Posts

There’s a special kind of sigh that comes out when someone says, “We’re rolling out OKRs.” It’s not quite despair. Not quite sarcasm. It’s the sigh of someone who’s seen this play before—probably in Q1 of last year, and Q3 of the year before that. The costumes are different, but the script is the same. Let’s get this out of the way: OKRs—Objectives and Key Results—can be useful. They’re not inherently evil. But like many good ideas, they tend to die in the wild. Not because the framework is flawed, but because the way most companies implement them is. ...

Scrum helped leave waterfall behind. Then it got in the way. It’s 9:57 a.m. and I’m watching someone update a Jira ticket to “In Progress” so they have something to say in the stand-up at 10:00. Welcome to Scrum in the enterprise: a play we all agreed to star in, long after the plot stopped making sense. I’m not here to burn Scrum to the ground. It served its purpose—just not the one it thinks it did. It helped kill off waterfall. It gave developers permission to talk to stakeholders more than once a year. And it introduced the radical idea that shipping working software might be more useful than delivering a 40-page Gantt chart with a straight face. ...

If you’ve ever deployed an EAR with a 12,000-line ejb-jar.xml, you deserve hazard pay. When one more JNDI lookup breaks you. For many of us still maintaining legacy Java EE systems, enterprise Java isn’t about sleek microservices or reactive streams. It’s about @RemoteHome , classloader gymnastics, and figuring out why a bean that compiles cleanly still can’t be looked up at runtime because it wasn’t bound under the “right” flavor of java:global. ...

Originally published on September 20, 2022 I was lucky. One of my first real gigs was at a startup that actually understood agile—not just the part where you stand around at 9:30 reciting what you did yesterday. No, I mean real agile. The founder believed in it. Not just the words, but the mindset. And in 2007, that made them look a little unhinged. Estimates in points instead of hours? Delivering small, usable increments? Developers writing tests on purpose? Teams interviewing and hiring their own teammates? It was weird. It was new. And it worked. ...

Originally published on August 26, 2022 In my last post, I described the arc most software companies follow—from chaotic startup to process-heavy bureaucracy. If you missed it, check it out here: An Evolution of a Software Company. Spoiler: there’s a pendulum, and it swings hard. Today, we’re talking about what happens next—and how to stop the pendulum from turning your team into a JIRA-powered approval queue. From Chaos to Control By the time the dust settles, you’ve landed in a company that’s seen one too many outages and decided that the cure is process. Lots of it. ...

Originally published on August 15, 2022 Let’s start with a story. It’s one you’ve probably lived through in some form, if you’ve spent more than ten minutes in the software industry. The Origin Story (and Its Inevitable Sequel) A company forms. There’s an idea, a goal, a dream. Usually it involves building the next great thing—a widget, a platform, a barely differentiated SaaS for onboarding PDFs. The founding team is small and experienced. They might be well-compensated, but more often they’re paid in the warm, fuzzy promise of equity. ...

I’ve been a commercial software developer since 2005. I’ve earned a BSc in Computer Science (mostly for the debt), and spent the last two decades writing code at big corporations, scrappy startups, and everything in between. WYSIWYG This blog exists because… I’m opinionated. I’ve seen “the next big thing” turn into the next big baggage. I’ve backed into a 12‑year‑old monolith more times than I care to admit—and lived to tell the tale. Here, you’re not getting hype, evangelism, or feel‑good manifestos. You’re getting: ...