Ever since I was a little kid I knew that I wanted to develop software. From what I hear that clarity of purpose is rare. It wasn't until college that I really felt I "got serious" about constantly improving on what I know. That learning has accelerated in the past few years since I finished my second Master's. 2012 has been a very good year for expanding my capabilities in software development.
However, I sometimes forget about the improvements I make and why I'm making them. I take some of what I've done and learned for granted.
Along the journey we commonly forget its goal.
This post is a summary of what I've accomplished and learned during 2012. Hopefully it helps me remember where I've been and where I may go next.
I also learned LESS and got much better with CSS3.
I did some work with the Urban Airship Python API binding library for a mobile app proof of concept to a client that unfortunately didn't pan out.
I issued a pretty simple but useful pull request on impress.js after I saw it on Hacker News but I found out it wouldn't work with my Logitech presentation pointer. All my 2012 presentations were created with impress.js. I never really felt like I had enough time to grasp the full power of the framework - but it's a really well written open source project. I'm working with deck.js for at least my first several 2013 presentations but I may go back to impress.js at some point.
I open sourced on Github my Visualization Portfolio side project. I also open sourced this blog and my Coding Across America blog. I need to figure out the licensing situations on those things though.
I launched (and relaunched) several side projects.
First up was ProofDriven, an attempt to create a full-stack Python web app that produces estimates of savings from software development upgrades (for example, going from no source control to using Git) so developers could give a financial report to their managers. ProofDriven produces PDF reports that can be printed out. Each report can be customized with various parameters, such as the total cost of the software development project, the team size, and other assumptions. The models and report outputs are based on input assumptions as well as other technology-inherent factors based on which report is chosen. I integrated Stripe and PayPal for payments on reports other than the free default one. I never marketed the idea but I may reboot it at some point and see if I can apply the Customer Development approach to find out if there is any potential market there.
I launched (and relaunched) this blog a couple of times on different platforms. More on that in the next section.
I created Visualization Portfolio and deployed it on Heroku. I'll begin adding more than the base three visualizations with d3.js soon once I get caught up with the 3.0 release.
I also just created but haven't finished (more on this in a few days) Full Stack Python. It will be a site dedicated to explaining everything you need to know to launch a Python-based web application from server to browser. It should also be useful to people who know other stacks (RoR, JVM, PHP) and what to compare them to the Python stack. It'll probably take me into 2013 before I finish that site up and officially launch it.
I moved from my original Tumblr blog to a statically generated Pelican. At first I hosted it on Amazon S3 but then I moved it over to Github Pages. Github Pages is absolutely awesome and I've been really happy with it ever since I made the switch.
My Coding Across America blog is also on Github Pages. Coding Across America should begin to get interesting in late January when I start talking to startups in Washington, DC. Then on March 9th I'll hit the road for a 5 month tour of the US. Any startups out there in any of the 30 cities I'll be in please contact me on Twitter @mattmakai or by email at firstname.lastname@example.org.
I gave a bunch of talks in 2012. Among my favorites were my April 2012 DC Python talk about using Urban Airship to send push notifications, my high-level Python Indoctrination class, and my discussion of Agile Software Development in the Federal Government.
I'm excited about my slate of talks for 2013, especially that many of them will be in cities I've never been to before: Memphis, San Francisco, etc.
I tried a little experiment in mid-September to write a daily learning log. Basically I force myself to go look at the documentation of a programming language or library I'm using and learn something new about it - however esoteric or at that moment seemingly useless.
For example, I'm been reading the d3.js API documentation to learn more about the shortcut functions that make it easier to manipulate DOM-bound data. Many of the functions aren't necessary but are really useful for cutting down on boilerplate code.
I really like the learning log and I'm going to continue doing it in 2013. The only major issue I've had so far is sometimes I get caught up in other things and I have to "catch up" - so I try to learn like 2 weeks worth at a time. That's less useful than learning one, two, or three new topics a day.
I've already written about what skills I'm currently working on developing to be a better full-stack software developer .
The majority of 2013 will be defined by my Coding Across America trip.
I'll write up a separate post sometime in the near future to cover 2013.