seafood

GPCE Post Mortem

Sorry about the lack of posts during the conference - too much was happening for me to slow down to write a thoughtful entry. What can I say? - it was a transformative (dare I say "life changing") experience. There was so much going on concurrently that sometimes it was hard to choose which talks to attend over others.

There were three conferences being held in Vancouver - OOPSLA: Object Oriented Programming Systems, Languages and Applications, ISMM: The International Symposium on Memory Management and GPCE: Generative Programming and Component Engineering.

They were all held inside the Vancouver Convention Centre, a truly lavish and splendid building. They must spend a lot of money on such conferences. You'd notice it in the little things - like the fact that the signs on the room where of considerable quality, yet obviously custom made. Either that or there's a whole lot of regular guests who like to use the MetaOCaml room.

Day 1 - Sunday

I quickly learnt that GPCE was meant to be pronounced Gypsy, but this conference wasn't due to start for another couple of days. Incidentally it's the one I gave the talk at. In the morning I attended the Source Transformations Systems Workshop and came to the realisation that I should have been talking to (or at least reading the papers of) these guys a long ago. I also quite liked the format - you had only ten or so minutes to explain you basic idea, after which there was extended discussion.

In the afternoon I attended a MetaOCaml tutorial and met Walid Taha, a former PhD student of Tim Sheard's. I've noticed something about researchers - there is a high correlation between their ability to explain themselves and the esteem with which they are held in the research community. This, I feel, is as it should be; people who are brilliant but nevertheless completely useless at explaining themselves have always irked me.

Walid was nothing short of clear. He was also extremely friendly; I went from feeling mildly intimidated to feeling right at home with him in very short order. I'd actually already read the slides he used to present the tutorial but of course, like most slide presentations there was much implicit in them. The opportunity to ask questions was also very valuable.

I also enjoyed the pace at which this tutorial was pitched. Usually I find conference talks to be delivered at a pace which outstrips my ability to comfortably keep up. But in Walid's tutorial I felt none of this; in fact at times I felt that maybe he should speed up a little.

Day 2 - Monday

I'd had a terrible sleep the night before. I think it was a combination of jet lag and the fact the people next door were making a lot of noise. Then an alarm went off in another room in which either no one was sleeping or they'd already left. I kept thinking that in just another five minutes it would stop. It never did. But somehow I slept despite the noise, jolting awake some time later to discover that it was 8:45 and that I'd be missing the first talk of The First MetaOCaml Workshop.

I actually ended up missing two of the talks. The rest were great and I didn't fall asleep during any of them. If I seem to be making a big deal about sleep it's because of how important it is to my concentration on difficult concepts. I'm well aware of how I work. Usually, even if my greatest wish is to stay awake, if I've had anything less than a restful sleep of at least eight hours the night before I'll find myself nodding off. At heart I guess that although I can learn from the passive lecturing style of presentation, I'd rather not. I want to be able to jump in at any point and ask questions. I've heard from a number of sources that even some of the brilliant researchers are like this. They get the most out of the conference experience by seeking out people and having one-to-one chats with them.

That night I prepared my slides further. I finished around 12:00am and got into bed. And promptly couldn't get to sleep. What!? I'm tired, I thought. I can feel it. My eyes are sore, my muscles ache. What I require is sleep. I finally managed to drop off at 2:30am after disciplining myself to stay still and breath deeply. In the intervening time I finished off the first Harry Potter novel and watched Canadian television. I even watched some of the French channels for fun.

Day 3 - Tuesday

GPCE began. I was hideously tired. I found it hard to concentrate. I went home after lunch and got two hours sleep thus missing three talks. Two of them I didn't mind missing but one of them was about a domain-specific language for scheduling OS process. Thankfully I heard from a fellow attendee, Ben Kavanagh, that although the paper was very good the presentation wasn't actually all that great.

I felt much better after the sleep and really enjoyed the final three talks. Then we attend the Turing Award lecture by Alan Kay, one of the implementors of Smalltalk. That's how you give a talk. It was titled, "A First Course in Computing should be Child's Play." In it he expounded upon his views of the deficiencies in teaching computing. He said some really cool things. First, he stressed the need for simplicity. He said something that I've believed for a long time and this is that if you can't explain something to the lay person, the novice or a even a child then you haven't spent the time simplifying your subject matter to the point where it is truly useful. Nor do you probably understand it all that well yourself. Alan also remarked that he always found it remarkable how much work you were required to do at University. He felt this distracted one from narrowing in on what was important and what was not. He also felt that simplicity was the panacea to this sisyphusean burden.

He demonstrated some software he'd been working on to explain fundamental computing concepts to children. As we watched he drew a childlike picture of a car and then proceeded to script its movement around the screen. He then drew a steering wheel and turned it into a movable object that would rotate like a steering wheel when you placed your mouse on it, clicked and then dragged. He then linked the behaviour of the car with the steering wheel.

Then he did something really cool. There was a little box in which he'd done the scripting of the car's behaviour. You could change the value variables and this would make the car do various things. He then proceeded to open up another box that described the variables of the first box. He then made the first box "drive" around the screen. He'd just demonstrated the essence of higher order anything.

He was given a standing ovation when he finished. It was the first one I'd ever been part of.

Afterwards Ben, Nathan Linger, others from Portland State University and I went to Steam Works, a micro brewery and restaurant, for dinner. I had an absolute ball with the Americans and managed not to embarass myself too much in front of them. At one point I castigated their country for inventing the franchise but I otherwise I behaved myself. Since it was a micro brewery it seemed only fitting that I try the beers. Man, these beers were weird! I tried a pumpkin and spiced beer and a honey lager. Bother were great although perhaps a little too sweet.

When I got home I practised my talk one more time before getting into bed and promptly having another freakin' bout of insomnia. One was not happy.

Day 4 - Wednesday

I gave my talk and it went much better than I expected. In fact, it was really well received. It's truly the case that my work probably surged ahead six months in a matter of three or four days. As the euphoria gradually wore off, and what a kick experiencing that was, I found myself feeling that I was starting to get tired of the whole conference experience. I'd absorbed a lot of the last few days and I felt it was enough - I now needed to digest it rather than imbibe more.

I went back to my hotel room and started finishing off some tutoring work that I'd promised I would. This meant that I missed post mortem discussion of GPCE. But second hand I heard that the Program Committee was happy to announce that there'd been a 100% acceptance rate from Australia. This wasn't saying all that much - there were only two of us - but it made me smile.

Despite the fact that I still had to write test cases for an assignment students were still struggling with back in Oz, that evening I went out to the social function that at the Vancouver aquarium. The highlight of that trip was definitely the Beluga whales. I've posted some footage of them in an early entry. I got home at a very reasonable hour and then proceeded to stay up until 5:00am writing the test cases.

Day 5 -Thursday

On Thursday I only attended one talk, partly because I didn't get up until 11:00am and partly because it was a short day anyway. As I said, I already felt that I'd had my fill of new material and I still hadn't seen Vancouver properly. This I managed to do in a large two and half hour walk. First I went to Gas Town, the Alt Stadt of Vancouver. It was also, as I discovered, adjacent to one of the rougher parts of the city. I then walked all the way to Stanley Park where apparently they filmed parts of Happy Gilmore. Good for them.

That evening I just chilled. I went for a gentle stroll to get dinner, and visited Doolin's Irish pub which is part of the hotel I was staying at. It had a great vibe and there were lots of people there without it being too full - I really dislike it when that happens. This was the first part of the trip I felt a little homesick. I was by myself and couldn't share the experience with anyone.

Insomnia struck again; this time I fell asleep at 1:30am.

Day 6 - Friday

Switching to the present tense now. I'm on my way to Seattle on a bus. The bus driver is a major dude, really friendly and very funny. A short while ago we went through U.S. customs. It was surprisingly straightforward and much like Australian customs - all forms of food and other agricultural products had to be declared. Still, it's a little intimidating - all officials are dressed in para-military-like uniforms and they don't smile all that much. Our luggage had to be taken off the bus, scanned, and put back on.

I've been enjoying the Northern U.S. geography out the window of the bus while I've been writing this. I just saw a bumper plate sticker on a car that said, "War is costly, peace is priceless." Not all Americans are war mongers.

It's time I signed off.