[miguel-heygen]

What

Two exports of one project started in the same second now get their own job and their own output file. The second takes a -2 suffix (then -3, and so on) when the id is already a running job or its file already exists.

Why

POST /api/projects/:id/render built the job id as the project id plus a timestamp with one-second resolution, and that id is also the output file name. Two exports within a second shared a key in the server's job map, so the second replaced the first job and both wrote the same file. The id is minted in one place, packages/studio-server/src/routes/render.ts. A single export keeps the same id and file name as before.

Testing

fails without the fix: packages/studio-server/src/routes/render.test.ts > "gives two exports started in the same second their own job and output file"

On the devbox: 90 passed in render.test.ts, 3 runs in a row; with the old render.ts the new test fails (1 failed, 89 passed).

Size

A short change to one id and its test; it has no other carrier.

[heygengenesis]

Requesting changes on e369f3e1 for one thing: the Comments check fails because of this diff.

RAISED comment share: packages/studio-server/src/routes/render.ts is 10.9% (45/414), at the base 10.5% (43/409) — 2 comment line(s) added.

The ratchet won't let a file's comment share go up, and the new two-line comment above baseId is what pushes it over. Delete the comment, or move what it says into a name, for example a small nextFreeRenderJobId(baseId, ext) helper. Once that check passes, I don't see another blocker.

The rest checks out:

• Fix: the id is chosen and renderJobs.set runs in the same synchronous step (adapter.startRender returns synchronously and nothing is awaited in between), so two concurrent POSTs can't both get the same id. Checking for an existing ${jobId}${ext} also stops a file from a restarted session being overwritten. A single export keeps its old id and file name.
• Consumers: nothing in studio/studio-server parses the timestamp back out of the id. The list, delete, view and download routes key on the full id, so a -2 suffix works with all of them.
• Fails without the fix: I confirmed it locally. At head, render.test.ts has 90 passing, two runs in a row. With the base render.ts and the new test, the named test fails (expected 1 to be 2, 1 failed / 89 passed).

[github-actions]

Edit accuracy: accurate 267 (base branch 267), smooth 97 of those

The gate passes.
Smoothness is reported in the artifact, not gated. A case fails only if it fails 2 of 3 runs.

Quarantined, measured but not gated (8)

• seqplainkeys-none-center-r0-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-center-r0-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): fail / pass / pass, passes
• seqplainkeys-none-center-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-pct-r0-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-pct-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-px-r0-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): pass, passes
• seqplainkeys-none-px-r30-nested-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run
• seqplainkeys-none-px-r30-root-z100 (fixed by studio: undo/redo on a keyframed plain element sometimes leaves the box 230-400px off (edit-accuracy seqplainkeys-none-*) #6989): not run

Unstable (1)

• seqfastkeys-none-px-r0-root-z100: tracking 0.07, pressJump 0.05, drop 0, reload 0, render 0.02, renderKey -, undo false, teleport true / tracking 0.06, pressJump 0.05, drop 0, reload 0, render 0.02, renderKey -, undo true, teleport true / tracking 0.06, pressJump 0.05, drop 0, reload 0, render 0.02, renderKey -, undo true, teleport true

[heygengenesis]

Approving at e7c88ab3. The only blocker from the last review is resolved: the comment is now a named helper, nextFreeRenderJobId, and the Comments check passes on this head.

The change since e369f3e1 only moves the loop into that helper. It works the same way, and the id is still chosen and stored in the same synchronous step, with nothing awaited in between.

I reran the named test on this head. render.test.ts passes 90/90. With the base render.ts, "gives two exports started in the same second their own job and output file" fails (1 failed / 89 passed).