STORY RECORD
VM architect calls out Rust's unsafe hatch
Filip Pizlo, a VM architect with over 8,000 followers, replies to PL researcher Kiran in an escalating exchange about Rust's claim to memory safety.
Pizlo analogizes Rust's unsafe escape hatches to C's pointer and array operators, arguing that if widely-used unsafe blocks are dismissed as optional, C would also qualify as a memory-safe language.
Kiran counters that Rust remains memory-safe because an escape hatch doesn't negate the language's safety guarantees.
Pizlo then quote-tweets his own reply as 'The trouble with saying your language is memory safe if it has widely used escape hatches, in a nutshell.' This exchange is a reaction_example within a larger live conversation, including Zig's Fil-C proposal that claims 'actually memory safe (unlike Rust),' Aram Havarneanu's Curry-style types critique of Rust's unsafe construct, and antirez endorsing making C itself memory-safe.
Why It Matters
This exchange puts a practitioner with deep VM expertise on the record challenging Rust's core marketing claim -- that the language is memory-safe despite widespread use of unsafe blocks in production.
It adds a credible systems-programmer counterpoint to the Rust-as-default narrative at a moment when alternative approaches (Zig Fil-C, Curry-style types, compiler-level C safety) are gaining attention.
The tension is not just academic: if the Rust ecosystem relies on thousands of unsafe blocks in practice, the label 'memory safe' becomes a contested claim rather than a settled fact.
The Facts
5Filip Pizlo (filpizlo), a VM architect with ~8K followers, replied to Kiran (kirancodes), a PL research scientist with ~1.5K followers, arguing that calling a language memory safe despite widely used escape hatches is flawed, using C as an analogy: the escape hatches are '*', '+', and '[]'.
strong · confidence 0.95
Kiran replied that Rust is a memory-safe language and the provision of an escape hatch does not change that fact.
strong · confidence 0.95
Pizlo quote-tweeted his own reply with the comment: 'The trouble with saying your language is memory safe if it has widely used escape hatches, in a nutshell'.
strong · confidence 0.95
The exchange is part of a larger live conversation on X about memory safety, including a Zig Fil-C proposal that claims 'actually memory safe (unlike Rust)' and a critique of Rust's unsafe from former Go compiler engineer Aram Havarneanu proposing Curry-style types.
contextual · confidence 0.75
Pizlo first challenged Kiran with the post: 'Do you have a single actual counter argument to anything I have said?' before Kiran's direct response about Rust's escape hatch.
strong · confidence 0.9
Still Open
3- OpenThe scope of Rust's 'widely used escape hatches' claim is not quantified in these posts; the exchange does not cite specific data on unsafe block prevalence in Rust codebases.
- OpenPizlo's professional background as 'Architekt maszyn wirtualnych' (VM architect) is descriptive but not independently verified beyond his X profile.
- ContradictionContradiction note is limited to snapshotted evidence.contradicted