if other rules would be defined.

It makes it possible to define some rules how to interpret the whole
thing for rendering and routing. It makes no real problem if you keep
the old big relations and replace parts one after another by checked
routes, add new ones, or even add single routes without knowing the
further context. You can instruct a renderer to ignore rendering of the
"basic_network" if a new named route runs over it. (Here the are just
signposting a new named route by adding route inserts changing "basic"
sections into "parts of a named route", you just could keep the relation
and check by an analysis tool whether they still share the same ways).
Ideally you get a set of rules which allow a router only to follow the
official bicycle network. (Actually you need the assumption that the
named routes are always part of the network, which is not completely
true...)

Finally the tag "basic_network" is only a part of the whole story, but
can be a very useful beginning. And you can use it as an additional
feature to the former taggings which has the advantage that there only
will be additional information but no conflicts. It can be used for
first experiences and quite easily be modified.

Sebastian