An Eventful Week

Lots of stuff going on, and I don't have time to go into all of it, so I'll just leave you with the highlights.

Moved into the new building at work, so now I have a 5 minute drive in the morning instead of a 30-45 minute drive, which is nice enough in and of itself, but on top of that I actually get an office. It's quite the change from being in a cubicle surrounded by the rest of the group I work with, but I think I'll adjust ;-)

Now I just need to finish learning my way around the new office. The place is quite nice, but it's also something of a maze. All the hallways look the same, and I'm constantly getting lost. I'm sure that will get better with time though.

In other news, my "Using the Apache Portable Runtime in a Non-HTTPD Application" talk was just accepted for ApacheCon, in Vegas in November. If you haven't guessed, I'm REALLY looking forward to this, the Apache guys were a lot of fun in Portland at OSCON, I can't imagine what they'll be like in Vegas ;-)

svn log -r 334

r334 | rooneg | 2004-08-21 20:40:15 -0400 (Sat, 21 Aug 2004) | 7 lines
Changed paths:
M /lucene4c/trunk/src/index/terminfos.c

Sometimes I'm really just too stupid to be allowed near a C compiler.

* src/index/terminfos.c
 (lcn_terminfos_idx_read): allocate the correct amount of memory for the
  index object.  this fixes the odd behavior i was seeing where additional
  allocation would stomp all over the fields in this object.

Why bother writing...

When others sum up the weekend so nicely.

Oh, and regardless of what Rob says, Alien versus Predator was still better than synchronized diving, which is what was on TV when we fled to the movies.

Damn

So it turns out that it's way too easy to do a "find -name \*.c -delete" when you intended to do a "find -name \*.o -delete".

Well, ok, there's no reason it should be easy to make that particular mistake, except that I'm a moron.

In related news, my "when do I check it in to version control" threshold has now officially been lowered to "there is any chance at all I might want this stuff later", since I clearly cannot be trusted to not accidentally delete several hours of work at any given time.

OSCON 2004 Day 6

The final day of OSCON was basically just the flight back.

I hung around the hotel in the morning, ran into a few people at breakfast, but basically I was just waiting to leave.

On the way to the airport I rode with one of the Postgres guys, although I never got his name. Had a fairly interesting conversation though, until he got caught in a conversation with someone who called his cell. I hate cell phones.

While waiting for the flight back I discovered that Casey West and Ricardo Signes were on the same flight. We talked about various OSCON related things while waiting for the flight to leave, and it seems that everyone had a great time at the conference. Casey showed me a new perl module he's working on for threading email, with a new algorithm that should be faster than the standard JWZ threading algorithm (it's 1 pass instead of 2 like JWZ's).

The movie on the plane was Jersey Girl, which kind of surprised me. Never thought I'd see a Kevin Smith movie on a plane, but I guess this one is more kid appropriate than the rest.

After the movie Casey and I took advantage of our airport cards and SubEthaEdit to pair program, working on some of the code for his new module. It worked quite well, despite the fact that we were sitting in separate parts of the plane. RJBS was too far back though, and kept getting cut off when he tried to connect to the ad-hoc network we had going.

I just barely made it to my next flight in chicago (15 whole minutes between flights!), and finally got home around 2AM.

It was a heck of a good conference, but I'm glad to be home.

OSCON 2004 Day 5

Started out this day totally stressed about the talk, so I skipped most of the first keynote and did a dry run in front of the mirror.

Wandered in most of the way through the first keynote, but I caught the end, which looked pretty cool. Others have already blogged it though, so I'll leave it to them.

Next was a talk from on of the Novell guys, which was theoretically interesting (it's neat to see a huge company like that making the switch to Linux on the desktop), but in practice felt like a sales pitch.

I decided to check out Paul's talk on Constructor Dependency Injection, since I'd been listening to him talk for much of the week, and he's a pretty neat guy. The talk was pretty cool, although the concepts were pretty simple when you come right down to it. Compared to the rest of the 'Enterprise Java' world it does seem like a good way to do things though, another of those things I'll have to keep in mind in case I'm ever in such an environment, like Groovy.

Now was my second talk, Subversion Best Practices. I think it went pretty well, although some people obviously were looking for something more like Greg's talk, an intro to Subversion rather than ideas on how to use it once you know how. I did have a number of people thank me afterwards though, and nobody told me it sucked (at least not to my face), so I guess it went about as well as I could expect.

Next up was the final keynote, with the head tech guy from Weta, the company that did the special effects for LOTR. They have an interesting set of problems (HUGE AMOUNTS OF DATA), and are being hit rather hard by RedHat's switch to per-machine fees. They need to use RedHat because it's one of the few versions of Linux that third party companies actually certify their apps on, but sooner or later they'll have to upgrade from the old RedHat version they're currently on, and the migration path looks like RedHat Enterprise Linux, but that has per-seat fees that would be a big problem for thousands of machines. On the other hand, he did have good things to say about Open Source in general, as they make use of a lot of it in house. My favorite quote was something along the lines of "Sometimes I wish I had an army of developers. Then I realize, I do have an army of developers". It's easy to forget that open source is just that, an army of developers, if you know how to use it.

After the keynote John Peacock, Justin, another Apache guy, and I went out to lunch at a seafood place near the hotel. It was good, both the conversation and the food. Then we wandered off to Powells to do the standard 'drool over the massive collection of techie books', and scattered off to our respective hotels.

I spent the rest of the afternoon taking pictures of the Red Bull Flugtog, that crazy thing where they launch home made man powered flying machines into the river. The event itself was saturday (after I was leaving), but I did get some good shots of the machines themselves as they were setting them up. I'll post them once I get a chance to get them off the camera.

Ironically this was the only time during the week I actually took any pictures. I love my camera, but it's just too big to comfortably cary around during the conference, at least if you're going to also carry a backpack and a laptop... I see one of those sweet little Canon Powershot cameras in my future, so I'll have both ends of the spectrum covered with regard to digital cameras.

In any event, I went to bed fairly early, since it had been one hell of a long week at that point.

OSCON 2004 Day 4

Ok, so I got distracted at the end of OSCON and didn't get around to actually blogging any of it in real time.

Here's thursday...

The keynotes were the Dysons (one of whom was actually the guy who came up with the Dyson Sphere, they actually showed a clip from that episode of Star Trek at the beginning of the talk), and Bdale Garbee, of HP and Debian. The Dysons were ok (I thought George Dyson's talk last year was better), but Bdale was quite good. He interspersed the standard open source stuff with bits about his experiences building amature radio satellites. It's nice to know that there are people who are realistically talking about putting a satellite in orbit around mars, without government funding.

Saw a talk on qpsmtpd, the Perl SMTPD. Cool app. Must look into it closer.

Andy Lester and Bill Odom gave a fantastic talk on Advice for Open Source Job Seekers. Not that I'm actively looking for a job, but things to keep in mind when one is looking are always good to know. One thing they mentioned which I should really do is to keep your resume up to date even if you're not looking, so you don't have to scramble if some kick ass job opportunity comes up and you need a resume. They also stressed the idea of building up your online reputation by doing cool stuff, so people are more likely to have a clue who you are when you call them up and ask for a job.

The talk on Wackamole, an open source high availability package, was quite informative. I will have to look at this closer. I was particularly amused by the program you use to control your Wackamole installation, wackatrl (pronounced, of course, "Whack a Troll").

Nicholas Clarck gave us a view inside the Perl 5 Porters with Perl 5.8.5 Was Boring, and in the process also gave a lot of useful tips on how to keep the stable releases of a software package actually stable and useful. I also managed to run into Peter Prymmer (the other $COMPANY person at OSCON) in this talk.

Next I went to the What Book Sales Tell Us About the State of the Tech Industry talk, by Tim O'Reilly and some of his hackers. It was interesting enough, but it strikes me that this info will be much more useful next year once they've done more with it. Key things (like linking in geographic data) are still missing.

Finally was the python lightning talks, which were supposed to feature Dan Sugalski being hit in the face with a pie for losing his bet over the speed at which Parrot would run Python code. Unfortunately Guido decided not to hold Dan to the bet. Amusingly it was the Perl people in the room who were most disappointed about this. The remaining talks were not overly memorable, I think the Python community needs to learn a bit more from the Perl community on how to do lightning talks. It's just not the same without the gong.

I've totally forgotten what I did for dinner on thursday, but eventually we wandered back to the hotel for the Novell "Free as in Beer" party. I skipped out after one drink, as I knew my talk was in the morning and I didn't want to be overly dead on my feet.

Unfortunately it turns out that the Perl Foundation auction was also thursday night. If I'd known I totally would have gone to it, but they had like no publicity so I missed out. It's too bad, I would have been willing to pay way more than 300 bucks for an hour of Pair Programming with Ward Cunningham ;-)