[OSM-talk] Highway=trunk : harmonization between countries ?

Greg Troxel gdt at lexort.com
Tue Aug 22 23:52:16 UTC 2017

Lester Caine <lester at lsces.co.uk> writes:

> http://map.project-osrm.org/?z=11&center=52.149501%2C-1.577225&loc=52.048797%2C-1.856918&loc=52.258912%2C-1.619625&hl=en&alt=0
> 43.4km 33min
> Dragged to normal route via Mickleton it becomes 32.3km 34min, but
> anybody who uses this section of the A46 will tell you that it's often
> stationary traffic even on the dual bits ... so the quoted times are
> simply wrong a lot of the time.
> To add to the fun Warwickshire has decided they will enforce a county
> wide 50MPH maximum overriding the national limit so some of the long
> straight sections we 'should' be slowing down while some sections of the
> A46 are 70, but the same 50 limit applies to other long stretches on the
> A46 as well ;)
> I have a smaller problem heading south to the M5 and again 'shortest'
> picks the quick route but then avoids the M5 as well :) Interesting ...
> checking OSRM is getting it right so I need to check OSMAND again on that.

If I drag the start point NE just a bit it flips to the route you like.

I don't have enough Copious Spare Time, but I'd like to see a way for
routers to export their estimates of time/distance per leg, and to match
that up with GPX, and essentially find legs with bad estimates, so they
can be looked at indvidually.  This is more complicated because there is
intersection behavior, so I think it needs to have time for the middle
leg out of 3, or something that I haven't figured out, but still, one
should be able to ask "Did what the router said would happen happen".

Dealing with traffic is much harder.  For that I lean to a separate
database pouplated with reports by time/date and some inference, but
that's hard.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 162 bytes
Desc: not available
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20170822/e37ccdbf/attachment.sig>

More information about the talk mailing list