← Feed

Posts

• View post

  Friday afternoon is a good time to slack off ^H^H make a tool I wanted. When code is heavily inlined and LTO'd it's not always obvious which functions are actually contributing to the code size. So, use `addr2line` to unwind the inline call stack at every single instruction, add up the instruction sizes attributed to each tree node, and scrape it into an HTML report. Nice to get an idea of "where should I look next"

• View post

  The fact GDB still uses `apropos` for help, but now prints an emoji when you ^C it, is such an incredible generational clash

• View post

  It's kind of sad when someone smart walks up to you and tells you about what they've been working on with Claude. Surely there's a German word for this feeling already?

• View post

  When I hear someone say "agent swarm" or "orchestrator", all I can think of is "we finally built the Man-Month, from the famous book, The Mythical Man-Month"

• View post

  🤔 GLP-1 general public license

• View post

  > wheels are intended to replace eggs completely normal statements from the Python community

• View post

  Is it "Unix guy coded" to invoke cargo from a Makefile?

• View post

  Idle thought in light of the Copilot GitHub thing: can you put yourself as Co-authored-by:? So I can have commits by Luke and Luke and Luke and Luke?

• View post

  When somebody tells me something is "complicated"

• View post

  Slightly annoyed at copy.fail releasing a PoC for an exploit still not patched in Ubuntu LTS so that they can advertise their AI thing

• View post

  Something I believe but never managed to put into words is that software/RTL tests lose some of their value every time you fail them. If you keep trying you will eventually find a hole in the test that lets incorrect code pass. Is there a name for this? This is one of the things that makes me skeptical of LLM-generated code. Thrashing around until the tests pass is more likely to leave you with fundamental correctness issues than if you'd thought about things up front, tried to make the...

• View post

  TIL: if node.js encounters an error in a minified source file it will print out the entire library as the "context line" for the error. For example the PDF.js worker is a 2 MB source file on a single line. If you're wondering why I'm trying to use PDF.js in node even though it uses browser APIs: I'm not, it's just there seem to be several ways to inject mocked test dependencies, one of them experimental and the rest deprecated or removed. I fe...

• View post

  Stubbed my toe on "JavaScript can't do SHA-256 if your web app wasn't served over SSL" and apparently this is a known rabbit hole, holy shit: https://github.com/w3c/webcrypto/issues/28 I would like to understand the thought process of someone who imagines any outcome other than people rolling their own crypto for locally served web apps

• View post

  When you see the word "orchestration" you should parse this as "the big ball of spaghetti that implements all of our logic and approximately 50% of our DB layer"

• View post

  I wrote a new open source licence for LLM-generated code

• View post

  This time of year is a good reminder to show gratitude, so I wanted to take the opportunity to say: "thank you for all the papers, science birb."

• View post

  Psst, want to see a funny GitHub issue? https://github.com/anomalyco/opencode/issues/18100

• View post

  @mntmn Gives new meaning to Pascal's "I have only made this letter longer because I have not had the time to make it shorter."

• View post

  I've seen things you people wouldn't believe. Cascaded ternary operators that continue for page after page. Nested twelve levels deep to force left-associativity. All gone, like tears in the rain

• View post

  ❌ "My boiler tripped my RCD" ✅ "My RCD passed its unscheduled test with flying colours"

• View post

  New post: A Love Letter to the Zbkb pack Instruction https://wren.wtf/shower-thoughts/a-love-letter-to-zbkb-pack/

• View post

  One thing I didn't know about the current LLM landscape: there are well-known techniques for completely zapping the RLHF refusals out of a model. It's not so much retraining as finding the direction that represents refusal and just... subtracting it. Qwen3.6 released today and the zapped model came out a few hours later. Zappy boi was happy to talk about some unspecified events in June 1989, and it ended the discussion on fuel-oxidiser explosives using household ingredients with...

• View post

  I pushed the Asciidoc source for my documentation through a local open-weight model (Gemma 4 31B Q8) and told it to check for errors. It found one typo and also noted I don't document the `tcontrol` CSR even though it's mentioned in several places, which I fixed: https://wren.wtf/hazard3/doc/dev/#reg-tcontrol I'm baffled by people trying to use these things to generate code, and I don't like branding the concentration of power in a handful of companies as &qu...

• View post

  Cool, a thing I worked on is public https://www.raspberrypi.com/news/rp2350-a4-rp2354-and-a-new-hacking-challenge/