Huge new Daily Parker release

You are looking at the Inner Drive Technology Journal project release 1.0.9723, containing the biggest feature set I've ever released in one week.

The last release came out only last Thursday and fixed a couple of bugs. Today's release, in contrast, started almost immediately after 1.0.9715 and contains an entire epic of features that brings advanced searching capabilities to the application. You can now search by project, event type, tags, specific fields, and geography. Want to see which shows I attended at Ravinia Festival in 2024? Or how about Daily Parker posts tagged with both Cassie and AI since last August?

The career-altering part: the Advanced Search epic only took 17 hours to design and implement, starting with Claude Fable 5's analysis task last Thursday afternoon and finishing this morning with a couple of tweaks to the Search page layout.

When I first put the Epic into Jira six years ago (!), I guessed it would take me about six weeks of after-hours and weekend coding to finish, something around 75-90 hours. Maybe longer. And a lot of the work would be tedious as none of the features is particularly groundbreaking in itself, just fiddly to implement. I also figured on spending a dozen or so hours just learning how to work with Cosmos DB's native searching and filtering capabilities so that I didn't reinvent the wheel.

Claude Sonata did all that stuff for me. I got to build the software, not the code (though I did review every line and made hundreds of small tweaks to Claude's work), letting me concentrate on what I wanted it to do and how I wanted it to work for the user, rather than how it had to work for the compiler. (And now I've learned how to work with Cosmos DB searching and filtering because I can read and step through working code.)

Throughout my career, I've experienced a couple of huge leaps in what I could accomplish. Reading Steve McConnell's _Code Complete in 1995 taught me how to code like a professional. Switching from Visual Basic to C# in 2003 taught me how to write performant, truly object-oriented code. Then in 2011-12 I moved from on-premises hardware into Microsoft Azure along with continuous integration/continuous delivery. Each step made it easier to spend time on the user and less on the technology.

So now here we are, with 30 years of professional software development experience and a $20-a-month Claude subscription, releasing an entire epic containing 11 features¹ in just 8 days and 17 hours of effort in my spare time.

I love living in the future.

One more thing: This experience has also gotten me thinking about the role of junior developers in the near to medium future. It's really going to suck for a lot of them, particularly the ones who took classes to "learn how to code." I graduated college with a degree in history and most of a degree in music and never thought I could turn my hobby of writing software into a career until the point in my second year of law school when a lawyer paid me to automate his real-estate closing documents. That was 1995, right before the first boom. I got very lucky. The kids going into software development today are already obsolete unless they've learned the kinds of things I learned reading history and music. I'll have more to say about this in the next couple of weeks.

For the moment, though, I'm going to lean back and enjoy today's software release. It's the first exciting deployment I've had in a long time that was exciting for the right reasons.