[Imports] How good can an import be?

[Anthony]

On Wed, Apr 6, 2011 at 4:24 AM, Andy Allan <gravitystorm at gmail.com> wrote:
> I'm not suggesting that
> we map everything with GPSes out on the ground, but I strongly believe
> importing data is detrimental to growing the community. Of course
> that's unprovable, but I believe there's only so many correlations
> before causation can be reasonably suspected.

I think the notion that importing data is detrimental to growing the
community is detrimental to growing the community.  And I think that's
a big part of the causation of the current stagnation of US interest.

The US is fairly unique in the world in that all the federal
government created databases are public domain, as well as much of the
state created databases.  Yes, there was a bad TIGER import, which
wasn't even that bad at first but was not designed to facilitate
updates (importing TLIDs was a huge mistake, for starters).

You can't build a competitive map in the United States without
utilizing imports.  It's just idiotic to throw away all that hard
earned public domain data.  You're giving up too much data that all
your competitors have, and that they can and do use without
restriction.  That may not be true anywhere else in the world (it's
certainly not true where the government data is copylefted), but it's
true in the US.

[quote]Make a mistake with the import, just one; duplicate too many nodes,
overwrite existing data, import from a low-quality or old source, fail
to connect it to existing data, and suddenly you are causing harm.
Net harm.  Not just "less good" but straight into "bad."[/quote]

That's absolutely true with OSM.  But it's not an inescapable fact of
life.  It's a fact derived from the lack of support of OSM for
imports.