Saturday I went up to the North Seattle Donor Center for the Bronies for Good Reboots Nurse Redheart's Roundup. That's me with the long hair, obviously.

A quick bit of background -- I've always been pretty into classical music. My mom introduced me to Beethoven, and hearing Bach's Die Kunst der Fugue pretty much sealed it for me, although it wasn't until much later that I started liking Stravinsky, Liszt, Bartok, etc. -- but it wasn't until very recently (two years ago?) that I started getting into metal, and that happened pretty much entirely because I started playing the guitar myself and they tend to be at the cutting edge of modern guitar technique in a lot of ways.

In my constant refactoring of Rice, I decided to break from Wheat in another way by changing the format of post metadata. Instead of plaintext key value pairs of the form key: value, I decided to just stick a JSON object at the start of the file.

I ran across a very interesting blog post while browsing Hacker News the other day. The author goes over one of his favorite intervew questions to pose, which is a really fun problem, because it's fairly simple, but has a lot of hidden complexities in it. Here's his actual definition of the problem:

If you're a fan of minimal desktop environments like I am, you may not want to install all of the dependencies for the VirtualBox GUI -- QT specifically. Not to fear, there are command line ways of creating, managing, and running VMs.

So one of the things I was concerned about was the fact that I sorted the index of articles to generate the home page, and then called array.reverse() on the resulting array. Well, it turns out it's better than the alternative of parsing the date string, according to this test I ran on jsperf.com.

So now that the initial featureset is complete, I wonder what step I should take next? Do I add tests for behaviour? Do I refactor the existing code? Or do I work on adding the missing features?

Today, I (re)-started my first real Javascript/NodeJS project. I'm calling it Rice.

So obviously the first part of recreating Wheat is understanding just what the hell Wheat does. Unfortunately for me, it seems the Git integration part of Wheat is the most complicated by far. I'm currently reading through Caswell's code following the nested function calls through various files and trying to figure out what exactly is going on in order to read a single file. I know the main file, wheat.js calls Git(process.cwd()), then something happens, and then Renderers.article, for example, is called, and then Data.fullArticle.

I've spent enough time recently installing and configuring a brand new Arch Linux system to be my development environment that I decided I might as well make a blog post of it.

Today, I have to implement a version of Hangman, without using the ACM Graphics package. That's about it. Everything else is flexible. Yay.

This assignment, I have to make a binary search tree class for a game of 20 questions (though not necessarily with a depth of 20), that can do the following:

If you're hacking around building a NodeJS server in Vim, the following (extremely dangerous, don't say I didn't warn you) keybindings in your .vimrc might help:

So right now I'm giving nvm a try, which is the node version manager.

For assignment #5 I'm supposed to make a game called (lazily enough) the Assassins Game. Basically, you have a list of random people, who have a random target, and a random person who is targetting them. They all kill eachother until only one person is remaining, and that person is the winner.