Five Proposals Hit TC39 Stage 3 in May: Error Stacks, Iterator Helpers, RegExp Anchors
The May 2026 TC39 meeting advanced five proposals to Stage 3: standardized error stack access, three iterator additions (chunking, includes, join), and RegExp buffer boundaries, joining earlier arrivals Source Phase Imports and Import Text.
The May 2026 TC39 meeting advanced five proposals to Stage 3, where implementations can start. Error Stack Accessor standardizes reading a stack trace as data. Today you parse error.stack strings with regexes that break the moment a browser changes format. A standard accessor means your error reporting and grouping survive browser updates.
Three iterator additions landed together. Chunking groups a stream into fixed-size batches, so you stop writing manual accumulator loops for API pagination and rate-limited requests. Includes gives you a cheap membership check over a one-shot iterator without collecting it into an array first. Join concatenates iterator values with a separator, deleting the spread-and-join pattern. RegExp Buffer Boundaries adds \A, \z, and \Z anchors, giving you precise string boundaries without lookahead gymnastics in multiline matching.
Two earlier arrivals are worth watching alongside the batch. Source Phase Imports, Stage 3 since 2023, lets you import a module's source without evaluating it, which kills the import.meta.resolve and fs-read workarounds bundlers and SSR tooling lean on today. Import Text, which advanced in March 2026, brings import ... with { type: 'text' } for loading file contents as strings with no build step.
No Stage 4 moves confirmed in this batch, and Dynamic Code Brand Checks stays at Stage 3. None of this ships in your runtime yet, but Stage 3 is the signal that polyfills get written and engines start building.