There is little objectivity around programming language debates, and AI will bring an explosion of new ones (I already made two). I took a shot at a more formal way to compare them with Language Explorer. cs01.github.io/language-expl…

Happy Opus 4.7 release day! If you want to share a Claude Code session publicly or privately (e2ee) check out sharemyclau.de

I like linkedin because there is so little content that your feed shows someone's post from 2 weeks ago. On x, your post is old and washed up after a couple hours.

I've never had so much fun with computers in my life. I literally do whatever the fk I want. Bending whatever stands in my way to my will. Tmux doesn't support mouse? Now it does. Compiled app looks cool? Reverse engineered, now it's "open source". Literally godmode.

Just when I got claude code cooking overnight for the first time, it shipped 27 PRs in about 11 hours, then BAM login issue and I still can't get back in. It would have taken me so much longer to pair program and fix all these legitimate quality issues. Total game changer.

Been heavily using Opus 4.6. > It is much more terse and to the point. > It never tells me I'm absolutely right (end of an era). > It is extremely good at coding, root causing multiple layers of code generation. Uses gdb for segfaults effectively. >It has corrected me on desgin decisions. > And it's only going to get better.

Self hosting achieved! Now to clean up and make public. npx tsx src/index.ts chad0 chad0 src/native-compiler.ts chad1 chad1 src/native-compiler.ts chad2 chad2 src/native-compiler.ts chad3 chad3 src/native-compiler.ts chad4 diff /tmp/chad3.ll /tmp/chad4.ll # no output

Now that everyone seems to agree that AI can write ~all code for us, where are the #remindme type threads over the last two years where AI bears have been proven wrong?

The singularity is near

The singularity seems near

I'm working on an AOT compiler of typescript code to llvm IR. It works, I'm now working on self hosting (compiling itself). Comment or DM me if you want access to the private GitHub repo to help get it across the self hosting finish line, or just to dogfood it/give feedback.

We’re making a fundamental tradeoff when we use these coding agents, but some of my engineering friends still don’t seem to get it. I see a lot of engineers make the argument “AI outputs still aren’t that great, code quality isn’t the same as writing it by hand.” And it’s all true, their argument is not wrong. But a coding agent can produce in a few hours what would take one of these engineers a month, or more. And I’ll even accept at face value that their well-thought-out, carefully written “artisanal code” would be better. But it would take them 10x longer to get out. So the fundamental thing we’re doing is to take this average code, generated in a day by an LLM, and try to make it a little above average. We are prompting our way onto something we can stand behind and support. And even if that extra work takes let’s say a week & it’s frustrating as hell to get it to something we can stand behind, a week is still better than spending a month or more. So that’s the tradeoff. And if you write code for money, pretty soon, nobody is going to let you take a month to write it carefully. Businesses all exist in a competitive space, if other businesses find a way to move faster by willingly making this tradeoff, they’ll all be forced to work this way. And it’s already happening and it will continue to happen. Then on top of all this, I notice many of us have this old mental model that we are writing code for other humans to extend, to work with, to add features to, etc. And in that world code quality matters a lot. But it’s unclear to me now that other humans will ever extend your code again. If these models improve even a little bit more, models will always be extending your code from now on. In that new world, you have to almost always make this tradeoff and you have to become great at it.

a great writeup of LLDB improvements last year by the lead lldb maintainer, Jonas Devlieghere, who reviewed every single one of the 2489 PRs jonasdevlieghere.com/post/ll… > I can confidently say that 2025 was an active year for LLDB. I counted 2489 commits in the lldb subdirectory from approximately 200 unique contributors. For comparison, in 2024 we had 1779 commits from 180 contributors. We had several new major contributors emerge that had a big impact on the project.

TIL you can mimic a statically linked binary by executing the runtime dynamic linker/loader with a library path to search. For example, exec ld-linux-x86-64.so.2 --library-path ../lib ./bin $@ And you can run libc.so --help to see its compatibility (but libm.so --help segfaults). Just used this technique when I had conflicting glibc versions and couldn't build a musl binary of bun . It worked nicely.