[OSM-talk] Surveying

David Earl david at frankieandshadow.com
Mon Nov 19 15:29:09 GMT 2007


On 19/11/2007 15:27, William Waites wrote:
> At 01:46 PM 11/19/2007, David James wrote:
>> In the area I've been looking at, I've been trying to locate speed limit
>> signs at the edges of villages, break the ways at the speed limit signs,
>> and then apply appropriate maxspeed= tags to the separate ways. Am I
>> wasting my effort here (as no-one else has yet mentioned speed limits).
> 
> Sounds useful, but perhaps there is a way to have much of this data with
> less legwork. Suppose that we have municipal and village boundaries. We
> know that the speed limit on the highway between villages tends to be
> standard (say 110kmph) and that within villages tends to be lower (say
> 60kmph) but likewise standard. Is it not then possible to tag the
> resultant segments with the appropriate maxspeed? It seems to me that
> in this way the dataset could be populated to a quite high degree of
> accuracy and completeness with relatively little data-entry work. Any
> errors that are subsequently found could then be corrected by people
> with more local knowledge and physical surveying capabilities as
> necessary.
> 
> In the case that there are worries about having potentially incorrect
> data, perhaps a tag such as maxspeed_bestguess=true could be applied to
> such line segments until they can be confirmed, without compromising the
> ability of route finding algorithms to function.

In general I've tagged streets within the urban envelope as 
highway=residential or highway=<whatever>;abutters=residential. I think 
that should hint to clients that the speed is the default. Then only 
when there is an exception need the speed actually be mentioned. (I've 
done that sometimes, but not been terribly diligent about it, though the 
info is all there on my dictaphone files).

David





More information about the talk mailing list