journal for 2003-07-08

oscon: day the second

Today was much better in general. I knew some people, I had some plans, and there was more to do. I’ll try to summarize well.


I ate this morning! I got up as early as yesterday and got the same things done. Then I headed downstairs. I saw Clint; I’d sat near him yesterday at dinner, so I said hello and then wandered over to the coffee bar and got an apple juice and a maple scone. Both were OK, although they were well overpriced at $4 each.

On a side-note, I’m not out of cash. I don’t want to use the hotel ATM, as it charges two fifty. I’ll need to go ATM hunting tomorrow.

Clint and I talked for a while. He finally got a registration and was going to be at the Damian’s Advanced OO talk. He said MST3K last night went really well and urged me to come to the next one. I said I’d do so.

After a while I headed down to the Rendezvous lounge and got on IRC to talk to Gloria. I checked with work, too. There’s stuff to do, but nothing that I’d be able to do right then. I have, in fact, not gotten to it yet. I should be able to tomorrow, when the day isn’t non-stop convention stuff. I think. A real workstation would help, too. I’d like to get a real mouse, which would be a huge step toward always-usefulness of knave. Clint has a really slick USB wireless. The USB dongle is stored inside the mouse. Removing it turns the mouse on. Very slick. I’d really like a Bluetooth mouse, but OS X doesn’t support them yet. Rumor has it that 10.3 will, so I’ll see what happens. I’m not in a big rush to get a mouse, especially one that will probably be expensive. I’m getting relatively proficient with the trackpad, right now.

After some talking and almost working, it was time for my first tutorial.

programming the apache lifecycle

This talk was much better than the mod_perl2 talk. While there was less actual new information, it was so much better presented that it just Won. In fact, this talk really helped me assimilate a lot of the information that I’d received and not processed yesterday.

I am filled with an urge to engineer things that use mod_perl. It is just so incredibly cool. I knew how much it could do, before, but now I feel like I really know how much it can do. The presenter (Geoff Young) said, roughly, “People sometimes say mod_perl is just faster CGI, or compare it to PHP. That’s just wrong, wrong, wrong. mod_perl is a complete Perl interface to the Apache API.” It fills me with a sense of wonder and playfulness.

I need to putz around with it; maybe I’ll make Debug a handler, like it really should be!

If, as Geoff said, most of the stuff he showed from mod_perl1 will really work in mod_perl2 with minor weaks, I feel well-armed to try to do some serious puttering.

During tea time, I ran into (among other people) Clint. He said he was recording the Advanced OO talk, and I said I was eager to get the recording. I hope to get it tomorrow. From what he (among other people) said, the talk was awesome and was not a rehash of the Damian’s book, which is what I’d feared it would be.

Toward the end of the mod_perl talk, I started to zone out a little, but not much. Geoff, like everyone, was mildly horrified by the lack of snacks at tea time. (Apparently previous cons have had them. I certainly expected them.) In reaction, he brought snacks of his own to distribute. I had a tiny Milky Way bar. “We presenters depend on the half-time break sugar rush to keep you audience members interested!” he said.


I met up with Clint, Joshua, Michel, and a pile of other people for lunch. We headed over to Quiznos. I’d been told that Quiznos was many things: “Subway++”, “Subway, but with good food and a toaster”, and “Once you eat Quiznos, you will never want Subway again.” So, I expected it to be a lot like Subway.

It was, and it was better. It didn’t change my life. I got a mesquite chicken and bacon sub. It was “regular” size, and it was slightly too much food. I continued to talk shop (and otherwise) with other geeks and then headed back to the hotel for my next tutorial.

programming SVG

Woah. So, first of all, the speaker was using Apple Keynote. It looked quite sweet! If I had to give more presentations, I’d consider it. I don’t, though, so it’d be $100 down the drain. Anyway, I saw some other free presentation software that looked serviceable. Keynote did this neat 3D surface rotation to change slides. It looked like what I expect Panther’s “fast user switching” to look like.

This talk was really up and down. In the end, it was way cool, but it had a lot of stops along the way at places like Boringville and Dronetown. For the most part, though, it was good.

Antoine (the speaker) spent too much time talking about stuff I already knew. This was annoying, but I know he had to do it because a lot of people weren’t familiar with the DOM and its methods. Still! Of course, I can’t expect many of these talks to be geared exactly to me, can I? Well, I can, but it doesn’t lead to much satisfaction.

He started with a cool flashy demo. (This is an amusing pun, because SVG can largely replace Flash.) It had stuff resizing and showing off textwrapping while playing video in a non-rectangular box with a translucent frame. Nice. Then he showed some stuff I think I could actually manage without rereading his notes a hundred times: simple boxes and text and ECMAscript.

The usefulness of SVG for doing things like SPC and live data-graphing is just off the chart. I need to talk to the engineers about this when I get back. It looks awesome. It could save us incredible amounts of time and just be Great. Of course, I doubt the engineers will want to learn it. Oh, the need we have for another developer! Meh.

Later, he busted out a demo of dynamically resizeable objects; the user could click on them to get handles, then drag those to resize. It was slick. He also showed off some stuff that allowed text to be freely rotated and still selected, copied, and pasted. Slick!


After SVG, I had a break until 1900, when the States of the Union would begin. I went down to the Rendezvous lounge and did a little socializing, mostly with Gloria. I also saw some nonsensical flooding going on on the Freenode OSCON channel and got ops, largely (I think) because I was sitting across from the only op there and because I knew how to use ChanServ. Woo!

I borrowed Clint’s iSight to record Larry’s remarks and then took the lift up to my room to drop off some crap. I ran into a guy in the elevator who was looking for notes on some of the toots I took, and I stopped to give him URLs. He asked for my notes from Stas’s mod_perl2 course, and I was happy to give them to him, but I’d dumped them in the trash this morning and housekeeping had already picked them up. Oops!

I called Gloria and we talked for a little while, but soon it was time to get to the ballroom for States.

states of the union

I got iSight set up, mostly, and it is freakin’ sweet. I’m a little disappointed that only iChat seems prepared to use it for much, but it was nice. I’ll try to find out, later, if iMovie or anything else can use it. I think I screwed up the first few minutes of recording, but most of the speech seems to have been recorded alright. I only got audio.

Perl (State of the Onion)

Larry’s speech was awesome, as expected. It was largely vague and motivational, but it was very clever and entertaining, and I think it really captured the Perl Attitude. It was a definite privelige to be here for it.

The biggest thing he explained was the migration-to-Perl6 strategy. Perl6 has a grammar written in Perl6 regex, and it runs on the Parrot VM. Since Perl6 needs Perl6, there’s a chicken-and-egg problem. The solution is to write Perl6’s bootstrapping grammar in Perl5, to which Perl6 regexen can be backported. They’ll be backported to Perl5 running on the Parrot VM, and this Perl5-on-Parrot project is being called Ponie. (“I want a ponie!” cried Larry.)

I hadn’t realized that Perl will be faster on Parrot, not slower, but obviously that’s the case. This means that Perl 5.10 will probably exist in C and as Ponie. Then, if Perl 5.12 must happen, it will only be Ponie. Hopefully by then Perl 6.0 will be available. Who knows!

other states

After Larry there were other States:

Guido’s “State of the Snake” on Python was really uninteresting to me. The material was OK, but I’m not a Python programmer. Guido himself was relatively dull. He spent a huge amount of his time talking about the decision to continue to exclude a ternary comparison expression from Python. He didn’t go into technical detail, so I’m not sure why this wasn’t just a one line mention. Other than that, nothing interested me.

Then was PHP. The PHP talk had some technical stuff, but mostly it didn’t interest me, either. The speaker was mediocre, and I just don’t use PHP. At one point, though, he said, “PHP is still that dirty little web language we all love.” He said it would stay that way. I’m glad he can admit it. It’s not a shameful thing. Still, with all the mod_perl fu I picked up, I feel even less likely than before to want to play with PHP again!

After PHP was MySQL. This talk interested me even less. I can’t remember much of anything they said. They were talking about how great MySQL is and their goofy dual-licensing scheme and various hard-to-swallow benchmarks. They encouraged people who want ANSI SQL procedures to be in touch with the dev team to help encourage and aid the development of them.

After MySQL was Apache, which was the worst yet. He just droned on and on about how many employees Apache had, and how many CVS committers, and how much money, and how many servers. It was pointless. Every single person in the room knew that Apache was a big project and that it was growing and that it was good. Why didn’t he tell us the state of the project? It was frustrating and boring, and it pissed me off.

state of the penguin

Ted Ts’o gave the awesome State of the Penguin address about Linux. It was just great. He mostly talked about the 2.6 kernel, which sounds fantastic. I’ll have to pull my notes out of IRC and put them on my OSCON page. It had crazy new things, much of which deals with serious improvements to real core kernel operations: the new scheduler, user-mode Linux, filesystem enhancements, ACPI and ALSA. Wow!

He also got some digs in at Microsoft (“NTFS support is much improved, if you’re in an environment with both Linux and legacy operating systems”) and ESR (“You’ve heard a lot about ESR’s new kernel build system. Well, ESR is very good at promoting his projects and himself. We’re not using it.”). The new kernel build system sounds great, although I was fine with menuconfig and the exisiting Makefiles.

the ActiveState party

After the States, ActiveState was providing free food and beer down the street. I stopped in, just to see what it was like. I wandered over with a huge group of people, chatting with Michel along the way. ESR himself showed up just as we were leaving, looking like shit (like usual) and being somewhat shunned. I thought it was kind of funny, and also sad. He was wearing a “unix is sexy” shirt.

The place was packed. It took minutes to get across the tiny upstairs room toward the buffet. I had some chicken wings, a gyoza, some cantaloup, and a mozerella stick. It was more than I needed. I haven’t worked out, still. At least I’m not eating much. I wish I had a scale.

I had a beer, mingled a little, and then shoved off. I gave Michel the rest of my Free (as in Beer) tickets so that they wouldn’t go to waste. The place was just too crowded for me, and I’m not big on giant groups and beer, anyway.

wrapping up

I got back and sat around for a while, chatting with mdxi and solios about geeky things. mdxi and I explained to solios that C is written in C, and that this should seem natural. I listened to iTunes and worked on this entry. Eventually, security kicked me out to the lobby (from the basement) so they could close the basement up. So, I’m sitting in the lobby now, really ready for sleep. Knave’s battery is near death.

I’ll save this and be done, then go turn in. Tomorrow: a few keynote addresses, Perl 6, lightning talks, Parrot, maybe some Subversion, maybe some DBI stuff, and the SCO vs IBM talk.

Written on July 8, 2003