Fe and the Mythical Compiler

Duration: NaN:NaN:NaN

Speaker: Micah Fitch

Type: Talk

Expertise: Some familiarity with the topic

Event: Devcon

Date: Nov 2026

An untold potential of compilers is being squandered right when we need it most. Prioritizing lowering above all else leaves rich representations of our programs to waste. All this information that could be used for next generation tooling and formal analyses, why just toss it out when the bytecode arrives? Fe's notoriety as an alternative EVM language overshadows a more interesting story: when the Fe team invested in a query-based, incremental architecture we got comprehensive, fast, high fidelity tooling in return. Fe's language server, code indexing and debug instrumentation are just thin, efficient, maintainable interfaces into the exact same machinery used for lowering and internal analysis. But why stop there, especially now? Let's reimagine compilers not as bytecode emitters but as open graph databases on program representations and the transformations between them, and in doing so unlock deeper analysis, safer code reuse and more fluid interfaces for formal verification.