Mike Kalvas
Mike Kalvas reposted
So my month of @jj-vcs.dev turned into a year+ and it’s definitely my full-time, not going back, always using it tool now. I never had a problem with git. It does what I need and I still use it sometimes, but jj is nicer to use. It’s not a wild game changer for me personally but it’s just… better.
Think I might make June my “month of jj” and see how it goes
Dang. This is going to be a real pain migrating away from @digitalocean.com. I have personal and professional stuff on there, but it’s all moving now and not ever coming back.
most expensive collection of shitty dotfiles and security vulnerabilities in human history
Mike Kalvas reposted
This isn’t fair, @oxide.computer’s branding was already top-tier and now they’re making it better? Leave some for the rest of us out here wallowing in SaaS blue!
Mike Kalvas reposted
You know what you want? You want a T-shirt with a crossover joke on it so incredibly nerdy that anybody who laughs it immediately becomes your new best friend... shop.codewithrockstar.com/product/line...
Mike Kalvas reposted
Mike Kalvas reposted
Mike Kalvas reposted
got frustrated recently at work about people writing using AI instead of using their brains! here's my take on it: sophiebits.com/2026/06/25/t...
I wrote a thing that I've wanted for a long time. There's a million things that can do this, but it was 45 lines of Rust and it's mine. It felt good to make something simple, complete, and useful using my brain and knowledge as opposed to dealing LLMs all day at work. github.com/mkalvas/lmk
Really liked this post, The Social Contract of Writing by @jola.dev. I had much the same thoughts and had even read the Oxide RFD referenced in it. Hopefully there are enough like-minded people who’ll be writing cool stuff for at least a little while longer. jola.dev/posts/the-so...
Mike Kalvas reposted
While his mountain of horrible personal conduct could fill multiple books, one fact in particular stands out: A year ago, Musk’s actions directly led to the deaths of hundreds of thousands of people. He did it knowingly. And, worse — gleefully. Read more: www.theverge.com/tech/949259/...
There are a lot of good books that I’ve enjoyed over the years, but Infinitely Full of Hope by @tomwhyman.bsky.social and No Straight Road Takes You There by @rebeccasolnit.bsky.social are the two books I’ve **needed** to read to meet the times we live in. I cannot recommend them enough.
Great to see that Zed is helping out and I still have options here, but Anthropic's TUI is just so, so, so bad compared to Zed's UI. Looks like I'll be canceling my Anthropic account this month and looking for something different.
Your Claude subscription still works in the terminal. And next week we're shipping Terminal Threads to Stable so terminal agents appear in the new Sidebar alongside your native agent threads.
Mike Kalvas reposted
As many standardized-test-takers know, there are two ways to score highly on a test: learn the material, or guess what the question-writer is thinking. LLMs have gotten really good at the second part!
There’s some cool recent research on this phenomenon! It turns out vision language models excel at image benchmarks *even when the actual images aren’t provided,* because the answers are implicit in the questions! arxiv.org/abs/2603.21687