I finally made it to a PgCon! I’ve been intending to go for a few years now, but have always had a scheduling conflict (like my birthday or some such nonsense.) This year the planets aligned, I had leave to attend, and my talk was accepted. In addition to speaking, I volunteered to work the registration desk, because it’s a great way to meet people*. I tend to stick to the West coast confs, so there were a lot of new faces here for me.
The conference location is wonderful: it’s an easy bus ride to & from the airport, there are nice hotels very close to the venue, and good eateries within a reasonable walking distance. I arrived in Ottawa a few days before the official conference, intending to take a day for some mountain biking. That didn’t quite work out as planned, but Dan Langille & I eventually made it over to Kanata Lakes for MTB-PgCon 2011, henceforth to be known as “The Great Bog Ride”.
Probably the most exciting announcement we heard last week: Postgres Open, a new conference that will be held in Chicago in September this year. I’m certainly hoping to attend!
The Royal Oak track was super-productive. I heard about per-user logging settings from Aaron Thul, but it turns out they don’t work quite like we expected. More on that in a later post.
Notes from scheduled talks I attended:
Josh Berkus 9.1 Mystery Tour. The two features I’m most excited about are Extensible ENUMs and triggers on VIEWs. Unlogged tables also sound interesting; you’d use them for high-volume tables that store ephemeral data, such as session ids and cookies, that maybe you don’t need to include in your data replication.
Stephen Frost – Review of Patch Reviewing. Stephen went over the review process and I am inspired to restart the PDXPUG Patch Review Parties as a recruitment strategy. (Now, to find the time to do so.)
Robert Haas – How to Get Your PostgreSQL Patch Accepted.
– run regression tests often (make check) so you catch problems early.
– don’t make superfluous changes. Sometimes stuff needs to be backpatched, and updating e.g. existing whitespace can make that process more difficult.
– Kevin Grittner recommends adding make check world, and something else I missed (can’t write fast enough!)
Tetsuo Sakata – NTT’s Case Report interesting stuff, but we can’t talk about it. ;) I am excited to try pg_bulkload and pg_statsinfo.
My talk on logging configuration and analysis went ok I suppose – I was told I went a bit fast, which is a problem I struggle with as a speaker. I talk about things that really excite me, and I tend to try to squeeze way too much information in. Next time I will find ways to enforce audience participation, which will slow me down a bit.
Last session of the day was the lightning talks. I learned about pgshark, which looks like a lot of fun to mess around with.
Jeff Davis – Range Types. Every time I go to one of his temporal datatype discussions, I understand a little bit more of it.
After that, I had to catch a plane. I wanted to hear about Bioinformatics in PostgreSQL, and I hope to get another chance to see that talk.
Thanks to the conference organizer & volunteers! See you next year!
*shameless volunteer plug: if there’s somebody you’re too shy to just walk up to, handing out registration packets is the perfect excuse to say “Oh hi! I like your work on [whatever].”