places.sql is hardcoded

[beusj]

place.sql, which includes prioritization has not been updated beyond 2010 census data. The priortization column appears to deal with different cities that have the same zip/postal code. See commit a3eeee. The appears to have come with the initial libraries imported from geocommons. I can't find documentation of the derivation anywhere, including in the original perl implementation.

[cole-brokamp]

This would be great to update. We've used the USPS city names scraped before, but have not purchased a comprehensive solution because agreements would prevent us from distributing it in source code. Is there an updated places name data source we could use?

In practice, we rarely ever rely on place names and instead match via zip codes. I think this is only called when using geocoding that doesn't include any matched zip codes or street names, which we don't utilize for geomarker assessment because it usually is not precise enough.