Today was Thursday of my second week at the Recurse Center as a participant in this year’s Summer 2 batch! I still haven’t decided how to number these blog posts going forward. I’m open to suggestions. Please message me with ideas on Zulip, or Hachyderm, Bluesky, Discord, or wherever else you can find me (even if it’s Twitter).
I’m running a bit short on sleep today after driving home from a family vacation late into the night.
Despite that, I had a pretty good and full day of RC activities!
The CLI Design reading group Erika has organized had a good discussion this morning and also wandered into some adjacent topics like Git workflows and the value of different types of commits. We talked a bit about how it can be helpful to make a lot of “piton” style commits while feeling out the design f something new or working out how to get things working at first, and how it’s also nice to rebase and clean up those commits into a sequence of better scoped tidy commits with good commit messages, but that it would be nice to preserve both and show that developments don’t always spring forth from the mind of super developers fully formed, but can have messy beginnings at first. We speculated on different ways to represent these different types or phases of work together using Git alone (rather than a web gui like GitHub).
After that I spent a good part of the day pairing with Rhea again, this time on a somewhat ambitious jq-like project she has been wanting to work on. This gave me an opportunity to write some Rust and to learn more about parsing. We used
serde_sexpr to create some test data serializations and then built a small S-expression parser using
nom. This was a lot of fun! Now that I know a bit more about
nom, I’m probably going to revisit a project I had in mind from years ago that can also make use of that crate.
Later I ended up taking a nap to catch up on a bit of sleep and then worked on some smaller tasks like tidying up dotfiles and testing some recent changes in the Media over QUIC (MoQ) project I’m contributing to.
Tomorrow I’ll probably do some more of my own work in Rust and pick up where I left off with my MoQ demo features.