Well, I got home last night after a day full of travel. We left the University of Houston around nine o'clock in the morning and I got in my front door around eleven thirty at night, almost ready to collapse into sleep.
Before leaving home on Sunday, I was a little worried about this YAPC. The general impression I'd gotten beforehand wasn't good: the ACT conference toolkit used to manage the site is still under heavy development, and it showed. There was a lack of communication about things that I thought was fundamental. Things kept getting sorted out just a little too late for my comfort.
In the end, there was a bit more confusion than usual, but it didn't matter. This was a great conference! Everything that mattered was really well done. There was great food (except for the traditionally mediocre and overpriced opening dinner), a good venue, good talks, and, of course, loads of really great people. A big topic of discussion, this year, was the fact that Perl's role behind the scenes of Absolutely Everything isn't very well published, leading to some companies keeping it on hand as their secret weapon. This leads to a lot of Perl People who are stuck hiding out, never sent to conferences on behalf of their employers. It's a shame: there's so much to be gained for everyone from a conference like YAPC. It's one of the few tech conferences I've attended where there are equal, large measures of fun and professional benefit.
Every year, I write less and less about what I got up to, largely because I'm busier and busier at YAPC, talking to more people about more and more interesting things. This year, the trend continued. I attended three excellent group meetings: the YAPC BOF (for people interested in hosting future Perl conferences), the PEP BOF (for the Perl Email Project, which was a great success, I think), and the Mason BOF, which has gotten me more excited about the prospect of a template-only Mason release in the future. I made a brief appearance at the Beers of the World BOF, but it seemed likely to grow out of proportion, so I ducked out. I love a few beers with a few friends, but drinking parties really aren't my thing.
I gave two fifty-minute talks, both of which went fairly well. I didn't have enough time to rehearse my email talk, so I didn't get quite as far through it was I would have liked. (I generally plan on having more material than I need, but with that talk, I didn't get through all the slides I thought I was sure to.) My Sub::Exporter talk went well, I think, despite being interrupted by a fire alarm. My first reaction was to just talk louder (having first asked the audience if they wanted to flee), but we got kicked out pretty quickly. I had a few people tell me, after the Sub::Exporter talk, that "this looked fantastically useful, but I'm not sure I really understood everything." This was my goal, more or less: I wanted to get people to see a lot of what it could do, so they would then check out Sub::Exporter::Tutorial and play around with it on their own. I hope this was a wise plan. I'd love to hear back from any attendees on that presentation.
I also gave a lightning talk on a domain-specific langauge I was cobbling together with the Parrot compiler tools. I had been thinking about doing this with a Perl 5 module, but on Monday I attended a talk by Patrick Michaud on the Parrot compiler tools, and I was immediately sold on them. They looked extremely powerful and easy to use, and so far they have been. I think that the hype may be true: the best thing to come out of Perl 6 may not be Perl 6, but rather Parrot.
There were some other great lightning talks, including a really fun talk about teaching a three year-old child to program Perl in Japanese, a silent talk given by Jason Crome (on the subject of hilarious fake module names, a favorite topic of mine), and a neat little demo of a tool written in Perl/Tk by Pip. (I hope he releases it, and that he does not give it a Time::PT version.) Scrottie showed off a little-looking continuations-based web framework, which I'll certainly want to look into.
I had some good barbeque (including BBQ duck -- who knew?) and some great beer. If you're in range, I heartily suggest you try Boulevard's Zon or Lunar beers, which were donated to the PEP crew by wirebird.com.
The network was lousy, which is to say that it was worlds better than previous YAPCs, which were atrocious. It wasn't quite good enough to support a video chat back home, but I managed to have a video-over-internet-but-audio-over-cellphone chat with Gloria, in which I got to look in on her and Martha. That was nice, but it would've been nicer to have done so more often.
I spent about fifty $2 bills in Houston, and I'm hoping to see a few WG hits in the next few weeks. Someone tried to give me change for a twenty when I paid with a two.
All my photos are up on Flickr, tagged with yapcna2007. I'll probably add more tags and descriptions and notes over the next week or so, but you can still see my whole travelogue.
It was a great YAPC. I look forward to next year's YAPC, and to any upcoming workshops I can get to.
Re:Continuity/YAPC
rjbs on 2007-06-30T17:57:31
I'm glad the email talk was not boring. I knew that was a serious risk.
Any time someone says "gopher" my interest is reawakened. I don't know why I like gopher. It sucks so much! I guess for the same reason that I like the Transformers cartoon: nostalgia.Re:Continuity/YAPC
scrottie on 2007-06-30T18:59:28
HTTP became quite complex, and HTML is primarily a language of kiddies making loud and obnoxious pages, and self-important companies making bloated, complex, machine and user unfriendly abominations.
I think Gopher appeals to me for the reason most people abhor it -- it has no provisions for making things pretty or controlling their presentation at all.
If I never saw 0 pixel tables surrounding content to control the document gutter or image cutups with a 100 columns for a patchwork of graphics and text, I'd be happy.
Gopher reminds me how nice it is to just have things work, and to simply not reach for that "higher" level of trying to do something so fancy, difficult, and contrived that it impresses people who don't understand what horrors went into making it impressive.
Back when HTML pages tended to have one image, it was fine. Animated gifs could at least be stopped with the Stop button in NS2/NS3 (God I miss that feature). Status bar scrollers, pop-ups, alerts... less so. And now a page can soak up all of your memory and CPU, just to display some information, ultimately. I can't even think of HTML as hypertext any more... it's now a terrible application programming language that just happens to be easy to start in in that constants on the language print themselves (text stays text). That there is a hypertext aspect to HTML is primarily reflected in image hotlinking but otherwise lost in the pile of features and misfeatures.
-scott
The general impression I'd gotten beforehand wasn't good: the ACT conference toolkit used to manage the site is still under heavy development, and it showed.
Re:Re ACT
rjbs on 2007-06-30T00:28:05
Yes, I spent time on the channel telling you about specific problems as I found them. Then, at YAPC, I made sure to talk to conference organizers and TPF reps suggesting that the attendees, who had mumbled on IRC and in person, send in all complaints.
What the hell do you want from my personal blog? Do I have to enumerate every flaw there, too? Or am I just not allowed to summarize my conference experience? I've tried to do what I can to help, and I thought that was clear.
