Field Notes
-
Aliasing Xor Mutability: A New Way to Learn Rust?
Many readers or one writer: using aliasing XOR mutability to reason about Rust's ownership and borrowing rules.
-
Bugs Rust Won't Catch: Lessons from the uutils/coreutils CVE Audit
A professional audit of the Rust reimplementation of GNU coreutils uncovered 44 CVEs — none caught by the borrow checker, Clippy, or cargo-audit. Here's what to watch for.
-
Enter MXC, Part 2: The Fence I Didn't Build
Part two of the MXC field report: before wiring containment into an agent harness, I mapped every place it actually executes something — and discovered my plan had a category error. What a review pass caught, what stayed open, and why understanding had to come before code.
-
Enter MXC, Part 3: The Fence Goes In
Part three of the MXC field report: the fence gets built and shipped, and the shell question answers itself — on Windows, bash cannot exist inside an MXC container at all. Plus the launch failure that every green test missed, because the tests were checking a fixture instead of the operating system.
-
Enter MXC: What Process Containers Even Are
Part one of a field report on Microsoft's MXC: what process containers actually are (AppContainer identities, declarative policy, learn-versus-enforce), explained from the underserved platform, doubts included.
-
De-mystifying `if-let`
How to think about the `if-let` syntax
-
De-mystifying JSON
How to look at a JSON shape and translate it to Rust structs and enums.
-
De-mystifying Slices (with a Table of Test Cases)
When to reach for a slice, what the two ampersands mean, and a table-of-test-data pattern that finally made them click.
-
De-mystifying Generics
One struct definition, many concrete types: where Message<String> comes from, and why it won't share a Vec with Message<i32>.