sanufar

yeah, I grew up on my dad’s old iPod and BlackBerry up until high school; a lot of people in my peer group also were exposed to “older” tech by way of parents, siblings, etc.

Yeah, for some reason their /blog/ is 404ing but for anyone interested in their other articles, https://github.com/zartbot/blog/ has all of them, just sub out anything after /blog/.* with the folder path (e.g https://zartbot.github.io/blog/arch/jalapeno/ from https://github.com/zartbot/blog/tree/main/arch/jalapeno)

yeah, i definitely feel like we're currently in a very time-sparse model for review, when a lot of changes can be condensed locally. it'd reduce a lot of friction and also save a lot of compute costs if we were able to left-shift a lot of our current review work

Looks cool! Chapters is definitely something I've been angling towards as well. Any plans on going in the other direction (directly incorporating rich feedback/review into the agent loop through Stage)?

Hmm, probably about the marathon they ran the other day

I can’t speak for the Go bindings, but using the Rust crate has been pretty clean. Haven’t really felt the need to dip into deserializing raw .scip file in the time that I’ve been working with SCIP.

Super fun, def gonna try this on my own time later

In your experience, what’s the best way to increase signal? I feel as though a lot of devs struggle with the initial process of getting past screening, drawing attention to projects, etc.

Jeez there’s a few all around my uni and surrounding areas, did not know about that at all.

67 has been searched 13k+ times, more than 69 and 420 combined

Times are changing

Also your promo video is really fun! https://www.omnara.com/assets/landing/video/movie.mp4

Woah, I had this exact idea, down to the tunneling and local machine! I basically just coded up a Tailscale + caffeinate harness for my agents and it's been working super well. Your UI looks great though, glad to see more players in this space!

Big fan of the bottom right snowflake on this page: https://pdr-assets.b-cdn.net/collections/japanese-snowflake-...

Transferring each snowflake by hand with tweezers is insane, what a feat

I’m new to CRDTs, but could we clean up tombstones if all replicas acknowledge? Not too sure, but could we persist snapshots of state at certain “compaction” points?

Huh, the checkpoint primitive is something that I've been thinking about for a while, excited to see how it's implemented in the CLI. Git-compatible structures seem to be a pretty big pull whenever they're talking about context management.

Haven’t heard of SwissGL before, but this render feels super performant. Nice work with the Barnes-Hut algo!

This is hilarious lol, it’ll be any day now before we get a full JS kernel. Garbage collection could be an obstacle, but I know there have been some kernels written in Go/Java before

Makes sense, it feels like this just codifies a lot of implicit standards wrt OSS contribution which is great to see. I do wonder if we'll ever see a tangible "reputation" metric used for contribs, or if it'd even be useful at all. Seems like the core tension now is just the ease of pumping out slop vs the responsibility of ownership of code/consideration for project maintainers.

The way hashing is used for tokens and for making a pseudo symbol table is such an elegant idea.

Works pretty nicely for research still, not seeing a substantial qualitative improvement over Opus 4.5.

Wow, the shelf UI is a real throwback to the old iBooks shelf design! Looks great. +1 on the bounding box issue, it takes me out of the flow when accidentally triggering a book below the current shelf.

Ooh, this looks fun! I didn’t know you could cat-file on a hash id, that’s actually quite cool.

Class, looks amazing. The embed in the writeup looks so cool!

Seems to have been hugged to death as of now

Highly recommend writing a BitTorrent client. The spec is easy to grok, it has a bunch of fun subproblems that you can go as deep or as shallow as you want into, and it's super rewarding being able to download something like the Debian kernel after all of your hard work. Magnet links and seeding are two fun things to tackle post basic implementation. It also got me really interested in peer to peer systems and DHTs like Chord!

I think for me, the difference really comes down to how much ownership I want to take in regards to the project. If it’s something like a custom kernel that I’m building, the real fun is in reading through docs, learning about systems, and trying to craft the perfect abstractions; but if it’s wiring up a simple pipeline that sends me a text whenever my bus arrives, I’m happy to let an LLM crank that out for me.

I’ve realized that a lot of my coding is on this personal satisfaction vs utility matrix and llms let me focus a lot more energy onto high satisfaction projects

Wow, looks really nice, and the README’s super comprehensive. I don’t know if I missed it or not, but do you support inter-note link navigation (like Obsidian WikiLinks)? I know markdown-oxide did this with LSP actions, but I’d love to know more about your plans for navigation/lsp-like features, if any

Could you elaborate more on continuing with research?

OCaml’s REPL is lovely, but I found myself having some friction with initial startup. The build/package system was pretty convoluted, and I ended up choosing JS’s Core stdlib for my needs (a simple compiler). With the new multicore release it’d be cool to see OCaml in some more practical projects!