島根 September — RubyWorld

The reason I went to Japan was to attend RubyWorld 2009, a conference centering around the Ruby programming language. The reason the conference was in this rather remote corner of Japan is that Matsumoto Yukihiro (known to geeks as “Matz”), the inventor of Ruby, chooses to live and work in Matsue, and the city and prefecture governments are trying to use his presence as leverage in diversifying the local economy into Open-Source software. While the conference was reasonably well-covered in the Japanese media, I’ve seen nothing in English, so I’ll try to rectify that here. With pictures.

(For context, see 島根 September — The Castle.)

Oh yes, the Japanese media; here’s one of them doing a video interview with Matz; they both seeemed to enjoy it.

Did I say Matsue was remote? There are places in Japan that are much further from Tokyo, but Shimane is definitely remote mentally; a recent poll found that of all Japan’s prefectures, Shimane was the least cared-about.

For a couple of years now, there’s been a “Ruby City Matsue Project” with offices, pictured below, just outside the train station. I don’t know how well that’s going.

Before I dive in, I have to send one more big huge domo arigato! to all the people who took care of us on the trip; not just the conference organizers, but Masaki-san from JTB and the Shimane University people who equipped us with students of English at the socials to help us get along.

It Was a Success · No matter how you measure it. They drew 590 people on the first day, 502 on the second, which I believe makes it the largest Ruby-centric event ever held in Japan, and a triumph given the location. Lots of people came from Tokyo, and I met others from Hiroshima, Gifu, and Fukuoka.

They had solid public-sector backing; at the opening ceremony there were speeches from Mayor Matsuura of Matsue city (pictured below) and Governor Mizoguchi of Shimane Prefecture, as well as the polished and impressive Director-General Ishiguro from METI in Tokyo.

Finally, the content was good. Matz is always interesting, the list of talks hit some important bases, and they flew in a bunch of gaijin to broaden coverage. This included Jeremy Kemper on Rails, Bruce Tate on team-building, Charles Nutter and Tom Enebo on JRuby, Evan Phoenix on Rubinius, and Stephen Kong of Shanghai on Rails to talk about what’s up in China. They had simultaneous translation on one of the two tracks, so I could follow along with the Japanese speakers.

I didn’t make all the speeches, since there were two tracks and I had to do some media work, but here are the impressions I took away from the ones Iw was at.

The Opening · The welcome speeches from Matz and the politicians were polite and polished and fairly content-free, with a couple of exceptions: I can’t remember whether it was the Mayor or the Governor who emphasized that in Matsue, around 5PM people knock off work and go home, which is notably unlike Tokyo. I think those familiar with Tokyo work culture will suspect that there’s no loss of productivity involved in the Shimane approach.

Second, someone pointed out that there’s a nuclear power plant right near Shimane, so electricity is cheap, plentiful, and carries a reasonably low carbon burden.

Speeches: Mr. Ishiguro of METI · He pointed out that Japan, a world-beater in so many sectors, has been mostly absent in the world of software. Since Japan knows all about manufacturing excellence, the obvious conclusion is that building software isn’t like manufacturing. Can’t argue with that.

He touched on the recent discussion of the failure of Japanese phone technology to travel offshore — people talk of a Galápagos syndrome — and suggested that one answer for both software and phones may lie in a more standards-based approach. It was an engaging and fact-filled presentation; in fact I have rarely seen denser PowerPoints.

Me · I’ll try to cover my material in a blog post here before too long. I will say that I mentioned Twitter (quite a lot), PDML-twit, Scala, PubSubHubbub, concurrency, threads, Functional Programming, Erlang, the MVM research currently going on in Tokyo and supported financially by Sun, my Android explorations, Ravelry, REST, Rack, the mobile-phone market, Wikipedia, and boring.com.

I had 93 (!) illustrations and several locals told me the simultaneous translation was very good, so at least I suspect they weren’t bored. Well, the businesspeople probably had trouble with the Erlang code.

Standardization · The Japanese Ruby community is marching ahead toward building some sort of Ruby standard in JIS and then perhaps taking it to ISO.

They think that this will make it easy for large public and private-sector Japanese organizations to adopt Ruby. The effort makes me nervous because I have intimate first-hand knowledge about how much work it is and how many ways there are for it to go off the rails. But if they’re right about the effect on adoption, more power to ’em.

Here’s a picture of a panel on the subject, with Sun’s own Takashi Shitamichi in the middle and Matz closest to the camera.

Matz’s Keynote · They defy summarization; just go to a conference where he’s speaking sometime and take in the experience yourself.

He gave a capsule history of programming, argued that we’re in the golden age, lamented the pervasiveness world-wide of lousy pay and working conditions, and made a general appeal for the integration of work and love.

On Rails · This was by Jeremy Kemper who works for 37signals, on their apps of course, but his fingerprints are all over Ruby on Rails.

I’ve been a bit out of touch but was super-impressed by his coverage of the currently-in-progress Rails 3; it’s been substantially re-architected to make it a little less opinionated on certain subjects. Also, it runs pleasingly faster than any current Rails when combined with the also-still-in-progress Ruby 1.9.

Amusingly, Jeremy was right on the 37signals party line in asserting that the speedup work was just for intellectual pleasure, because “Rails is fast enough”. For them, it is.

I hadn’t heard Jeremy speak before, but now I’ll go out of my way to do so; his presentation was graceful and stylish as well as being full of good hard information.

Ko1 on Ruby.next · “Ko1” is Sasada Koichi, currently pretty busy with his day job as a U.Tokyo professor, but the primary author of much of Ruby 1.9, and still involved with its development.

1.9 continues to move along nicely, and is showing signs of stabilization, with snapshots due at the next RubyConf and at Christmas. This is a good thing, because once the word gets around that Rails runs faster on it, there’ll be a flood of adoptions; it seems there are some people in the world who think that Rails isn’t always fast enough.

Could It Work? · I mean, could they really succeed in building a hotbed of open-source software around Matsue? It’s certainly not impossible. If there’s a parallel, perhaps it’s with Portland, Oregon: Green, a bit remote, high quality of life, lots of cheap electricity.

This was my first venture much outside of Tokyo and, as I think this series of blog posts should suggest, Matsue and Shimane are pretty easy to like, at least on first acquaintance.

What would I do if I were them? I’d probably focus more on the local University, trying to build up the CompSci department by recruiting some hotshot faculty. That, along with some judicious tax incentives and keeping the quality-of-life drumbeat loud and steady, could go a long way. Your average Tokyo commute is around an hour; Matz’s is like ten minutes and there are lots of people who are closer in than he is.

Schmoozing and Closing · There were parties and, this being Japan, there was lots of good food and drink at the parties. I got to hang briefly with Shimane Gov. Mizoguchi, the first time I’ve ever met a Governor of anything, as well as various assorted consultants from big companies and tiny Rails consultancies. Business for the latter is not great but not bad, which is actually pretty good given the current economic situation.

To close this out, here are pictures of two guys who I suspect may have done as much work as anyone and more than most to pull the whole thing together. First is Maeda Shugo; he was listed as “Secretary general of Executive Committee” for the conference, which I take to mean “Guy who did tons of work”. Thanks Shugo!

Finally, here is Inoue Hiroshi, CEO of NaCL, a Matsue-based company which among other things employs Matz. Inoue-san is a charming host, an expert aficionado of Pentax lenses, a good businessman, and nominally Matz’s boss. Which is to say, a very special person, and one to whom I think the RubyWorld 2009 conference owes extra thanks.

They’re going to run it again, and if you’re anywhere near, I’d go if I were you.

---

---

Contributions

---