[OSM-talk] Highway = Motorway automatically oneway ?

[Ulf Lamping]

Dermot McNally schrieb:
> On 14/10/2007, Abigail Brady <morwen at evilmagic.org> wrote:
>
>   
>> Any competently written tool will assume oneway=yes on motorways, if only
>> because of the vast quantities of data that does that.
>>     
>
> I think we need to be careful what we infer in cases like this
> (especially if we infer it without telling anybody). While I can't
> think of any current stretches of motorway that are 2-way, Germany as
> one example is full of 2-way stretches of motorway_link, as here:
>
> http://www.informationfreeway.org/?lat=48.19084402854966&lon=11.704346359824747&zoom=16&layers=0B00F000
>
> >From what I can see, osmarender is inferring oneway=yes on the
> motorway_link sections, even though only the parts explicitly tagged
> with the property are thus tagged.
>
> I've never been that comfortable with the assumed-one-way status even
> on roundabouts (where I've yet to see an exception), but I'm fairly
> opposed to it in this instance (at least for motorway_link).
>   
You are making a lot of assumptions with saying that motorway is 
automatically oneway.

I know some parts in the alps (italy or switzerland) where motorways 
don't have strictly seperated lanes for both directions, so the oneway 
tag is not the right choice here AFAI understand this tag.

I don't know if it really fits well in other parts of the world at all.

So because your idea of a motorway is only oneway (with two seperated 
lanes), you shouldn't assume that this is the case anywhere else in the 
world.

And because of that, any competently written tool that assumes 
oneway=yes for motorways is just making a mistake IMHO.

Regards, ULFL

P.S: motorway_links (at least here in germany) are *very often* not 
oneway, already noted!