Atul Varma

Ok cool, that makes a ton of sense! I've looked at the `.sql` files and they seemed pretty straightforward. My one main discovery was that some of the derived tables...

Hmm, some more thoughts: * It looks like materialized views are actually literally the same thing as just creating new tables, only you have to manually issue a `REFRESH MATERIALIZED...

Ohh interesting, I didn't realize the CSV schemas themselves change! Oof!

Haha yeah :) So yet another alternative is just to have a `pristine` schema that has the result of the load without any transformations applied. Then an update could just...

Oh, here's what the comment looks like in Postico, courtesy of Mr. Dan:  Unforuntately I was hoping the comment would also show up in the column's tooltip when viewing...

Er, also, the script I made has a lot of limitations: it assumes you're creating a dataset based on a single CSV file, so it won't help as much (and...

I think that [NYC Planning Labs GeoSearch](https://github.com/NYCPlanning/labs-geosearch-docker) is somehow able to figure out what the latest PAD URL is, so that might be a place to look for fixing this....

This looks great! I am strongly for adopting semantic versioning and this changelog file--the change in #140 would have necessitated a change in the major version number, which (along with...

An interesting thing I just noticed is that [`first_last.sql`](https://github.com/aepyornis/nyc-db/blob/master/src/nycdb/sql/hpd_registrations/first_last.sql) seems to explicitly define functions in the `public` schema. All the other SQL files and operations _seem_ to be schema-agnostic. Should...

I believe it does mean that the datasets effectively aren't updated, but this error has only occurred once (last night). If it keeps happening then that is definitely a problem!