[Imports] How good can an import be?

[Richard Weait]

On Wed, Apr 6, 2011 at 11:10 AM, Katie Filbert <filbertk at gmail.com> wrote:

[ ... ]
> http://osm.org/go/ZZcc60T6
>
> Is this crap??? Are my efforts crap? Is there any reason to stay involved
> with OSM?

I haven't looked at Andy's example, but should a rectangular building
have 47 nodes?  Everybody makes mistakes.  Dedicated, experienced
mappers who foot-survey make mistakes.  No problem.  But with an
import or bulk edit of some form, mistakes are invariably multiplied.
And, because we are looking at larger areas when we import and check
our import, it is more difficult to see a problem immediately and
resolve it.

Nobody decided, "I'll deliberately import simple rectangles with ~10x
required nodes."  But nobody caught that and fixed it before the data
went in either.  That's not because the contributor is bad, or
thoughtless, or mean.  It's because imports are hard.

Your efforts are not crap.  My efforts are not crap.  Some of my
import-results are crap.  So I've stopped importing until I can be
better at them.