This repository was archived by the owner on Sep 1, 2026. It is now read-only.
Repository navigation
Independent cross-platform benchmark of TypeScript 7.0.1 RC vs TypeScript 6.0.3 #4410
Unanswered
Aleksey Safonov (safal207)
asked this question in
Q&A
Replies: 1 comment
|
You also might try passing a larger value to |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
This is an external benchmark report, not a performance regression report.
I built a reproducible harness comparing "typescript@7.0.1-rc" with "@typescript/typescript6@6.0.3" on GitHub-hosted Ubuntu x64, Windows x64, and macOS arm64 runners.
Workloads
Each scenario used 2 warm-up rounds and 15 measured rounds in a deterministic randomized order. The report includes medians, P90/P95, coefficient of variation, MAD, IQR outlier counts, and non-parametric 95% bootstrap intervals. Outliers were reported and retained.
TypeScript 7 default median speedup ranges
The defensible claim is limited to the documented fresh-process, warm-filesystem protocol; this is not intended as a universal TypeScript speed claim.
Output verification
After CRLF-to-LF normalization, TS6 and TS7 produced identical paths and SHA-256 content hashes on every tested OS for:
".tsbuildinfo" files were excluded as compiler implementation metadata.
The project-reference graph was intentionally parallelizable. Increasing "--builders" from 1 to 4 reduced median build time on all three platforms, while the TS7 default remained close to the best observed configuration.
Reproduction and full results
One remaining limitation is that the workloads are generated. A useful next step would be running the same protocol against one or more pinned real-world repositories.
All reactions