On the road with the community
By Giuseppe Maxia on Nov 26, 2008
Notes of one week with MySQL community in US and France
I wanted to attend the first edition of the Open SQL Camp in Charlottesville, from November 14 to 16. For some mysterious reason, a four days plane ticket to any place in the US costs EUR 2,500, but if I stay 8 days, it costs EUR 800. Considering that I was in Frankfurt until November 8, the most sensible thing to do was flying to the US as early as possible and meet as many communities as I could. Among the ones I asked, Boston and New York answered enthusiastically, and then, serendipitously, I found an unexpected group in Baltimore, just before the last leg to Charlottesville.
I sent all groups a list of ten topics to choose from:
\* MySQL 5.1 features
\* Using MySQL partitions in practice
\* testing with MySQL Sandbox
\* MySQL Community How To
\* Recent community additions to MySQL code
\* Creative cross-language programming with MySQL
\* MySQL Proxy wizardry
\* advanced replications techniques
\* advanced Lua programming for MySQL Proxy
\* How to fake a list of ten with only nine topics
Boston MySQL meetup
The Boston user group has a clockwork organization. They were able to accommodate my request for a meeting quickly and efficiently. The meeting takes place in one of the MIT buildings, and a large load of pizza and soft drinks was delivered on the spot, so the meeting started with everybody happily fed.
The topic chosen by this group was MySQL Sandbox. The presentation lasted 50 minutes, followed by 30 minutes of discussion.
The MySQL user group in New York is quite large. The venue for the meetings is a pub in South Manhattan that can accommodate 40 people comfortably seated, or 60 if there is standing audience.
The room was filled to capacity for the talk about MySQL Proxy. The participation was intense and widespread. We spent about 2.5 hours discussing the subject and related topics. There is much interest in the Proxy as a component for HA solutions.
Host of the evening was Philip Antoniades, who is widely known in the community.
The second day in New York should have started with a presentation at Columbia University, which was postponed, and thus the ad hoc presentation made by Dups for this occasion will have to wait for another event.
Dups joined me in late morning. We had lunch with Ronald Bradford, and then spent the afternoon discussing the status of the community.
Dups presentation at the PHP group was a success. The room (the same as the previous day) was packed. The contents were very much appreciated and discussed afterwards for a few hours.
Philip did not attend, because (lame excuse!) his wife was delivering his second child, just while Dups was delivering his first presentation as Community Manager.
Baltimore MySQL meetup
Transfer to Baltimore using the Greyhound. We found out that the Amtrak train cost $400 for the ride, while the Greyhound was $70 for the two of us. The ride was very useful. It allowed us to discuss the community needs in more detail and to make plans for a MySQL Proxy extension that will deal with memcached transparently (or at least Dups says he will do it. We'll see how good his Lua skill will be in a while).
Some excitement came in the cab ride from the bus station to the hotel, when the cab driver got lost twice and had to ask for directions.
Fortunately, we paid a flat fee for the ride, and no damage was done. The host of the MySQL meetup is Greg Haase, who is known in the community as the author of a convincing use case for MySQL 5.1. He is also a supporter of MySQL Sandbox, for which he is creating some additional features.
The topic in Baltimore was MySQL Proxy, with a very passionate debate following, and I am happy to say that I was able to address all the concerns.
We ended the evening with a dinner at a popular restaurant, packed with very noisy football enthusiasts.
The locals were genuinely surprised - no, better, amazed! - that we came by bus, and we actually survived!
Charlottesville Open SQL Camp
Public transportation from Baltimore to Charlottesville was not an option, so we rented a car. While I drove, Dups interviewed me efficiently, and took notes about the wishes and needs of the community team. It resulted in a tentative plan for engaging the community in a better web environment. Stay tuned for news on this subject in the next weeks.
The initial day was about registration, introductions, and finalizing the schedule for the next day. The efficiency of the organization was impressive. Considering that Baron has done most of the work on his own, with little help from others is really commendable, and an example for who wants to organize something similar.
The schedule was hand made, quite useful, and fit for this kind of event.
There were about 60 attendees, out of 100 registered people. The level of the audience went from intermediate to advanced, with very few beginners.
The meat of the conference was on Saturday. The efficiency of Baron's organization (with some sponsors help) was evident with the huge breakfast provided to the attendees.
Th event was dominated by MySQL/Drizzle presence. Brian Aker's keynote on the ubiquity of open source databases caught everyone's attention.
Some of the sessions were very innovative (libdrizzle, with its evolutionary features and backward support for MySQL, is one of the breakthrough topics), others, like the PostgreSQL ones, were more traditional and down to earth.
The lightning talks were much appreciated.
After the normal sessions, there was planning for the hackfest on Sunday and an unplanned prize giving, when we awarded Baron with a T-shirt signed by all the participants.
The inner force behind the organization was Sheeri, who has organized the wiki, the donation page, a legal entity for donations, and a sergeant major attitude which, together with her personal example, involved many people in the organizational tasks.
Networking was the major activity of the day. The lunch (not enough food for the latecomers, I was told) and the dinner (too much food) catalyzed the attendees together in temporary bonds that often led to serendipitous findings.
Met a lot of people, and expanded my Facebook and LinkedIn portfolios.
The hackfest was a collection of coding efforts in different directions. Several bugs were found and fixed during the day, in both MySQL and Drizzle, while an ambitious group of web developers tried to create a startup in 6 hours (they failed, but just barely, and it was a splendid effort!)
The travel to Washington brought more brainstorming, facilitated by some traffic jams. Dups drove, while I collected the ideas, and we arrived to Washington with a fully detailed plan about enhancing the MySQL Users Conference with some surprising things that you will hear about, probably next year.
That was the end of the good news. The meeting with the local Sun rep was canceled, and we were on our own. Following ill advice from Jay, we left the car at the airport and went looking for a locker for our luggage. There was none, and we were sent to a museum downtown that had lockers. After a quick train ride, we were at the museum, where our luggage was minutely and messily inspected, only to find out that the lockers did not have locks.
So we had lunch at the museum cafeteria and went back to the airport, where we scanned the ether for free wifi, and spent some productive hours online.
We parted company in the afternoon and that was the end of my US campaign.
Paris meetup is held very far away from the center. It's a different town actually, which gave me time to catch some sleep in the cab ride.
The attendance was not high, but everyone was interested, and all resisted
until the end. How many? We were 10, which is decimal ten, not binary two. The funny thing is that the next day, at least ten more people told me that they would have come to the meetup, if it hadn't been so far away from the center!
The topic was, once more, MySQL Proxy, and the questions asked were a bit different from the usual ones. Someone wants to use MySQL Proxy as a front end to hide the complexity of using different versions of MySQL in a sharded environment.
After Proxy, Nat Makarevitch, a local celebrity presented a long topic on server optimization, which lasted until 10:30pm, when finally we had food and drinks.
More talks, met more people, was asked to explain Proxy features again, discussed the Infobright roadmap, collected more business cards, and fell asleep in an armchair, only to be woken up suddenly for an incoming meeting.
Oh, well! it was an interesting exercise, but don't ask me to do it again soon!