[Tagging] Highway classification in Antarctica

[Frederik Ramm]

On 25/04/2024 16:16, Fernando Trebien wrote:
> I also think that such changes also imply corrections to the following
> section 

...

> "In a region with poor infrastructure, a road of highest importance,
> forming the main road network there, should be highway=trunk,
> regardless of being a high-quality wide asphalt road or a low-quality
> narrow track worse than highway=service in other regions."

Yes. I would agree with "trunk doesn't necessarily have to be a 
multi-lane road built to western motorway standards", but "trunk in one 
region might be worse than service in another region" is somewhat 
stretching the relativistic approach.

> What is missing is a statement about whether one starts judging such
> networks by importance from the top (trunk downwards) or from the
> bottom (tertiary upwards).

As an aside, are there even any routing engines which by default will 
use different standard speeds depending on country? Or do they all 
assume you can go 80 km/h on a trunk road even if that trunk road 
happens to be "worse than service in another region"? I know that at 
least GraphHopper and OSRM have the capability of using different speeds 
for different regions but AFAIK that isn't used by any of the default 
profiles. Should the OSM community compile a list of sensible 
assumptions of "this is how fast you'll go on average on type X highway 
in country Y"?

-- 
Frederik Ramm  ##  eMail frederik at remote.org  ##  N49°00'09" E008°23'33"