On 2018-08-10 15:35, Mark Goodge wrote:
> On 10/08/2018 13:14, Colin Smale wrote:
>
>> Who is the arbiter of relevance? I think for any given "mapper" or
>> "consumer" 99% of the contents of OSM is not relevant. People are mapping
>> the nuts and bolts of the insulators on electricity pylons.. I can't see
>> that being relevant to most people.
>
> Can you see the nuts and bolts?
>
> I don't think there's any real argument about whether or not we map things we
> can see. There may be disagreements about *how* we map them, but the basic
> principle that we map what is visible is, I think, pretty firmly established.
>
> The basic question here is how we go about mapping things which you can't see
> - intangibles, such as administrative boundaries, postcodes, road numbers,
> etc. And that's where questions of relevance come into it.
And by extension the dilemma where what is visible is demonstrably
wrong, e.g. a typo on a street name sign. If we stick to exactly what we
see, we propagate the error. If we apply a bit of QC and make the street
name consistent, we have a better map.
> The basic principle of OSM is that it is free, in all possible senses.
> It's free, but it isn't unrestrained. You can't just make up entries. You
> can't put Ambridge and Hogwarts on the map (although you can, now, include
> Platform 9 3/4). You can't label a road as a river and a wood as a
> skyscraper. To be useful, we have to agree to a common set of principles and
> then stick to them.
Yes you can, until and unless it gets noticed. There are no barriers to
creative, erroneous or downright mischievous tagging - except "social
control" by others, which is very hit-and-miss. In some areas mappers
may have "adopted" their town and monitor every change in a defined
area, and in other cases people might monitor the whole world for a
specific object type. But my suspicion is that most objects in most
places are scarcely "policed" in any way. It would be nice (IMHO) if
this ex-post moderation itself were to be monitored, to help ensure that
every little corner of OSM has somebody/something keeping an eye on it
to detect anomalous mapping, and all the "moderators" (human or
otherwise) worked to consistent standards. But then we get back to the
nub of the problem: who defines these standards, and thereby codifies
what is right and wrong? Nobody wants to burn their fingers on this
proactively, so we are stuck with a patchy, reactive system and the most
incredible inertia which kills many attempts to improve data
consistency. Sacrificing the good on the altar of the perfect...
_______________________________________________
Talk-GB mailing list
Talk-GB@openstreetmap.org
https://lists.openstreetmap.org/listinfo/talk-gb
