This is a personal blog. Unless otherwise stated, the opinions expressed here are my own, and not those of my past or present employers.
Intended audience Software engineers working on large source-controlled repositories, particularly Git. Origin Originally presented at Build Meetup Seattle (July 2026) Experience working with Git , git-branchless , and Sapling . Experience optimizing operations on large monorepos. Mood Determined. Hoping to inspire people to rethink the status quo. Table of contents: Slides Links Related posts…
Intended audience Software engineers who already write test plans in commit messages or code review descriptions. People working with patch stacks or stacked diffs. Origin Private correspondence re Julio Merino's post A markdown-based test suite . Mood Practical. Why How Patterns Script Related posts Comments Julio Merino recently published A markdown-based test suite , about using Markdown itself…
Intended audience Linguistics enthusiasts. Autistic individuals working in a professional setting with access to emoji. Occasion April Cools' 2026 — late as usual! Origin Original research from 2025-02-28 Informal surveys from a few of my group chats. Mood Curious. Table of contents: Question Results Responses Chat A (n = 3) Chat B (n = 4) Chat C (n = 2) Chat D (n = 2) Related posts Comments At…
Intended audience Software engineers interested in logic programming or static analysis, and actually using it in production. Origin Originally presented at Seattle Build Enthusiasts Meetup (Autumn 2025) Frustration at work due to not having a hermetic build system. Prior experience with static analysis techniques. Mood Excited to share a useful tool with like-minded colleagues. Table of contents:…
Intended audience Back-end software engineers, particularly in big tech. Origin Experience in big tech working on highly-scalable developer tooling. Sophie Alpert's blog post Materialized views are obviously useful . Mood Frustrated. Determined that we should do better as an industry. Table of contents: Related posts Comments When I was a kid, I wrote a little social networking site for my school…
Intended audience Static analysis enthusiasts Logic programming enthusiasts Origin Recent experience with Soufflé Datalog. Discussion with Max Bernstein about his blog post Linear scan register allocation on SSA , resulting in his blog post Liveness analysis with Datalog . Mood Liberated Table of contents: Soufflé Rapid prototyping The curse of brevity Related posts Comments Soufflé I recently…
Intended audience Build system developers and enthusiasts Source control developers and enthusiasts Origin Attempts to generalize the library behind git test Originally presented at Seattle Build Enthusiasts Meetup (Summer 2025) Mood Instructive Enlightened Table of contents: Slides Speaker notes Related posts Comments Slides This talk was given at the Seattle Build Enthusiasts Summer 2025 meetup…
Intended audience Those with a passing interest in linguistics. Occasion April Cools' 2025 — a little late! Origin Personal experience. Mood Amused. These are some anecdotes from my various experiences learning languages (including Urdu , Polish , Latin , and American Sign Language ). The order of entries is mostly arbitrary. Hopefully you find some of them entertaining! Miscellany We all scream…
Intended audience Build system developers and enthusiasts Developers of tooling that doesn't fit into the build system Origin Lessons synthesized from 8-year career in incremental computation and build systems Originally presented at Seattle Build Enthusiasts Meetup (Fall 2024) Mood Instructive Passionate that people should avoid my mistakes Table of contents: Slides Speaker notes Related posts…
Intended audience Developers of version control systems, specifically jj . Those interested in the version control pedagogy. Origin Reprint of research originally published on Google Docs . Surveyed five participants from various "big tech" companies (>$1B valuation). Mood Investigative. Table of contents: Methodology Results Remarks Conclusions Related posts Comments Methodology Q: I am doing…