Runtime Arguments

Jim McQuillan & Wolf

Conversations about technology between two friends who disagree on plenty, and agree on plenty more.

  1. ٢٥ يوليو

    33: Perl Refuses To Die

    Wolf and Jim open with two AI stories from the week. First, the report that an OpenAI model "escaped its sandbox" and went after Hugging Face — a story Wolf finds infuriating precisely because the words don't go together. Reading the actual joint statement, it turns out the incident happened during an internal evaluation explicitly designed to push models toward advanced exploitation; the model found a zero-day in the package registry cache proxy, escalated privileges, and moved laterally until it reached a node with internet access. In other words: they built a gun, and it fired a bullet. Then Jim covers Linus Torvalds' LKML post declaring that the Linux kernel is not an anti-AI project and that AI is a useful tool for developers — following similar comments from Andrew Tridgell of Samba and rsync fame, and drawing similar backlash. That leads into a candid discussion of AI's loudest critics, why Jim (a skeptic himself as recently as February) thinks most of them are working from an outdated picture, and where "adapt or die" lands when it's your own brother's freelance copywriting career that dried up. The main topic is Perl — not how to write it, but why a language that once owned the world became one you'd never pick today. Jim traces his own path from a million-plus lines of COBOL to Perl in 2000, including the genuinely clever trick of linking the Perl runtime into the AccuCOBOL C bridge so COBOL programs could talk to Postgres — which let him replace the system one piece at a time over about nine years without ever stopping the world for a rewrite. That sets up Wolf's rule (if your plan is to rewrite everything, you've already lost, see also: Netscape) and the argument that Perl 6 broke exactly that rule, fracturing the community for years before finally becoming Raku. The hosts dig into the numbers: of ~17,500 GitHub projects created in 2026 with 100+ stars, exactly three had Perl as their primary language — 49th out of 116. Along the way: CPAN as possibly the first package repository, why Python is like coming to work to play with a puppy and Perl is like petting your porcupine, the languages that actually protect you (Rust, Swift, TypeScript) versus the ones that just let you do anything, the companies still running on Perl (Booking.com, Craigslist, cPanel, DuckDuckGo, IMDb, Ticketmaster, Shutterstock, Slashdot, Bugzilla), and Wolf's verdict — Perl is dead in the way COBOL is dead — followed by a genuinely warm case for why that leaves Jim in an enviable position. Links: OpenAI / Huggingface incident: https://openai.com/index/hugging-face-model-evaluation-security-incident/Linus Torvalds LKML post about AI being a useful tool: https://lore.kernel.org/linux-media/CAHk-=wi4zC+Ze8e+p3tMv8TtG_80KzsZ1syL9anBtmEh5Z40vg@mail.gmail.com/The Tiobe Index showing Perl's ranking: https://www.tiobe.com/tiobe-index/ Hosts: Jim McQuillan can be reached at jam@RuntimeArguments.fm Wolf can be reached at wolf@RuntimeArguments.fm Follow us on Mastodon: @RuntimeArguments@hachyderm.io If you have feedback for us, please send it to feedback@RuntimeArguments.fm Checkout our webpage at http://RuntimeArguments.fm Theme music: Dawn by nuer self, from the album Digital Sky

  2. ٢٧ يونيو

    31: Local LLMs: Good Enough Might Be Enough

    Jim shares his adventure into running LLMs on his own hardware. For him it's less about saving money and more about privacy — working in healthcare, he can't send patient data to the cloud. App vs. model: Claude Code and Codex are applications, not models. Features like plan mode come from the app. (Wolf's "Opus Plan" is a Claude Code mode that uses Sonnet 4.6 for most work and Opus 4.8 for planning.)Ollama makes local models easy — ~15-min install, runs on macOS/Linux/Windows, and exposes a REST API. Not to be confused with Meta's Llama models. Example: `ollama run llama3`.Parameters & training: Think of an 8B model as "8 billion knobs." Training randomly initializes them, then refines predictions over billions of iterations. Wolf ties this to Markov models (parameters ≈ weighted edges) and the Bayes episode (random init = priors).Fitting big models in memory: Quantization shrinks 32-bit parameters down to ~4 bits. Mixture of Experts (MoE) keeps only part of a model active (e.g., Llama 4 is ~108B params but ~17B active).Jim's tests (M1 Mac Studio, 64 GB), asking why H₂O is liquid: Llama 4 Scout took ~10–15 min and maxed out RAM/swap; Llama 3 (8B) answered in ~31s; Qwen (36B) gave the best answer in just 34s.The open question: Is local "good enough"? Wolf's real test isn't trivia — can a local model write and iterate on an 8-page implementation plan? (Homework for Wolf's 128 GB MacBook Pro.)Build your own: Fine-tune an existing model or train from scratch. Jim's dream: a local model fine-tuned on his DB schema + 2,000 SQL queries so users could ask in plain English and get runnable Postgres — no cloud required. Browse Hugging Face for specialized modelsHosts: Jim McQuillan can be reached at jam@RuntimeArguments.fm Wolf can be reached at wolf@RuntimeArguments.fm Follow us on Mastodon: @RuntimeArguments@hachyderm.io If you have feedback for us, please send it to feedback@RuntimeArguments.fm Checkout our webpage at http://RuntimeArguments.fm Theme music: Dawn by nuer self, from the album Digital Sky

  3. ١٣ يونيو

    30: Available compute: way more than you need, right up until you need it!

    You almost never have exactly the right amount of compute for the job. Either cores are sitting idle while your code runs on one, or you've got more problem than machine. This episode is about the two fundamental tools for closing that gap — and why picking the wrong one makes things slower, not faster. Topics covered: Concurrency vs. parallelism — the core distinction: concurrency is a scheduling problem (you're waiting a lot); parallelism is a compute problem (you need more processing). They are not interchangeable.IO-bound vs. CPU-bound — how to identify which problem you actually have before writing a line of concurrent code. (And a third case: memory-bound, where the fix is data layout, not more cores.)Threads aren't always what you think — system-level vs. user-level threads; why JavaScript's `async`/`await` is single-threaded concurrency and not parallelism; why Go routines can be either.Colored functions / async infection — why `async` spreads through a codebase the way `const` does in C++, and why Go sidesteps it entirely with `go func()`.Go channels and Rust ownership — why these two language designs are the cleanest modern answers to shared-state problems.Python's GIL — what it was, why it hurt, and why 3.14+ removes it (with caveats for single-threaded performance).Amdahl's Law — the mathematical ceiling on how much any parallelization effort can help, and why it's specific to your problem.Hidden parallelism — CPU pipelines and branch prediction run in parallel below your abstraction layer, and you can't see them without special tools.Communication is the real enemy — GPU bus bandwidth, cluster fabric, Apple Silicon shared memory vs. NVIDIA CUDA: the cost of moving data often swallows the benefit of more cores.Fork and copy-on-write — how Unix `fork` got fast, and why Python's reference counting undermines it.The actor model — how Erlang (and now Swift) solve the ownership problem by letting the object own the data, not the caller.Heisenbugs — the bugs that live in parallel code and only appear when you least want them. Examples: Pixar render farms — 130,000 frames × 24 hours each, solved by embarrassingly parallel independent frames`make -j` — the classic CPU-bound parallelism win; why it only helps when you have real cores, not threadsTrolltech's distributed C++ build system — compile-farm tied to a specific commit, object files cached and sharedJavaScript worker threads and Web Workers — breaking out of the four-query Node.js limit Link to Wolf's dap-mux presentation at mug.org: https://www.youtube.com/live/iyAk8-oE6cM?t=1725 Hosts: Jim McQuillan can be reached at jam@RuntimeArguments.fm Wolf can be reached at wolf@RuntimeArguments.fm Follow us on Mastodon: @RuntimeArguments@hachyderm.io If you have feedback for us, please send it to feedback@RuntimeArguments.fm Checkout our webpage at http://RuntimeArguments.fm Theme music: Dawn by nuer self, from the album Digital Sky

  4. ١٨ أبريل

    26: Why You'll Never Switch Editors (And What You're Missing)

    You already have an editor. You already love it. Nothing we say is going to change that — and we know it. But your editor shapes how you think about editing, which means there are problems it could solve for you that you've never even thought to have. We walk through the real differences between editors, IDEs, and the technologies underneath them — TreeSitter, LSPs, the Debug Adapter Protocol — and talk honestly about what actually matters: syntax awareness that doesn't break, language intelligence that works across editors, and where AI fits into all of it. No editor wars, no winner declared. Just two guys who've been doing this for decades explaining why the landscape looks the way it does. What we cover: The spectrum from Notepad to full IDE — and where VS Code actually fallsModal editing (Vim, NeoVim, Helix) vs. chord-based editing (Emacs) vs. point-and-click (the entire rest of the world)TreeSitter: why regex-based syntax highlighting is broken and what replaced itLSPs: the protocol that turned simple editors into language-aware toolsEditors as complete environments — Emacs, Smalltalk, and the "world" conceptAI integration: editor-first (Cursor, VS Code) vs. AI-first (Claude Code)Muscle memory, sunk costs, and why switching editors is like moving to AustraliaCan JetBrains (or any company that lives on editor/IDE sales) survive when free tools keep getting better? Links: TreeSitter (https://tree-sitter.github.io/tree-sitter/) — incremental parsing library, originally built at GitHub for AtomLanguage Server Protocol (https://microsoft.github.io/language-server-protocol/) — the protocol that decoupled language intelligence from editorsHelix (https://helix-editor.com/) — modal editor with TreeSitter and LSP built inAlabaster theme (https://github.com/tonsky/sublime-scheme-alabaster) — Tonsky's minimalist syntax theme that highlights what mattersXKCD #927: Standards (https://xkcd.com/927/) — the comic about inventing yet another standard (re: IPv8)Hosts: Jim McQuillan can be reached at jam@RuntimeArguments.fm Wolf can be reached at wolf@RuntimeArguments.fm Follow us on Mastodon: @RuntimeArguments@hachyderm.io If you have feedback for us, please send it to feedback@RuntimeArguments.fm Checkout our webpage at http://RuntimeArguments.fm Theme music: Dawn by nuer self, from the album Digital Sky

التقييمات والمراجعات

٥
من ٥
‫٤ من التقييمات‬

حول

Conversations about technology between two friends who disagree on plenty, and agree on plenty more.

قد يعجبك أيضًا