Joel Berger

Good morning Chicago. I’m back from a wonderful trip to YAPC::NA 2012 and while its nice to be home, I’m really sad that the conference is over. It was my first YAPC and I’m sure that it will not be my last.

Before I get to my reflections, I want to say that the job that JT Smith(blog post picked because I like it), the MadMongers and the many volunteers and UW-Madison folks who made it all happen. I was so well organized and run, I’m astonished.

Moving on.

I started out a little starstruck. I haven’t been a programmer for that long (my first Perl scripts are dated 2009). I haven’t been around to watch the community develop, so I learned the names of major Perl contributors/community members from the CPAN modules etc. In my mind they sprung from the earth as fully-formed Olympian gods of programmers. It took until about lunchtime before it struck me that they all were really down to earth people and I started talking to as many people as I could. Pretty soon I was having lunch with Andy Lester, Sinan Unur, and Gabor Szabo among other perhaps less well-known, but no less interesting Perl users. I love that this conference gave me these opportunities! SawyerX is totally awesome! MST is hilarious (though I didn’t talk to him in person) and Larry even stopped by our impromptu “Perl Scientists” gathering the first night. I’m sure I’m missing some names that I could drop, but I think you get the point: the big names were there and were as great of people as they are coders/contributors/teachers/etc.

I learned some great things too. I saw Devel::NYTProf in action for the first time and I think I might see if it could speed up some of my more, lets say, “interesting” scientific scripts. I am going to add Data::Printer “awareness” to some of my classes, and I’m going use the concept of a specially-named-semi-private-method-for-awareness the next time I think about the interoperability of different modules (if that doesn’t make sense, look into the _data_printer filter option). I think I have a better handle on when to use Moose Roles. I’m excited about new Perl features (__SUB__++).

I even have a new project; a tryruby-like web-based “Try PDL”, which might-or-might-not use vti’s showmetheshell and might-or-might-not be hosted on Stackato (and even if it isn’t I learned about LXC for sandboxing). Would more people use PDL/Perl for science if they could play with it before installing (the much improved, but still rather tricky to install) PDL module itself? I also need to make more websites for my modules (as mentioned in the last lightning talk session).

Finally, I gave two talks myself and was so excited about the attendance and reactions. The first was about my reseach, and more specifically the methodology that I use to make modeling easier both for the module-author (in this case: me) and the script-author/user (me and others). The room (235, the best room IMO) was full, probably about 30 or so people and they seemed very interested and engaged. Reini Urban even paid me a great complement in front of the whole room (!!!). The full model is available here.

Then I (immediately, though in another room) gave a how-to-get-started talk on XS. I’m glad I didn’t have more time to worry, because the room was bigger and full too, I don’t want to guess how many attendees there were, but it was a lot. I really hope people learned and will be able to use this knowledge as well. The feedback I got was very positive.

So thanks to all of you who attended my talks! You really made my conference even better! The links above contain the PDFs of my slides. If anyone has any questions or would like to hear more about my research, please contact me. (As an added bonus, the LaTeX-Beamer source used to make them is also there! The PGF/TikZ/Beamer modules have started a similar renaissance in the LaTeX world like Moose etc have started in Perl FYI.)

I guess this has become a really long post, but I had such a great time. I can’t wait to relive it, and get even more knowledge (!), once the the recorded talks are posted.

Cheers!