Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
| @@ -0,0 +1,1558 @@ | |||
| // Bun Snapshot v1, https://bun.sh/docs/test/snapshots | |||
There was a problem hiding this comment.
If you're not used to snapshot testing, this file is automatically generated with --update-snapshots - it's supposed to be checked in. Then when you run tests in the future it checks them against this snapshot. And if it's different than what comes out it errors. If you intend for the change to occur, then you run --update-snapshots again and check in the change. Now all changes to the AST output come with a visible representation of the diff in output.
|
|
||
| ```bash | ||
| # From the package directory | ||
| bun test --update-snapshots |
There was a problem hiding this comment.
Should we let the AI know about this? Could that result in them seeing a failing test and then going oh, well, i'll just update the snapshots to fix the problem?
There was a problem hiding this comment.
these are like an industry standard, so AI should know when to do it and when to not.
Improves the packages/trees file-tree profiler startup path for large pre-expanded/open trees. The change stack removes redundant prepared-input validation, represents profiler workloads with initialExpansion: 'open', slims PathStore node/storage shapes, reduces cold startup writes, tunes visible-child chunk summaries, and removes redundant FileTreeView mount/render work. Experiments: kept runs #1, #16, #18, #24, #25, #26, #28, #36, #41, #42, #43, #44, #51, #65, #67, #68, #71, #74, #80, #82, #137, #139, #140, #142, #143, #147, #149, #153, #155, #163, #165, #171, #180, #182, #186, #196 Metric: visibleRowsReadyMs 334.8ms -> 253.8ms best (-24.2%); final kept sample 267.8ms (-20.0%).
Improves the packages/trees file-tree profiler startup path for large pre-expanded/open trees. The change stack removes redundant prepared-input validation, represents profiler workloads with initialExpansion: 'open', slims PathStore node/storage shapes, reduces cold startup writes, tunes visible-child chunk summaries, and removes redundant FileTreeView mount/render work. Experiments: kept runs #1, #16, #18, #24, #25, #26, #28, #36, #41, #42, #43, #44, #51, #65, #67, #68, #71, #74, #80, #82, #137, #139, #140, #142, #143, #147, #149, #153, #155, #163, #165, #171, #180, #182, #186, #196 Metric: visibleRowsReadyMs 334.8ms -> 253.8ms best (-24.2%); final kept sample 267.8ms (-20.0%).
) Optimize file-tree startup for large open trees Improves the packages/trees file-tree profiler startup path for large pre-expanded/open trees. The change stack removes redundant prepared-input validation, represents profiler workloads with initialExpansion: 'open', slims PathStore node/storage shapes, reduces cold startup writes, tunes visible-child chunk summaries, and removes redundant FileTreeView mount/render work. Experiments: kept runs #1, #16, #18, #24, #25, #26, #28, #36, #41, #42, #43, #44, #51, #65, #67, #68, #71, #74, #80, #82, #137, #139, #140, #142, #143, #147, #149, #153, #155, #163, #165, #171, #180, #182, #186, #196 Metric: visibleRowsReadyMs 334.8ms -> 253.8ms best (-24.2%); final kept sample 267.8ms (-20.0%).
Gettin' it goin'
bun run diffs:test