Update | 2025 Mar 23

Hello!

Embedded Stuff

I’ve picked back up some embedded systems projects I was started to explore some months ago. Getting closer to the bare metal is fun, and it’s really interesting to learn the raw mechanics/protocols of some of the most fundamental electronic/computer systems that are used all over the place, from the internet to elevators.

I recently reproduced Ben Eater’s reliable data transmission video. It starts with simply printing a message to an LCD, then upgrades to sending a message over a serial connection where the computers are synced up based on simple internal clock delays, then ends with the transmitter sending a clock signal alongside the message to overcome clock differences.

Here’s a video of the simple message transmission working for 2 letters “H” and “e” but then failing (because the clocks get out of sync):

Here’s a video of it working with the clock signal:

Generally going forward, I’m looking to better understand the physics side of things, and continue doing more embedded systems + circuit projects. Hopefully in upcoming blog posts I can share increasingly interesting stuff.

Math Academy

I’ve been continuing with Math Academy! Finished linear algebra, and currently relearning multivariable calculus. I didn’t really feel that enthusiastic about learning this subject back in high school, but I think this second time around it’s going to be a lot more fun. Before I was just getting a grade in a class, but now the math is fun for the sake of itself, and I can see all these different subjects I care about that will become easier from understanding this better. And math academy makes learning it so much more effecive, and the progress is more concrete + rewarding.

I haven’t been doing it everyday, but I’ve gotten into a solid habit of doing larger chunks Friday-Sunday, and maintaining a place in the second to the top league.

I still really wish there was a math academy for other subjects. The confidence I have in my progress while learning is very motivating, whereas spending most of my learning time reading vs. less time doing problems with a textbook feels much less concrete. When it comes to learning more of the physics fundamentals of electricity/magnetism that I’m starting to get into, I would love something math academy-esque. But for engineering projects, building things is so hands-on I feel like I can get into the progress/learning feedback loop pretty well already.

Noobular

Not much of an update lately. Got some basic functionality for auth working but have since been focused on other things.

I’ve been learning to better distinguish when I should rewrite or refactor something.

On rewriting:

• Writing from scratch is fun, and teaches you a lot.
• If something was half-baked, it can be better to start from a blank slate than clean up a mess.
• Rewriting something you’ve written before at the start of a new project can let you tackle things with fresh eyes, e.g. john carmack rewriting a game engine from scratch for each new game.

On refactoring:

• Part of the beauty of software is its malleability, and each time you refactor you pick up more ways on how to make your software more adaptable. And most of the time you don’t know upfront how everything is going to look, and structures naturally emerge as you learn more about what you’re making. In this case rewriting doesn’t really make sense.
• Often people wrote things in a particular way for a reason, and rewriting it ends up either throwing out important things unintentionally, or the rewrite ends up looking really similar to the present thing. I continue to see situations where instead of thinking “huh this is weird, it could be written in a nicer way,” it’s better to think “huh, the author came to a solution I didn’t expect, what led to this being written this way?”

In hindsight I don’t think I really needed to rewrite noobular. It’s too early on to need a rewrite, and the structure that emerged already worked well for what I needed. Sure I can make some of the my HTTP structures slightly nicer, but really that doesn’t really matter all that much (and ended up being pretty similar to what I had already). There are many things that needed to be cleaned up, but could have been done perfectly fine with refactoring. The one thing I found was kind of valuable to do from scratch was the DB stuff. I am not using an ORM and am forming my own sort of structure/wrapper around the DB, and haven’t done this many times before, so the change is big enough that it might as well be rewritten.

Anyway, as I do more embedded projects, I’m not sure how much time I’ll be dedicating to noobular! Only time will tell.