Today I LearnedRSS

March 2026

2026-03-16
Nobody Gets Promoted for Simplicity

It comes back to selling your work. That said, the example paragraphs they wrote for selling your work building simple solutions are atrocious. It's not that they're wrong about their thesis, just that the examples provided to help the audience isn't going to be of much help. That's because getting recognition for your work is a performance. You have to learn how to sell your work effectively, not just at all.

Never just list features you built, nor the ones you considered. Instead, always talk about your work in the broader context of the impact to the business and its customers. I've deleted a few dozen lines of code that saved a company a 6 figure sum. Sure 99% of the work was being completely sure those lines weren't load bearing. You think I sell it by saying I deleted a little bit of code? No! I talk about how I just saved us all a small yacht's worth of cash.

Making more money requires people thinking you're too valuable to not pay for. Meritocracies are a utopian myth. If they really existed, companies wouldn't be spending trillions on advertising. Your skills and abilities are as valuable as people think they are. Part of that is genuine demonstrated ability. But people pay more for Coca-Cola over the store brand because it gives them a good feeling. Blind taste tests show store brands and home made soda tastes better. I'd link a study here, but there's hundreds of them and to my limited knowledge, no good meta analysis. Just go search around.

Think about it like balancing the five P's of marketing: product, price, place, people, and promotion. You want your price to go up, and let's assume you're working as hard as you already can on your product, that is, doing your best work. You can still impact place, people, and especially promotion (because so few engineers are thinking about it). Place here is where you spend your time. Are you working on the problems your company cares about? If not, you have to then spend extra time to convince them that the work you're doing even matters. Why fight an uphill battle into a fierce headwind? People is about soft skills. Knowing your audience, presenting yourself professionally, and having great customer service; that is being attentive, friendly, and communicative. Lastly is promotion, and that's a big part of building the feeling people have about the work you do.

That feeling matters. In all but the most Byzantine systems, promotions happen way before any promotion package or formal process. Even in large companies, people know roughly who they can count on to get things done well. You're basically playing dodge ball in gym class again and managers are picking team members. They know the kid who can run fastest isn't always the best at the game. But that kid's at least got a reputation. The weird kid that keeps to themselves is an enigma though. Especially if they mostly just stand there and maybe walk back and forth a bit. They may be a secret martial arts master able to dodge anything, but it's not obvious if they're downplaying it.

One of the missing pieces in this discussion is how much harder simple solutions are to make. This is going to hurt your perceived velocity. What's most likely to hold you back is if you're spending your time simplifying everyone else's projects. You need things you can take credit for. If you're actively leading projects and able to deliver working solutions under budget—in both time and money—then it becomes easier and easier to demonstrate a track record.

My strategy here is have long term goals for the business and your customers. Push for the time to action on the things that would lead to demonstrable improvements in business value terms. Not all of your ideas will be tractable. Some of the simplifications I've been pushing for, I'm still pushing for over five years later. Many of the simplifications I wanted have already been done. Doing those and demonstrating the value gave people confidence in me to go work on bigger projects. Past performance inspires confidence in future results but people only know about your performance if you talk about it. If you can deliver concrete value, you'll be trusted with the power to do more.

2026-03-13
Lecture Friday: A Swift Introduction to Geometric Algebra

I don't have a lot to add here, only to say that I keep coming across geometric algebra and it keeps seeming like something I need to really dig into and learn. I've started going through a bunch of the resources from bivector.net. Lots of things to unpack in this. Not fully sure if this is a better basis for a 3D renderer than quaternions, but it's definitely worth playing with given how much more intuitive the math notation seems to be.

If you're also going to look into studying this, another term you want to know is Clifford Algebra.

2026-03-12
Stop Telling People To Sanitize User Input

Signal boosting this. Always manage user data transforms for safe encoding on output, never input. Your input code shouldn't know how the data's going to be used. Maybe it'll be put in an email, maybe it's going on a webpage, maybe you're printing it. Each of those have very different sanitization requirements. Your input code should not be trying to account for all the different ways it'll be displayed.

Data given to the system should be treated like your handling explosives. There are ways to safely move it, ways to safely store it, but you don't go messing with it prematurely. Definitely don't trust, touch, look, or lick it unless you absolutely have to.

2026-03-11
Dynamicland

I know I missed posting my usual Lecture Friday, so I come bearing great gifts. Today I was reading Dave Gauer's book review on The Art of Doing Science and Engineering, and he opened by talking about Bret Victor's forward and noted that his site has an amazing collection of all the best research papers in computing. Upon reading that I wondered, "Huh, what's Bret up to these days." I then proceeded to fall head first into Dynamicland. I'm still not sure what to make of it all but I'm really excited by everything I'm seeing.

So what can I say except, "You're welcome?"