<rant>
Recently I read Derek Webber's post about Spotify, in which he says:
I am paid $0.00029 per stream of a song on Spotify, and even this amount depends on whether the song is being streamed by a paid user or someone using the service for free. This means it will take upwards of 3,500 streams of a single song on Spotify to earn $1.00 versus that same revenue for one iTunes song purchase
His article made me think of a time years ago when I wanted to be in the music business.
About a month ago I migrated my blog from Tumblr to Octopress hosted on github-pages. I migrated because of Tumblr's slow page loads and frequent downtime. I picked Octopress because it is built on top of Jekyll, and it comes with a reasonable default stylesheet. Jekyll is a static site generator, which ensures I can deploy my blog on any old server in the future, and continue to use these same old tools (because I have them, rather than tumblr having them).
This XKCD comic struck a chord with me. I feel that it presents a false dichotomy. Either miss deadlines to design it right or go wild west cowboy coding and meet deadlines. I think there's a middle ground between those options. Design it correctly, but deliver less functionality, and add the rest of the functionality at a later date. While that's a simple solution, this comic touches on other related issues.
There are plenty of micro-frameworks out there for web development. Some ruby based ones are Sinatra, Padrino, and Camping. There's also Flask and web.py for python. These frameworks promise quick ramp up time and simplicity. They are small and do not provide much functionality, which is sold as "not getting in your way".
I have a running series of blog posts on learning emacs. It is a life-long endeavor. In this post I'm going to share some of the emacs resources that I have come across over the years.