2013/1/10 SomeoneElse <li...@mail.atownsend.org.uk>:
> Christian Quest wrote:
>>
>> I that case nodes are not shared by the ways, but are duplicate (same
>> lat/lon) and as Clay mentionned, this will ring some other alarm.. I
>> would move one of them a little bit to avoid it.
>
>
> I can understand why you're saying this, but isn't education of remote
> "fixers" that an alarm isn't necessarily an error the better option here?
> If we can help them to think a little more, and interact with local mappers
> a little more, then surely that's better for everyone in the future? *
>
> The aim of OSM isn't "no errors displayed by QA sites", it's "a map that
> best reflects reality".
>

I agree, even more when the QA alarm could be considered as a false
positive: 2 overlapping ways with different layer=* tag should not be
considered as duplicates as they clearly represent different things in
reality one above the other.
It is useful to report false positives on tools like Osmose, as these
can be reviewed to improve the algorithms.

For example JOSM validator rings a lot of bells that do not always
require a fix without a possibility to have false positive feedback
due to its local nature.


> Cheers,
> Andy
>
> * as an aside I'd also suggest the same approach with new mappers - give
> them a bit of time to get the hang of things, then after that try and help
> them with issues that they might be having trouble with, rather than just
> diving in and fixing stuff.
>

I also agree.

I'm contacting many new mappers in France, and have a look at their
first edits (when I have enough time).
If I find errors like unconnected ways or very bag tagging, I also
mention it in this welcome message (as a "post-scriptum" because I
prefer to have the welcoming part of the message being the main one).
I may fix most errors not to discourage new mappers with too many
required fixes but usually leave at least a few as examples with
explanations on how to fix them.


-- 
Christian Quest - OpenStreetMap France - http://openstreetmap.fr/u/cquest

_______________________________________________
talk mailing list
talk@openstreetmap.org
http://lists.openstreetmap.org/listinfo/talk

Reply via email to