Repository navigation
chore(site): upgrade to Vite 8 - #23485
Merged
Merged
Conversation
DanielleMaywood
commented
Mar 24, 2026
Comment on lines
-70
to
78
| const [first, ...others] = WorkspaceTimingsResponse.agent_script_timings; | ||
| const failedScriptTimings = WorkspaceTimingsResponse.agent_script_timings; | ||
| export const FailedScript: Story = { | ||
| args: { | ||
| agentScriptTimings: [ | ||
| { ...first, status: "exit_failure", exit_code: 1 }, | ||
| ...others, | ||
| { ...failedScriptTimings[0], status: "exit_failure", exit_code: 1 }, | ||
| ...failedScriptTimings.slice(1), | ||
| ], | ||
| }, | ||
| }; |
Collaborator
Author
There was a problem hiding this comment.
Something in the dependency chain is unable to handle const [first, ...others] = WorkspaceTimingsResponse.agent_script_timings;. Unfortunately it generates invalid code
DanielleMaywood
force-pushed
the
danielle/vite-8-upgrade
branch
6 times, most recently
from
April 1, 2026 14:37
ea45fc7 to
8ae16b5
Compare
Upgrade the frontend build toolchain to Vite 8, which replaces the dual esbuild+Rollup pipeline with the unified Rolldown bundler. Package changes: - vite 7.2.6 → 8.0.2 - @vitejs/plugin-react 5.1.1 → 6.0.1 - vitest 4.0.14 → 4.1.1 - @vitest/browser-playwright 4.0.14 → 4.1.1 - vite-plugin-checker 0.11.0 → 0.12.0 - rollup-plugin-visualizer 5.14.0 → 7.0.1 - @storybook/* 10.2.10 → 10.3.3 Config changes: - Rename build.rollupOptions to build.rolldownOptions - Migrate manualChunks to codeSplitting.groups Production build time drops from ~56s to ~12s (~4.8x faster).
DanielleMaywood
force-pushed
the
danielle/vite-8-upgrade
branch
from
April 1, 2026 14:42
8ae16b5 to
affc30b
Compare
jakehwll
approved these changes
Apr 1, 2026
jakehwll
left a comment
Collaborator
There was a problem hiding this comment.
Script available on-request but this seems to be doing things faster and more consistently 🙂
% pnpm run bench:vite-upgrade -- --base main --candidate danielle/vite-8-upgrade --runs 5 --warmups 1
> coder@0.0.0 bench:vite-upgrade /home/coder/coder
> node scripts/benchmark-vite-upgrade.mjs -- --base main --candidate danielle/vite-8-upgrade --runs 5 --warmups 1
Vite upgrade benchmark
repo: /home/coder/coder
base: main
candidate: danielle/vite-8-upgrade
runs: 5, warmups: 1
install: yes
temp worktree root: /tmp/vite-upgrade-bench-ZMvt6Z
[base:main] pnpm install --frozen-lockfile
[base:main] warmup 1/1: 49212.3ms
[base:main] run 1/5: 46274.3ms, 134.65 MiB
[base:main] run 2/5: 74953.8ms, 134.65 MiB
[base:main] run 3/5: 69332.3ms, 134.65 MiB
[base:main] run 4/5: 48958.7ms, 134.65 MiB
[base:main] run 5/5: 43710.6ms, 134.65 MiB
[candidate:danielle/vite-8-upgrade] pnpm install --frozen-lockfile
[candidate:danielle/vite-8-upgrade] warmup 1/1: 35982.2ms
[candidate:danielle/vite-8-upgrade] run 1/5: 16879.2ms, 133.44 MiB
[candidate:danielle/vite-8-upgrade] run 2/5: 16820.8ms, 133.44 MiB
[candidate:danielle/vite-8-upgrade] run 3/5: 16368.9ms, 133.44 MiB
[candidate:danielle/vite-8-upgrade] run 4/5: 16698.7ms, 133.44 MiB
[candidate:danielle/vite-8-upgrade] run 5/5: 16406.3ms, 133.44 MiB
Base (main)
avg build time: 56645.9ms
min/max build time: 43710.6ms / 74953.8ms
avg output size: 134.65 MiB
min/max output size: 134.65 MiB / 134.65 MiB
Candidate (danielle/vite-8-upgrade)
avg build time: 16634.8ms
min/max build time: 16368.9ms / 16879.2ms
avg output size: 133.44 MiB
min/max output size: 133.44 MiB / 133.44 MiB
Delta (candidate - base)
build time: -40011.2ms (-70.63%)
output size: -1.21 MiB (-0.90%)
DanielleMaywood
marked this pull request as ready for review
April 1, 2026 15:04
DanielleMaywood
enabled auto-merge (squash)
April 1, 2026 15:06
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Summary
Upgrade the frontend build toolchain to Vite 8, which replaces the dual esbuild+Rollup pipeline with the unified Rolldown bundler.
Package changes
vite@vitejs/plugin-reactvitest@vitest/browser-playwrightvite-plugin-checkerrollup-plugin-visualizer@storybook/*@rolldown/plugin-babelConfig changes
build.rollupOptionstobuild.rolldownOptions.manualChunksto RolldowncodeSplitting.groups.@vitejs/plugin-reactinlinebabeloption to standalone@rolldown/plugin-babelwithreactCompilerPreset. Babel was removed as a dependency in plugin-react v6; the react-compiler integration now runs as a separate Rolldown plugin.const [first, ...rest] = arr) don't get properly declared as variables.Build performance
Measured on the same machine, back-to-back, three runs each.
End-to-end production build (includes
vite-plugin-checkerTypeScript checking):Bundler-only (checker disabled, isolates Rollup→Rolldown improvement):
The gap between the bundler-only and end-to-end numbers comes from two sources:
@rolldown/plugin-babeladds ~8s runningbabel-plugin-react-compileronsrc/pages/AgentsPage/. This is the cost of migrating from the old inlinebabeloption (which ran inside plugin-react's own transform pipeline) to a standalone Rolldown babel plugin. As more of the codebase opts into the compiler, this cost will grow, but it can be mitigated by narrowing the filter or by switching to the SWC-based compiler plugin when available.vite-plugin-checkerruns TypeScript in a worker thread during the build. On Vite 7, the ~48s Rollup bundling dominated so the checker finished for free; on Vite 8, the bundler finishes in ~11s and then waits for the checker, adding ~3s of visible overhead.