Eric Radman

My workaround was to restore the schema to an [ephemeral database](http://eradman.com/ephemeralpg/) before running the comparison tmp_url=$(pg_tmp) pg_dump -T 'log_*' -s $db_url | psql -q $tmp_url migra --unsafe $tmp_url ... Here...

Thank you--partitioning support is even more useful than an ignore option.

Right, `pg_tmp` doesn't test to see if the destination directory already has a database running. I thought about implementing a "status" subcommand, but I'm not sure it would make this...

I don't know of a utility similar to entr(1) for Windows, but I'll keep this open for a little while in case someone else knows.

Cool. Thanks for the lead @franciscod

> > would it be possible that patroni doesn't demote master at all when etcd cluster is unhealthy? > > No, it absolutely not safe. It's not obvious at first...

> > @SourceCipher Do you have any recommend? > > We have replaced with the https://numbrojs.com library most of our code base Numbro also has a fatal flaw, namely rounding...

@deecay do you have time to rebase this change? There are quite a few conflicts, but all of them appear to be easy to resolve

> Awesome! 😄 > > @eradman Any interest in reviewing this PR? Yes, I'll take this change for a test drive

I ran into some trouble with larger data sets on Chrome and Firefox. With as few as 500 entries both browsers would spin at 90% CPU for several minutes. Now...