Okay, here's my take on OSCON 2007. I had a great time. It was good to be surrounded by professionals who do and care about many of the same things I do. Most notably: software development and Open Source. I met several folks and, being my first real Open Source event, I saw quite a few famous people whom I'd only seen in pictures or Revolution OS.
Tuesday: Travel
I skipped the tutorials (though, not completely willingly). That was mostly because I didn't want to be away from my wife and 7 month old son for longer than 4 days. I traveled to Portland direct from Manhattan, Kansas. Unfortunately, my flight arrangements were such that I was on three flights rather than just two. (My flights were not made by me.) I do thoroughly enjoy the first 40 minute jump from Manhattan to Kansas City on a Beechcraft B1900 turboprop. I also like to laugh at the n00bs on the plane that are a little nervous about riding a plane only holding 16 passengers. "Oh my, it still has propellers." N00bs.
My trip from Manhattan to Kansas City to Phoenix to Portland was pretty much uneventful. I caught the Trimax into downtown and walked over and checked into the hotel. I then met Lance for some Chipotle and headed over to the Doubletree where we met up with Seth and hung out in the bar for a bit. Then back to sleep a bit.
Wednesday: Sessions Day 1
The first keynotes of the day weren't very interesting. I now know what Mark Shuttleworth and Tim O'Reilly look like, but the only talk I found very interesting was the one on Transactional Memory and that for technical reasons I won't elaborate here. As with most conferences, they always try to pack as much of the best stuff into the first day of sessions since that's the day most people stay for. There were at least 3 sessions during each period I wanted to attend.
After the keynotes I hit the expo hall and started picking up free T-shirts, pens, and other swag. Here I met with James Turner, who's been my editor for a couple of my OnLamp.com articles. He was showing off his work on his comic at the Watering Hole. I also saw Mark Tiemann, current OSI President and founder of Cygnus, here, whom I recognized from Revolution OS, handing out OSI T-Shirts in exchange for donations.
For my first sessions I went to How to Herd Cats and Influence People with Jono Bacon of Canonical. The basic recommendation of the session is that if you want a successful community make it very easy to participate. Period. There was a comparison with McDonald's and how they've made it so easy for morons to make burgers and that's the secret to their success. In Open Source, we should learn to do the same.
The second session I attended was Care and Feeding of Large Open Source Applications by Perrin Harkins, which was basically a talk on the project management tools available for Perl development. It was largely a review of what I already knew on the subject, but he did provide some good suggestions that I hadn't really worked with before, such as using Test::Class.
I ate lunch with Seth and a couple other attendees and then moved on to the afternoon sessions. I nearly went to Beautiful and Unique Snowflakes: Cooking with Catalyst with Matt Trout, whom I've chatted with on IRC but missed shaking hands with at OSCON, unfortunately. However, I saw that his session was introductory and I've already built Catalyst applications (including a blog I used on this site briefly), so it probably wouldn't have been the best session to attend. Instead, I went on to Who Am I? The Age of the Digital Identity by Robert Richards, which was about OpenID and InfoCard. His talk was interesting, particularly since I'm writing an SSO application, CAS+, but I had a hard time following the talk. I hope to implement OpenID in my application, but I can't just yet (or at least, not as the primary login mechanism) as our clients aren't quite ready for it.
After this I went to the Perl 6 Update with Larry Wall, which was interesting as ever. One reason I got into Perl 5 was because of what I'd read about Perl 6. I decided on Perl as my language of preference during my first year as a grad student. I had hacked around in Perl before while working on Anomy Mail tools when I was with Network Resource Group, but I didn't take any particular liking or disliking to the language at that time. As I started my Masters Degree I was trying to decide which language I liked for building web applications since I already new I didn't like Java and while I liked C++, it's not considered an ideal language for web development. I tried out Python and then taught myself some better Perl and felt an affinity to Perl immediately. I can think Perl. I read the first couple Apocalypses and was sold when I found out what Perl 6 would (eventually) be. Larry Wall and the Perl 6 language team continue to impress with more good ideas and the elimination of more bad ideas.
After the Perl 6 Update, I went to the Programming with SQLite talk by Michael Owens with Lance and Seth. The main thing I learned from this talk is that SQLite isn't like other SQL databases and requires a little more manual intervention to make things work just the way you want, which can be seen as an advantage or disadvantage. That is, if you want to use SQLite efficiently, take your locking seriously.
Lastly, I went to Myths of Innovation by Scott Berkun and then promptly went down to the book store and bought his book. I don't agree with everything in the book (particularly the veiled hostility toward spirituality), but it is a good book overall.
At the end of the day, I took Lance and Seth out to the Rock Bottom Brewery where they had a couple pints and a microbrew sampler. I even drank a couple of the samples and enjoyed them as much as anyone who doesn't drink can. We then headed over to the OpenSolaris party at the Doubletree and I met Donnie, who's a PhD student at OSU, one of the top guys at Gentoo Linux, and a friend of Lance's. After chatting at the party for awhile, Lance and I headed back to the Red Lion to end another day.
Thursday: Session Day 2
The Thursday keynotes were very interesting. The keynote by Robin Hansen on Bias was simply depressing, but basically stated a fundamental Christian belief, people have a veil pulled over their eyes and refuse to believe truth when it whacks them in the face. Of course, he didn't make any spiritual remarks and my statements would be construed as my own biases in his view. He also made some good points that the key to discouraging bias in the work place is to use competition or even gambling as a mechanism for encouraging truthful predictions of project timelines and such.
Bill Hilf of Microsoft announced that they were sending their Shared Source licenses to the OSI for approval, which got applause. I don't really remember anything else about his talk. He was followed by Rick Falkvinge of the Swedish Pirate Party, which is a political party in Sweden based on the single issue of IP rights. He made some very good arguments about why copyrights are being abused to harm consumers and individual liberties while working to aid corporate and governmental power. His platform is very interesting and I've added Sweden to the places I'd like to visit someday.
The best talk of the morning, though, was Steven Yegge's talk about marketing and branding. He works for Google and was basically saying that OSI needs to remember to be focused on branding. Our brand is very important and the brands of individual products and even developers help us market our products to the wide world. He was very entertaining as well, especially since I think he was probably best representation of the audience that we saw give a keynote. He was one of us and I think he was persuasive only because of that. If a marketing guy had stood up there and said the same thing, he probably would have been written off in that crowd.
After hitting the Expo Hall again for yet more free stuff, I went to the morning Lightning Talks. The talks I noted most were the talk by Donnie on the state of Gentoo and the talk on the current state of Linux, which is mostly funny because the speaker (whose name has escaped me) and his daughter give the talk together. He reads off every patch and improvement made in the last year and she holds up signs summarizing things and then mocking him and penguins toward the end. The most important statistic: Linux receives an average of 3.8 patches per hour all year long.
After lunch, I did want to go to Sam Vilain's Next Generation Version Control systems to learn more about GIT, but I decided to go shopping with Seth instead to pick something up to bring home. I ended up with a shadowbox frame with the cow jumping over the moon for Gabe's room.
When we got back, I went to straight over to Jesse Vincent's talk on Domain Specific Languages, which is interesting talk for me since it was about software I used regularly. It's also interesting because he was discussing work he and others have been doing on Jifty to create mini-languages in pure Perl. The declarative language used by Jifty to create models, action definitions, dispatchers, templates, and (coming soon) web application tests. All of these are written using pure Perl, but it doesn't look like traditional Perl because of how he's stretched the syntax a little bit. I like the ideas a lot and I've always been a fan of DSL's (I did write one for my multi-agent system for my MS a while back). This was first opportunity to do more than just shake Jesse's hand as we hadn't met but over IRC previously. I also stood by and watch Jesse, Larry Wall, and Matt Trout discuss some backport issues from Perl 6 to Perl 5, which was interesting and a little amusing.
Afterward, I went back to the Expo Hall to hang out with the OSL guys for a bit before the evening sessions. In the evening sessions, I went to the Perl Lightning talks. Andy Lester gave his talk on ack (and alternative to grep) and his assertion that we should call Perl work "programs" rather than "scripts" (which got cat-called by Larry Wall). There was also a rendition of a song by Bare Naked Ladies with new lyrics, Perl in a Nutshell. After the lightning talks, I went to the rest room and met Chromatic (another of my OnLamp editors) on the way out and chatting with him on the way back to where Larry was going to give the State of the Onion. At one point, I had both him and Jesse in front of me so I suggested that we needed a Jifty book. Perhaps we will someday soon. I also got to meet a few of Jesse's "minions" and others here, though I confess that my brain has blurred all the names other than Kevin Falcone.
Outside we waited for the Perl Foundation auction to end while everyone made fun of the proceedings inside. However, we were shortly inside and listening to Larry Wall deliver the State of the Onion address, which is something I'd been looking forward to. Larry gave a summary of his history with language and reasserted his view that programming is hard and scripting is easy, so he strongly disagreed with Andy Lester's statement about what we should call Perl application components. He went on to discuss all the different areas where a scripting language can vary: strong typing or weak typing, early binding or late binding, etc. He also showed why he thinks Perl 6 is going to win in the end because Perl 6 isn't so much a new programming language, but the root of a family of programming languages, which played nicely into Jesse's earlier talk about DSLs.
After this, I dropped my stuff off at the hotel and hopped on the train to get over to the Thirsty Lion to attend the Open Source Lab party, sponsored by Jive. We chatted some and then went to the Irish Pub across the street to get some dinner. Lance needed to get back to Corvallis so he checked out and headed home and I spend the evening in the hotel room half-working and half-relaxing.
Friday: Session Day 3
Unfortunately, the last day had come. I wasn't really ready for it to be done. But it was. The keynotes today were certainly the funniest and most memorable of all. Philip Rosedale of Linden Labs gave a talk about Second Life. The most interesting thing about his talk was that Linden Labs does all their meetings in Second Life, which I think would be cool if they had better tools for collaboration that were easier to get to, but with everything going Open Source, perhaps they will.
Jimmy Wales of Wikipedia fame gave the second talk about Wikia which is the "Wiki to End All Wikis" (or he would like it to be) and also about his hope to provide a true Open Source search engine where everything from the algorithms to the data is all open. Which is a good idea, if it can get enough support to work.
Then came the "Keynote Resignation" of Simon Wardley where it was announced that Fotango (though he didn't use the company name) has refused to let it's subsidiary Zimki Open Source it's web development platform, which Wardley was to announce during his keynote. Too bad. It looked interesting and I'm guessing that if they'd Open Sourced it Zimki could have had a chance at becoming a bigger player, now they will be relegated to the same obscurity as the other hosted frameworks. However, his replacement talk on IT Commidization was very funny, though I have no recollection of what points he made other than he likes ducks.
Open Source therapy by the conference MC, Nat Torkington, was basically just an attempt to tell us all to get along with one another. Which I certainly agree with. I may mock Python, Ruby, Java, and PHP every now and then, but I don't mock those that use them (unless they really are worth mocking for other reasons). Anyway, there are certainly lessons that Perl can learn from each. PHP, for example, teaches that ease of deployment can make all the difference in the world to popularity. It also reminds of how bad things get if a language doesn't provide namespaces.
Pimp My Garage by James Larrson is easily the most humorous of all the talks. I hope he posts his videos online, but he basically like to take household junk and revamp it so that it's either funnier or more dangerous. He had some excellent visuals and I think he has a thing for tall leather boots.
For the last two sessions, I thought about going to Hack Your Manager, but ended up in Adventures in Copyright Reform with Karl Fogel, which was somewhat enlightening as to how the RIAA and MPAA are trying to restrict civil liberties in order to help their clients (record companies and movie producers) continue making money the old fashioned way rather than adapting to the new media. The new media tends to favor the artist over the producer, which producers don't like very much since there's a lot of money to be earned in production.
Finally, I went to Running Your Language on Parrot with Chromatic and Patrick Michaud. One of the things I regret leaving K-State for is the fact that I no longer get to play with Parrot as much. I was hoping to use Parrot more to teach my class on Computer Architecture, but, alas, I don't have much need to play with it as a web developer. However, the talk showed me all the new cool toys they've added in the last two years since I've played with it and it's getting to be very nice indeed. If you have a chance to play with Parrot, I highly recommend it.
The End of the conference. I didn't go to the last talk since I wasn't very interested in Open Source Hardware (I'm really not much of a hardware guy) and worked on some stuff for Jifty instead.
Friday: Corvallis
After the conference ended. Seth and I grabbed some lunch at the Doubletree, picked up a rental and headed south to Corvallis. Corvallis is the home of Oregon State University, where Lance now works for Open Source Lab. He gave us a tour starting with meeting his boss who was still around waiting to start an after hours meeting. He then showed us the lab itself including many of the Open Source projects they host. He took us in to his office where one of his students was still working to fix some DB problems with Drupal.
I'm a little jealous of Lance and all his toys. However, I enjoy software development far more than I enjoy systems administration, so I'm not that jealous (but a little). OSL has a very nice facility, much nicer than anything I had at K-State. Lance also has some phenomenal students working for him. I am jealous for that being the lone wolf at my current job.
After the tour, we went out to Newport and had some seafood on the bay and then headed back to Portland. Seth and I said farewell and so ended my time in Portland.
Saturday: More Travel
My trip back to Manhattan was pretty uneventful, so I won't bore you with anymore details.
I had a lot of fun on my trip to OSCON and I look forward to making the trek again next year.
Cheers.
Update: (2007-09-30) Corrected references to Andy Lester. Thanks for the correction, Andy.
