[learning cs] computer science tooling

28 Sep, 2025

I prioritize learning and trialing out different sde tools.

This week, I've had quite a bit of fun configuring and test driving iTerm, Sublime Text, and Zed.

With iTerm, I also spent time configuring local gen ai integration, which meant I fiddled with Ollama.

From it all, I got a taste of the history of Unix and its derived environments. Simple example, ctrl + (key) in terminal versus cmd + (key) in MacOS.

I learned what mouse reporting is and how it affects the interplay of terminal emulation and workflow. Upon reflection, I'm astonished I quickly I've gotten used to certain hot-keys and overrides.

Setting up Ollama, I learned some very basic parameters of gen ai.

I learned how to file a good bug report in Zed. No small part in thanks to great contribution and bug reporting documentation from Zed--which Claude Sonnet 4 was able to wonderfully reference.

I'm also getting better at, on my individual level, context engineering and prompting. Though my biggest take away when it comes to gen ai is to just turn it off completely¹.

Probably, the best case scenario for me is that gen ai answers some troubleshooting info much quicker than reading (often times) convoluted docs.

Overall, I'm enjoyed being on the command line. I feel like there's much more to explore and discover about the history and complexity of computing from that cursor.

Not bad for a week. Vim, next?

While AGI or even improved models may not be solved with scaling, it's clear that dependency on gen ai is not "augmenting" me. It's supplanting the struggle in execution functioning: planning, thinking, managing my reactions. Gen Ai gives that dopamine hit way too soon and in too large a quantity.

One simple analogy is we don't just teaching our children how to use a calculator and call it learning mathematics.

#learning_cs