I am out at MountainWest RubyConf and so far, it’s been a great time. I didn’t take detailed notes, but I’d like to share a few highlights from the first day here. The talks have been great, the conversations even better, and we even have wifi + power!
The talks
Because these comments are sort of spotty and experiential, you should check out the speakers lists for a big picture of what talks went on. :)
Ara T. Howard - Ruby Queue
Ara’s talk melted my brain. There is something beautifully hacky about rq in that it’s super pragmatic. He had given a lightning talk about this at RubyConf and I’ve been anxious to try it out but with absolutely no application for it. It basically allows you to create massively parallel linux clusters without doing much of anything. Maybe I can build a mini cluster in my basement and start calculating large prime numbers or something. :)
Gregory Brown - Pragmatic Community Driven Development in Ruby
I gave a talk on pragmatic community driven development in ruby. That’s a long title. I renamed it “Rolling the Dice with Ruport”. I wasn’t sure how this talk would be received because it was mostly about my experiences with developing Ruby Reports and the community responses to various things we tried out. However, it turns out this is a common problem in Ruby and lots of people said my ramblings were helpful. Some even said it made them want to start their own projects, or get involved in some project they’ve been playing with. I purposely made the talk run short, to encourage discussion. There was no shortage of that, luckily. It was a great experience to see people get excited about a talk about community ,it shows that people do care about it.
Carl Youngblood - Simple Bayesian Networks with Ruby
Carl’s talk on Bayesian Network was another brain melter. Lots of juicy equations for math nerds like me, good primer on the fundamentals of Logics, and neat implementation of some Bayesian Network stuff. I think it might have scared the non CSish crowd, but hey, that’s a part of Ruby Conferences, isn’t it? :)
Alternative Implementers Panel
I missed part of the implementers panel due to hallway track discussions with Ara about RubyGems packaging edge cases. I hope to catch up with him and take some notes on this, because it’d be a good article for my Digging Deep series for this blog.
The panel was interesting as it brought folks together from JRuby, Rubinius, RubyCLR, and Cardinal.
It was interesting to see the different takes on things. I was one of the few people who actually have used &nil before, which surprised them :)
Jeff Barczewski - MasterView
MasterView seems interesting, I sort of wish Jeff would have put more examples up earlier in his talk, he gave a fairly comprehensive background of the project, which is tough to do in a short talk. He said it doesn’t depend on Rails, but he hasn’t worked on it with other frameworks. This blog entry will serve as a reminder to myself to check to see if there are any interesting integrations with Camping, but if someone wants to save me the effort and point me to some examples, that’d rock too :)
James Britt - Black-boxing with Ruby: Adding Ruby APIs and Front-ends to Existing Software
James Britt’s talk on blackboxing showed things that I literally need. The Tracula stuff is particularly interesting to me. One of his points was that mechanize and hpricot rock for adhoc ruby wrappers of websites. I can’t agree more.
Lightning Talks
The lightning talks rocked. I don’t have good notes on them at all except that CruiseControl.rb looks very cool, but not available as a gem. They mentioned it was because it was an application and not a library, but I wonder if spec.bindir is all they need. If I find better links to summaries of these talks, I’ll update the post here. Also, if you copied down the list of topics and names of speakers, please send them along.
Chad Fowler - Keynote
I’m not a big fan of keynotes. They always seem to be sort of meta-talks and something about them just rubs me the wrong way. Chad’s talk was actually pretty good though, and I think it’s important to talk about the things he did. He basically talked about where we might head now that Ruby is becoming increasingly mainstream. Striking the balance between maintaining the culture we have developed as well as responding to growth and change is a difficult process, and I think some of the points he made in his talk helped get the ideas flowing for folks. So you all can flame him, he said (perhaps just for reaction) that he thinks Ruby should have optional static typing. I think it’s funny because this brought up a mini holy war which I was regrettably a major part of on our back-channel. One thing I don’t agree with him on is that good Ruby is slowly emulating the works of DHH. That is true in certain domains, but I don’t think it’s a panacea, and I’m still a curmudgeon that wishes to de-emphasize the importance of the ‘Rails Effect’ on Ruby. I don’t mean that we shouldn’t recognize it, but that we shouldn’t be bound by it. Some of the things Chad said also seemed to support that kind of thinking, so I’m not sure what to think. I guess that these issues are just complex and need to go through the evolutionary process that’s inevitable in any growing community.
Other Conference Activities
A few people were interested in picking mine and Pat Eyler’s brain on user groups kind of stuff. We suggested a little get together after dinner, but after popping a balloon to silence the room at dinner time, our announcement got together a large group of folks interested in starting user groups or making improvements to them. We basically just shared experiences from our groups with each other, and really came up with a lot of good ideas. Yet again, I didn’t take notes. Notice a (bad) pattern here? I guess I should keep vim open at these conferences for more than just code!
I’ve also got myself involved in a couple projects. I might be putting together some reports against Tattle using Ruport, which could be cool for community demographics. I’m also hacking with James Britt on a quarky little ‘ri Method of the Day’ thing that we might hook into Growl.
I ended up going to bed early last night because I think I’m the only one on EDT out here. So that’s the end of the Day One thoughts. I’ll try to get a summary out later of the second day. :)
I'd love to see someone post a transcript of Chad's keynote and discuss the merits of optional static typing (along with the other points he brought up).
Personally I favor it, mainly for the multi-dispatch benefits (i.e. overloaded methods), not the type enforcement. But, I'd also prefer that it be implemented the way Evan Phoenix did it in Sydney, i.e. via Behaviors, where you would have to explicitly require an "overload" behavior in order to use optional static typing.
Others oppose it. One of the dangers, of course, is that "optional" static typing becomes mandatory in practice once the C/C++/Java programmers get ahold of it. Still others say we should use some form of type inferencing, which may be the best route, but is probably the most difficult to implement, and will likely be confusing to people who really aren't that familiar with type inferencing (like me).
Personally I favor it, mainly for the multi-dispatch benefits (i.e. overloaded methods), not the type enforcement
Didn't really think about this one, Dan. That would clean things up a bit, but idk, I think he was mostly just playing devil's advocate to prove a point :)
If a transcript shows up somewhere, and I hear about it, i'll update the article.