Jon Atkinson

I'm a technologist, with a long and varied history in high-end technical delivery and infrastructure management. I'm particularly interested in managing software teams, rapid application development, and scalability challenges.

New Years Resolution: Stop switching tools

1st January 2010

Like nearly everyone else on the planet, I've spent some of today thinking about how I can do better next year. My resolutions are roughly split into two categories; tech stuff, and everything else, which makes sense considering I spend at least half of my waking hours in front of a computer.

Since I can remember, I've compulsively tinkered with the computer in front of me; changing hardware, changing operating system, experimenting with different editors, different languages and different organisational systems. I'm a classic computing dilettante, I have a very broad knowledge and very shallow of many areas of programming and administration, but no real expertise anywhere. This bleeds into my choice of environment; I'm forever switching tools and languages trying to find the perfect tool. After years of searching, I've not come any appreciable distance nearer to that goal, so it's time to stop trying.

From what I can tell, this is a fairly common disease among programmer types. Some may tinker with software, some with their time management system, but the problem remains the same; we are wasting time trying to perfect a system which will never be perfect.

So, the main tech resolution I've made is to stop trying new things. This has been happening naturally over the last few years anyway, it's probably a combination of age and habit, but I'm going to endeavour to ignore new software, new methodologies and new trends in the coming 12 months. For the next year, I'm going to use the following tools, without deviation:

  • OSX Snow Leopard on my laptop, Debian Lenny on the servers. I already know these platforms, they're both stable, and I have no reason to try anything else. I'm often tempted to try the latest releases of operating systems, which inevitably results in a week of lost productivity. As far as I know, there is no scheduled OSX release in 2010, and no release date has yet been set for Squeeze.
  • TextMate. I've spent about half of 2009 using Emacs, but the power of Emacs comes at the expense of distraction; the hours I've spent writing elisp to try to save myself 10 seconds here and there far outweigh the actual time saved. TextMate is stable, the next release is a long way off (in fact, the TextMate 2 FAQ says it's coming when we get Duke Nukem Forever, so I guess that means never), and the lack of customisability compared to Emacs means there's just less to tinker with.
  • IPython. I've spent far too long investigating other Python Repls, and while bpython is nice, IPython and ipdb.set_trace() have become so ingrained in my muscle memory that changing to anything else isn't worthwhile.
  • Django. I'm not sure exactly how the roadmap for Django is laid out, but I suspect the 1.2 release will be with us for most of 2010, and Django is adequate for all the work I'll be doing with 84labs in the next year. I've toyed with Kohana, Rails and Seashore in the last year, and none of those brought any significant benefits over Django that would be worth holding on to.
  • Safari. Safari is the most pleasant browsing experience to use on OSX. 2009's development of the Web Inspector removed any real dependency I had on Firebug, and combined with ClickToFlash and /etc/hosts level ad blocking, I've no need to use any other browser. Thankfully, I very rarely need to test my work in other browsers, so while I'll probably keep around a Firefox installation, it's very unlikely that it'll become my browser of choice again.

So, that's it. We'll see in 2011 if I'm still using this software (that's where the 'resolve' comes in), but I'm looking forward to gaining at least a week or two of productive work simply by keeping a static environment.

Will anyone else commit to joining me?