Nov 2025

This is a Jukebox of the best blogs I have read so far in November. My plan is to update it whenever I read something new that is memorable.

Super Specific Feedback: How to give actionable feedback on work output

A great blog about what feedback helps you grow. Worth a read for everybody who gives feedback and wants to receive quality feedback. I even wrote my own post summarizing what I read.

Note To My Younger Self

And Note to my slightly older self. The posts go hand in hand. I love reading lessons learned from other people.

From Vim to Zed

This is a good break down why you should consider trying other tools from time to time and why one can switch from Vim after 20 years of heavy use.

A Few Words on Testing

This is how most people should see software testing.

If you don’t tinker, you don’t have taste

Memorable quote:

when you tinker and throw away, that’s practice, and practice should inherently be ephemeral, exploratory, and be frequent - @ludwigABAP

“This does not mean I spend every waking hour fiddling with my neovim config. In fact, the last meaningful change to my config was 6 months ago. Finding that balance is where most people fail.

Over the years I have done so many things that in hindsight have made me appreciate programming more but were completely “unnecessary” in the strict sense.”

Fake it until you make it

Rousseau invites endless introspection. Franklin invites progress. The first is about how you feel; the second is about what you build.

Design the right data model

Very interesting view on the importance of data models. This got me thinking…

Skills

Claude code has skills. Skills are a nice addition to the coding agent workflow. Very cool that I don’t have to use my Tmux plugins to pipe prepared texts into coding agents any more.

You should write your own Agent

This is one for the LLM haters and the lovers. As I read this I thought of course it works like this I have already written something like this by accident. Great post. And it throws the question: Is MCP the right way for Agents/Clients to communicate with Tools/Servers?