[OSM-talk] non-temporary usage of highway=road

Warin 61sundowner at gmail.com
Tue Sep 27 12:03:41 UTC 2016


On 27-Sep-16 05:44 PM, Martin Koppenhoefer wrote:
>
>
> sent from a phone
>
> Il giorno 27 set 2016, alle ore 08:07, Simon Poole <simon at poole.ch 
> <mailto:simon at poole.ch>> ha scritto:
>
>> Use of it, IMHO, doesn't really make sense at all any more, particularly
>> if you have imagery available, a conservative best guess plus a fixme
>> tag would seem to be much more useful.
>
>
> I tend to agree, but would limit this to areas where you know more or 
> less the context (regions where you have been on the ground).
>
> When mapping far from any place you have been to, it seems more 
> sustainable to not infer tags related to aerial imagery and your own 
> context, which might have little to do with the actual situation on 
> the ground. Adding suppositions coupled with fixme tags is not what we 
> generally recommend, and will only work in areas with good coverage of 
> active contributors on the ground.
>
> Better less but reliable data, than more but with a significant amount 
> of serious errors.
>

A 'conservative best guess' would normal place the type of highway at a 
lesser level (lower capability) than what it could actually be.
Any error that has a lesser level (lower capability) would not be 
'serious' in a safety aspect.
Additional clues can be had from existing tagging of similar features in 
the area. In this way the result will be consistent and can be readily 
interpreted by the user.

Most people when faced with things outside their range of knowledge will 
take a very conservative approach to assessing capabilities.

-------------

suppositions: a belief held without proof or certain knowledge; an 
assumption or hypothesis.

As the assessment would be made with an image .. and that image can be 
taken as proof the use of the word 'suppositions' is incorrect.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.openstreetmap.org/pipermail/talk/attachments/20160927/2ef25603/attachment.html>


More information about the talk mailing list