[Imports] How good can an import be?

[Steve Singer]

On Wed, 6 Apr 2011, Andy Allan wrote:


> Great - a complete duplicate. And have a look around at the other
> buildings - a broken import, a nightmare to fix, and impossible to
> imaging anyone making such a hash of it by attempting to map things
> the OSM way. And here I remember the excitement about importing DC
> building outlines, people twittering about it and congratulating each
> other - it makes me despair that such poor standards are acceptable
> and zero effort goes into fixing the mistakes. I mean, it was me who
> first started the discussion on fixing tiger county boundaries, years
> after the import. It was me who lead the first effort to connect the
> east and west coasts of america (and the second, and the third),
> again, years after the import. I don't want to be the guy organising
> "wikiproject $IMPORT fixup" for ever more.

I think a big part of the problem is that import guidelines set a fairly low 
bar for an import and most imports don't even follow those and 95% of the 
time there are no real concequences for not following the import guidelines 
if you can 'complete' your import before anyone notices.

Most discussions on imports tend to focus on banning imports rather than 
trying to come up with ways of making them work in practice.

Another set of guidelines might be

* All imports must involve submitting a proposal that must be approved. 
Approval will take time.  If your not in this process for the long-haul then 
you shouldn't be importing data.

* Import proposals must demonstrate that their is concensus in the community 
to import the type of data being imported.

* The proposal must address how the import will try to detect or avoid 
duplication with objects that are already in OSM

* The imported data must be approriatly licensed.

* A sample of the import must be performed on a dev instance using the 
proposed toolchain to allow the community to asses the tagging of the import

* The person doing the import must demonstrate that they have sufficient 
experience & skills to deal with problems that arise during/post import.

Yes I completely avoid the question of who 'approves' these proposals, and I 
realize we culturally prefer optional guidelines to rules but that hasn't 
been working so well (with respect to imports).

The other half of this is that we need to be willing and able to quickly 
detect and revert imports that people conduct that fall outside of the 
guidelines (ie haven't yet received approval).



>
> Cheers,
> Andy
>
> _______________________________________________
> Imports mailing list
> Imports at openstreetmap.org
> http://lists.openstreetmap.org/listinfo/imports