Perl Conference 5: Day 1
Sunday evening, I had dinner with Eric Raymond and Chris di Bona, two people I'd been meaning to meet up with for a long time; we had a good chat and an excellent Mexican meal down in San Diego old town. However, I had to get back early to finish off my preparation for my tutorials in the morning.
Monday was a very busy day for me - I was giving two tutorials, which meant six hours of talking. My first tutorial was about Unicode and Perl; around 30 people came to hear what Unicode was and how Perl supports it - I was excited to hear that a fair number of them were already using Unicode or were planning to do so immediately.
The afternoon session was a lot more fun - I was teaching Perl internals, with a class which really wanted to listen. This meant we could make the tutorial pretty interactive - I got asked a couple of surprising questions I didn't know the answer to, so I dived into the Perl sources and together we figured it out. Some people came away with ideas for things they wanted to hack on, and everyone came away with more confidence about diving into Perl's guts and knowing where to find what they're looking for.
In the evening, we had the perl5-porters meeting, a fairly busy town-hall style meeting of the Perl developers. People got up to talk about various things: Jarkko talked about what was new in 5.7.x and where 5.8 is going, and Sarathy talked about the maintainance of 5.6. (Once again trying - and failing - to give up the pumpkin!) Hugo, who Jarkko had nominated the 5.10 pumpking, talked about what was going on with the regular expression engine. Dan Sugalski and Nat both talked about Perl 6 - Nat from the project status point of view, and Dan from the internals design. Jeff Pinyan talked about proposed new regular expression features, and Jeff Okamoto asked us how we would like to support IPv6.
Finally, Adam Turoff kicked off the big debate - how we should "recommend" a set of modules for Perl users, and how we should define what goes into the Perl core. The main consensus should be that Perl should allow the user to select, just before installing itself, a set of "personalities" or "tasks", which define modules which it will then download from CPAN. Although we were pretty much agreed on what to do, there wasn't really an agreement on how to do it or on the details, so we decided to push back discussion onto perl5-porters.
It was a busy day for me, but definitely an enjoyable one. It's such an excellent feeling to sit down and talk face to face with people you've been working with for years; it's also great fun to spot your Open Source heros as they wander around and relax. I'm very much looking forward to tomorrow, when Mark-Jason Dominus is teaching all day on advanced Perl topics, and Larry Wall will be giving us his State of the Onion presentation - where we can expect something hopefully not quite as mind-blowing as last year, but pretty mind-blowing nevertheless...
Simon Cozens is an Open Source programmer and author.
Comments on this weblog
Return to weblogs.oreilly.com.