Closed philbooth closed 5 years ago
Yea, not convinced that the renaming is worth it at this point if all we need to do is wait 2 weeks until we get to real 10/50. It would just make more work for you to re-re-name after the re-import.
I already told the other data scientist who would be likely to use these tables about this issue so I believe we won't have any issues there. If anyone else is interested in the data, they will likely come to me or Alex and we both are aware of the problem so can give the proper heads up.
In that case I'll close this one, thanks @irrationalagent!
Related to #129.
Contrasts to #131, where that approach fixes the off-by-one error but this approach accepts the presence of the error and just renames the tables instead. Would need to be applied in conjuction with a bunch of
ALTER TABLE
queries:But, @jbuck and @irrationalagent, are we really sure it's worth doing this if we're also doing #131? This forces people to edit their broken queries twice, once now to pick up the new table names and then again later when they want to update to the fixed schema. Does it make sense to pay that cost for 2 weeks fewer of a bug that has been hanging around us for literally years at this point?
(I'm happy if you think the answer to those questions is "yes" btw, just wanted to make sure we were asking them before taking action)
r?