[Imports] HIFLD

[stevea]

I echo what Mike and Greg have said here.  If there really are 570 separate datasets here, our community needs what is effectively that our Import Guidelines be followed 570 times.  That is such an overwhelming and daunting task that it "stuns at encounter."  Even as James says he has been preparing for "months," such a massive undertaking must at least begin with a (public, here is OK) recognition that there truly are 570 distinct Imports, and (I'm guessing) that some, many or even most of these (e.g. "Public Refrigerated Warehouses") are what I'll colloquially call "crap data" on both data quality (including its vintage and lack of being updated in many years) and "worthy of being in OSM as a distinct category" aspects.  The community really must be allowed to chime in on ALL aspects of this proposed Import, and we've only scratched the surface here, with the result that these data rather quickly seem to be characterizable as "not especially worthy of entry into OSM," at least at first glance.

It's been said many times, but it bears repeating here:  just because a government agency publishes geo data does not mean they should be Imported into OSM.  Far from it.

If there truly are worthy subsets of this gigantic dataset, they must be carefully selected for relevance (first and foremost), quality (must be OSM's declared-to-strive-for "very high"), freshness (2014 really is OLD in this realm) and wider community acceptance.  This is deliberately a high bar, maybe even a very high bar, but it is what OSM has repeatedly expressed before an Import happens, even a relatively small one (which this isn't).

I don't wish to sound deeply discouraging (and maybe I do), but the vetting of these data must continue at a rather close-scale granularity, even as some might say I'm being harsh by insisting upon this.  James, you are courageous to even begin to undertake this, but you must realize you are biting off a mountain as you chew, chew, chew.  Digesting this (digestion begins with mastication) will be a long process, let's not kid ourselves.