[Talk-us] Starting OSM Trail Map Initiative In US

Kevin Kenny kkenny2 at nycap.rr.com
Sun Jul 22 15:20:44 BST 2012


On 07/20/2012 09:27 PM, Richard Weait wrote:
> I can't think of a good substitute for a motivated local mapper.
> Daniel Begin said this recently on talk-ca:
>
> "You'll find that there is nothing better than an active community to
> find odd features in authoritative data!"
>
> Daniel knows of what he speaks, he works at the Canadian National
> Mapping Agency and has been participating in OSM for several years.
> He publishes authoritative data collected by paid professionals, with
> what I presume is top notch equipment.  And the OSM community of
> enthusiastic amateurs finds and fixes odd features and errors.
>
> I don't think that you can get that nearly obsessive attention to
> detail across a large area. It takes a _personal_ interest in the
> data.  I think of that area of obsessive interest and perhaps
> encyclopaedic knowledge as the natural range of a mapper.

I'd argue that this is entirely the right model. I'm sure all will
agree: it's hard to recruit individual mappers - although that has
to be a primary goal, because only someone with detailed local
knowledge can do the fixes just mentioned.

But by the same token, it would be foolish to discard the work -
often of extraordinarily high quality - that has been done at
taxpayer expense to curate the large data sets. For the most part,
it's better than what we'd get from a horde of novice mappers.
In any case, the government agencies that administer our public
lands are the ultimate sources of information like "{horses,
mountain bikes, skiing, snowmobiles, ATV's} are/are not
not permitted on this trail," or "camping is permitted on this
trail only 200 feet or more from the trail at elevations below
3500 feet."

So I'd argue that selling the OSM model to the authorities is
just as important as recruiting OSM mappers. If we can get the
odd features and errors fixed upstream in the official data, we've
essentially recruited the professionals as mappers. We need both:
the professionals who have boatloads of high-quality data (with,
of course, the inevitable errors and omissions) at their disposal,
and the enthusiastic amateurs who will find and fix the errors.

I wonder if the people who argue that all the work has to be
done by individual mappers have simply despaired of ever getting to
the point where corrections can flow into the official sources.
I could argue that in many localities, that's the reason that
imports are problematic: they can happen once only, because the
public data are set in stone. The result is that when the public
data are improved - as with the latest version of TIGER - there's
no way to reimport without a tremendous manual patchwork to merge
the new data into the work that individual mappers have done.

-- 
73 de ke9tv/2, Kevin



More information about the Talk-us mailing list