Here’s what I released:

1. Update to web app for client M
2. Update and open up for Twimmetry
3. Repurposing of a deployment script for use with Twimmetry

The web app update was straight-forward, although it presented a few small hassles.

Likewise the Twimmetry mods largely involved turning off the beta-user requirement (now anyone can use the site) and fixing a few bugs here and there. Technically I did two releases of Twimmetry over two days, but that didn’t sound fun to count toward my goal. My most embarrassing bug wasn’t actually a software bug at all. I misspelled “achieve”. ‘I’ before ‘E’, right? Terrible.

The toughest bit of code to release actually turned out to be the deployment script. That should have been the easiest, since I was totally cheating and just copying an existing deployment script and making a few small changes. Or so I thought. I’ll spare you the long version and just sum up by saying that I need better standardization on how I store my Django projects in Subversion and how they are structured for deployment. I also shouldn’t debug code at 2AM… ever.

I need to find a nice push-button, multi-platform deployment solution that works the way I think and doesn’t do a bunch of other crap like server admin which is nice but overkill for an apps shop like mine. One nice thing about the deployment scripts that I’m using is that they are totally customizable and they can reproduce any action that I take at the command line. (I’m using Python and pexpect if you’re wondering.) I pay a hefty price for that power sometimes though. Also, while I have crazy wizard-like powers at the command line, GUI-driven, non-Web Windows deployments are gumming me up. But I digress.

Next challenge: grow from 3 in 3 to 7 in 7. Although I may give myself a day off because my neighbors are currently reenacting the Revolution, the Civil War, and the Dresden bombings – simultaneously – in my neighborhood. It’s a tad difficult to concentrate.

Happy 4th!

I once read a short story about a circus performer who made a deal with the devil. In exchange for his immortal soul the performer would become the greatest acrobat in the world. I don’t remember many details, but the gist of the story was that the Devil could give him the ability to withstand a fall from any height. And sure enough, the Devil kept his word. The Devil offered no magic though, only training. The training was simple: To start with, the guy had to roll out of bed and fall onto the floor every morning when he woke up. After he’d done that for a week, the devil had him practice falling off a ladder until he could do that comfortably and without hurting himself. This pattern continued. Eventually the performer worked his way up to being able to fall from 30 feet without hurting himself.

“So what you’re saying is that you made a deal with the Devil and/or you want to be an acrobat?”

Uh… no. (wtf?)

What I’m saying is that I’d like to revisit my 30 in 30 project. Much like our circus performer, I aim to take baby steps. If you’ve been around me long enough you’ll eventually hear me go on and on about kaizen and baby steps. And that is the tool I’m whipping out here.

Excuse me while I whip this out.

I’m going to scale things down to a more manageable size via the time-honored Order of Magnitude reduction. That is to say, instead of 30 releases in 30 days, how about I shoot for 300 releases in 0.3 days?

Wait. Taps calculator.

How about 3 releases in 3 days?

Sounds fairly manageable. Same terms as before. Dirty tricks? Allowed? Hand waving? Totally cool. Victory declaration, regardless of outcome? Obligatory.

So. We’ll see.