[Talk-us] [Talk-us-newyork] Highway classification guidelines for New York State

[Eric Patrick]

>
> I would not focus too much on routing outcomes, but more on whether
> produced
> road structure is sane/reasonable.


If trunk wasn't placed so high in the routing engine, then I can understand
that argument. Since motorway is the only road type above it, there is a
lot more importance placed on trunk than the other road types. Why does
road structure become important over functionality? I understand where one
may not want to be routed onto dirt roads (Georgia's FC has a lot of that),
but what's the real difference between a rural 2 lane road which connects
two medium sized cities over an urban 4 lane road within a metropolitan
area?

I'm a router. My mind thinks like a routing engine, I'm always looking to
see how things can be routed more efficiently, it's just how I am.

On Tue, Sep 28, 2021 at 3:26 PM Mateusz Konieczny via Talk-us <
talk-us at openstreetmap.org> wrote:

>
>
>
> Sep 28, 2021, 19:14 by txemt1 at gmail.com:
>
> I'm not suggesting that a router needs to apply different heuristics for
> a highway=trunk in an urban area versus a highway=trunk in a rural area.
> Rather, I'm pointing out that Principal Arterial covers a variety of
> roads that *should* be handled differently by renderers and routers, so
> we should represent that distinction in the highway=* tags we choose.
> Assigning these roads the same tag is what would lead to funky things,
> as I have already illustrated. If the FHWA functional classification
> system doesn't help us make the distinctions that matter for our use
> case, then unfortunately we'll have to base our decisions on something
> else.
>
>
>  There's a multifaceted problem here. We're trying to please everyone,
> which won't happen, but by not doing something, we're still not pleasing
> everyone. There has to be some give and take between all of the uses for
> OSM. Changing PA to a trunk may work for vehicle routing purposes, but it
> may not work for bicycle routing purposes.
>
> I would not focus too much on routing outcomes, but more on whether
> produced
> road structure is sane/reasonable.
>
> Note also that mapping more info about road will reduce impact of such
> assumptions based on highway=* value.
>
>
>  Why should this particular functional classification
> system enjoy prominent placement in the highway=* key while another,
> perhaps better suited system should go untagged?
>
>
> Is there a better suited system?
>
> +1
> _______________________________________________
> Talk-us mailing list
> Talk-us at openstreetmap.org
> https://lists.openstreetmap.org/listinfo/talk-us
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk-us/attachments/20210928/164f9321/attachment-0001.htm>