[OSM-talk] definition of the main highway-tag

[Martin Koppenhoefer]

2009/7/31 Dermot McNally <dermotm at gmail.com>:
> 2009/7/31 Greg Troxel <gdt at ir.bbn.com>:
> Well, you could argue that it would be valid to adopt this standard in
> a country where it was deemed useful. But that's not how it is here.
> Ireland has two grades of National road, primary and secondary
> (corresponding fairly well to how the UK has two types of A road).
> Like the UK, we use trunk and primary to differentiate between the two
> (trunk for primary, primary for secondary, and yes, I know this isn't
> how you'd ideally design the terminology...). But primary and
> secondary are measures of the route's significance, not of the actual
> build standard, which can vary widely.

This is exactly my point. The highway class already represents the
importance of the road, not it's physical build standard, but the wiki
defines the latter to be relevant. I was suggesting to update the
definition according to best-practise, not to change the meaning of
existing tagging.

If the administrative class in your country coincides with the
importance: fine. Nothing changes. Unfortunately this is neither in
Italy nor in Germany the case: some roads have been downgraded /
passed to a lower maintenance entity for administrative reasons (now
somebody else pays and cares for the maintenance, what was before a
nation / federal road has sometimes become a regional / Landstraße).
Others, like Kreisstraßen in Germany (comunal roads) have been
upgraded and are now almost Autobahn-Standard.

As result of this, it has been agreed not to corelate administrative
status and highway-class. But there is a problem with tagging pure
physical state as well: it depends on the context. In a rural area a
secondary or primary street will be much smaller than in a highly
dense urban area. This is why importance of the road seems most useful
(be it for routers or to structure visually and according to
significance on rendered maps).

> And going with your suggestion would leave us without a useful
> differentiation between the primary and secondary national roads. What
> we have works, and build quality can be inferred by other means.

again: build quality is what the wiki _already defines for 2 years
now_, it is not what seems reasonable or actual practise, that's why I
started this thread.

cheers,
Martin