Tech Scene

Today was the Stack Overflow Dev Days Toronto event at the St. Lawrence Centre for the Arts, and all in all, it was a good day.  First of all, it was a one-day, $99-conference, and for that kind of money, I’m not necessarily expecting it to come out at the same standards as, say, QCon.

Registration
It started with registration, 8am – 9pm.  This was reasonably well-organized, and there were some muffins and coffee/water available, although both the muffins and coffee were average at best.

Joel Spolsky, Keynote
Joel talked about the paradox of choice and the tradeoff between the power and the simplicity of a software application.  This is ground he’s covered before elsewhere, but Joel is an entertaining speaker, and I’d be happy to hear him give the same talk again tomorrow.

Barry Gervin, Joey DeVilla: ASP.NET MVC
Barry Gervin from ObjectSharp did a live coding exercise with Joey DeVilla as backup.  Joey’s always good for some entertaining quotes, and it wasn’t a bad introduction to ASP.NET MVC.  Live coding exercises are risky, and I appreciate the effort that went into it.  It gave me a better understanding of ASP.NET and ASP.NET MVC, but I have to admit my mind started to wander after a bit — this is at least in part because I don’t expect to be doing any ASP.NET in the near future.

It was nice to hear that Microsoft has a free version of Visual Studio that is sufficient for doing all the ASP.NET work they did in the coding exercise.

Jordan Baker: Python
Jordan Baker was slightly less comfortable with public speaking than most of the speakers, but I think that with that in mind, he did a reasonable job of introducing python to people and walking through an interesting example of Peter Norvig’s 21-line probabilistic python spell-checker.  His explanations of the benefits of Python tended to cover well-trodden territory, and there are some trite bits in there that I disagree with, but if you haven’t already been exposed to these arguments, it’s probably good that you hear them here.

I wouldn’t have minded hearing a little more about Python 3.0 or seen a little more Django than the App-Engine Django that I’ve done myself, but there’s only so much time.

Joel Spolsky: Fogbugz
Joel covered some of the new changes in FogBugz 7.0; some of them bring FogBugz closer in line with a solution I’d use.  It was interesting to hear Joel invoke agile terminology several times in his description of how Fog Creek works, which seems a bit of  change from how he talked when he came to town for FogBugz 6.0.  He also showed off a code-review and web repository hosting solution using Mercurial, which was interesting to see.  It sounds like Fog Creek will be offering FogBugz-integrated mercurial hosting in the near future.

This is essentially an ad, so if you’re not interested in FogBugz, this could be a wasted session.  But as before, Joel is an entertaining speaker, and I’m happy to hear him talk about FogBugz.

Lunch
Sandwiches, wraps, orzo salad, green salad, some cookies, some pop, water and coffee.  Nothing really stand-out, but it was all edible, and free, so I think most people were satisfied with it.

Ralph Whitbeck: jQuery
This session would have been better if it weren’t right after lunch.  Ralph Whitbeck seems like he knows jQuery well enough, and his content was fine, but he was one of the less energetic speakers (and probably one of the less-experienced ones) and coupled with a little post-lunch sleepiness, I had a really hard time focusing, drifting in and out a few times.  Picking a strong and energetic speaker for the post-lunch session would probably have been a better fit.  Even swapping his time slot with one of the other presenters might have made the problems go away.

I like jQuery, and I hope this session helped other people to appreciate it, although I’m not sure that jQuery really needs the ‘help’.

Greg Wilson: Evidence-Based Software Studies
Easily the talk that got the most applause of the day, and the one I heard the most attendees talking about.  @gvwilson is a professor at U of T, who I’ve met before at the Developer Lunch that Krispy Uccello holds.  He argued that:

• Software arguments are often made with little to no evidence (e.g. Ruby makes me 50X more productive!)
• Software arguments are often made with citations that don’t bear up under examination (particularly if you consider the sample size of the study, and so forth)
• We should in fact be conducting far more research to gather evidence about the effectiveness of alternative software practices, and making arguments on the basis of strong studies — we should be experimenting and learning.
• As people who are likely to be exposed to these arguments, dig deeper, find out of the evidence is really there, and if it is, hang on to that evidence and use it to improve your craft.

He also touched on the fact that these problems extend beyond software, but that software is his area of focus.  The audience responded very well, both during the talk, and with long and sustained applause.  It was also my favorite session.

Reginald Braithwaite: Ruby
Reg / @raganwald gave a modified version of his rewrite ruby presentation, talking about metalinguistic programming and with specific examples in Ruby.  Reg is well-respected, had good content and is a reasonably good speaker, so this was an above-average session for me.  If you tend to read what he writes anyway, I’m not sure there would be anything very new for you here, but it was still worth watching.  Reg isn’t quite as polished as Greg Wilson or Joel Spolsky with the public speaking, but that also lends him additional authenticity.

Next Time
Wifi that works, and stays working.  Joel should know better.

Even more power strips.  Better coffee.  More challenging/interesting sessions.  Good public speakers, and slot the less-entertaining speakers mid-session.

Keep the cost low to help make it an affordable local tech conference.

In Summary
I enjoyed myself.  I’d go again.  If you’re interested, just set your expectations low, and hope to be pleasantly surprised.  Keep your eye out for some of the slides of the presentations on meta.stackexchange.com.