(First rule of blogging: sometimes you have to write that post you promised yourself you’d write, even if you’re exhausted and don’t feel like it.)
I can’t describe how genuinely good it felt to be back at OSCON. Quite possibly the best part was all the people that recognized me, and made sure to say how glad they were to see me again. It’s nice to feel like people enjoy having you around!
Started the day with the keynote talks, though I missed most of Allison Randal’s (as I was getting my registration packet and all). Of the talks, the high point (upheld by the amount of Twitter traffic it generated) was Tim O’Reilly’s announcements of http://opensourceforamerica.org and http://www.data.gov — both parts of the open government initiative he’s been participating in. The goal is to move the (U.S.) government towards both open source (to save on expenditures) and open data reporting, formats, etc. (to promote transparency and openness). While Tim cited that both the President and his designated “national CTO” were in strong support of the plan, he also feels as though if the program doesn’t prove itself within the first year, it won’t be able to survive or make a difference.
For the first round of sessions, I took in Jacinta Richardson’s comparison of Perl MVC frameworks, and Tim Bunce’s overview of Devel::NYTProf. Both sessions were excellent, and I was sure to rate both highly on the conference site. But Tim’s work on the profiler is just stunning. I was tempted to skip lunch in favor of playing around with the module. (But I didn’t… the food smelled far too heavenly.)
After lunch, I took some time to check out the vendors. There seems to be a lot of focus not just on cloud computing in general, but in making it as easily accessible and easily manageable as possible. Several vendors seemed to be competing in this space, whether offering their own complete stacks or just offering slick UIs on top of Amazon or Google clouds. I wandered back to the sessions rooms in time to catch a very nice presentation on sharing and serializing complex data with YAML. As a format, it is capable of a lot more than I have given it credit for. Clearly, I could be getting much more and better use out of it, if I can apply some of the things I learned.
After the last break, I took in two sessions. Or meant to, at least– the first of these was “Perl, UTF-8 and You”. I have a lot of interest in learning how to properly manage character encodings, but the planned speaker was apparently unable to make it, and the replacement person just stood up there and read the slides off to us. So I left and went ahead to the room where Allison Randal would be doing a talk on the Parrot VM. Much of it was material I had seen or read before, but much of it was new to me, as well, in particular the way she explained the register-based nature of Parrot, versus the stack-based nature of most existing hardware and virtual machines (including the JVM and .Net’s CLR). The explanation/comparisons of closures vs. continuations vs. coroutines was also very illuminating.
At this point, fatigue started to catch up to me strongly in the form of a sudden and powerful headache, so I reluctantly passed on the evening’s parties and BoF sessions. Tomorrow is another day, and better rest should help me get more out of it…