[OSM-talk] Google fumbles again in latin america

Craig Wallace craigw84 at fastmail.fm
Fri Nov 5 17:15:55 GMT 2010


On 05/11/2010 16:54, Serge Wroclawski wrote:
> On Fri, Nov 5, 2010 at 12:33 PM, Apollinaris Schoell<aschoell at gmail.com>  wrote:
>>
>> I consider it improving osm by a human mapper according the spirit of the project instead a container full of imports with not much value. If a human surveys on ground or based on personal knowledge and image tracing it has 100 times more value than any imported data
>
> We're not talking about human surveyed data- that is already addressed
> by tiger_reviewed- we're talking about disassociating the original
> feature from an import from its current incarnation. This doesn't
> improve anything, it just makes it harder to associate the data from
> the source.
>
> Also, in the case of image tracing, one is recommended to mention a
> source= tag. Not everyone does that all the time (I don't do it often,
> when I should), but the idea there is the same- to illustrate the data
> lineage.

Probably more useful (especially for imports) to tag the source on the 
changesets, instead of the objects. As the source is specific to those 
changes/import, and not the objects themselves, which may later be 
edited using data from a variety of other sources.
The changeset tags can be retrieved from the history if necessary.



More information about the talk mailing list