[Tagging] Proposed relation give_way

Philip Barnes phil at trigpoint.me.uk
Sat Mar 16 21:41:13 UTC 2013


On Fri, 2013-03-15 at 11:53 +0100, Simone Saviolo wrote:
> 2013/3/15 Pieren <pieren3 at gmail.com>
>         > It is for navigation. Imagine a corner right before an
>         intersection: the
>         > navigator would warn you that you'll have to give way or to
>         stop, and you'll
>         > prepare the corner accordingly.
>         
>         
>         I don't know any commercial car navigation system that warns
>         you about
>         give-ways or stops. 
> 
> 
> This doesn't mean that future navigators shouldn't. 

I would very much disagree with you here, it is up to the driver to
remain alert and to drive within the limits of what they can see. The
golden rule is "never drive so fast that you cannot stop, safely, within
the distance you can see to be clear". 
On single track roads half that distance.

>  
>         Perhaps because it has no influence on the route
>         itself. 
> 
> 
> Don't mix routing with navigating. 

Give ways and Stops do influence routing as they add time  to a journey.
A longer route that avoids give ways and stops can often be quicker.
Particularly turns which require a gap in traffic in both directions on
the road you are joining. (A right turn in drive on left countries, and
a left turn in drive on the right countries.

Most drivers will select routes avoiding some junctions due to the time
taken, the queues that can form etc. Providing this information to
routers will help them produce better routes. 

A junction with stop sign will take longer to accomplish as the presence
of the stop sign indicates reduced visibility.

>  
>         And that the decision to stop or go depends on the actual
>         traffic. 
> 
> 
> Also the decision to stop because a traffic light is red. Also the
> decision to stop because there's a queue. This doesn't make it
> pointless to signal that there is a traffic light ahead, or that
> queues may be possible (there's even a road sign for that). 
> 
Again traffic signals affect the time taken to negotiate a junction, it
may be more or less time than a give way, but it is still part of the
algorithm.
> 
>  
>         And you cannot rely on OSM for your driving decisions.
>         Imagine that your navigator warns you 4 times to give way and
>         the
>         information is missing in OSM for the fifth... 
> 
> 
> Still better than having none. 
> 
Absolutely wrong, obeying traffic rules is 100% the drivers
responsibility. There already enough numpties on the road, without
giving them another reason to blame the satnav.

A satnav should only use the presence of give ways and stops to
calculate routing times and hence the optimum route.

Phil (tripoint)






More information about the Tagging mailing list