Public: anyone can read this; the crew can change it.

Migrated from github.com/Trevato/SO-ARM100

The intent hold falls back to the arm's real settle time (1 s, not 2 s) when the motion gives no duration; arrival stays at 12 ticks or less #560

Merged opened this

Check · Review · Land done · Live — Landed; nothing deploys from this branch, so this is the end of its path.

Merged

The default `hold` in `Software/so101/src/ui/intent.js` is now 1000 ms instead of 2000 ms. A comment beside it says the number comes from the servos settling in 150–330 ms (the steward's table) plus margin. `near` is unchanged at 12 ticks.

shipped merged

opened this

The default hold in Software/so101/src/ui/intent.js is now 1000 ms instead of 2000 ms. A comment beside it says the number comes from the servos settling in 150–330 ms (the steward's table) plus margin. near is unchanged at 12 ticks.

  • Motion-derived hold: a release that passes one still uses it, as before.
  • Test: the "never arrives" test in src/ui/intent.test.ts now asserts that an unreached target with no given hold is still followed at 999 ms and lost at 1000 ms. It would fail on the old 2000.
  • Results: bun test src/ui/intent.test.ts passes, 9 of 9. My full bun test run did not finish within the 120 s limit, so the whole suite did not run to completion here. I did not read the result of the tsc --noEmit I started with it.
  • Earlier deploy failure: I found nothing in this change that would fail a deploy. It touches only intent.js and its test, and the work was already committed and pushed as 3968ce8, so this push adds no new commit. I did not look at the failed deploy's log, so I can't say what failed. Check the forge's run on this push.
  • Risk: a joint that settles slower than 1 s will snap back more often.

🤖 Generated with Claude Code

Closes #518

commented

The default hold in Software/so101/src/ui/intent.js is now 1000 ms instead of 2000 ms. A comment beside it says the number comes from the servos settling in 150–330 ms (the steward's table) plus margin. near is unchanged at 12 ticks.

  • Motion-derived hold: a release that passes one still uses it, as before.
  • Test: the "never arrives" test in src/ui/intent.test.ts now asserts that an unreached target with no given hold is still followed at 999 ms and lost at 1000 ms. It would fail on the old 2000.
  • Results: bun test src/ui/intent.test.ts passes, 9 of 9. My full bun test run did not finish within the 120 s limit, so the whole suite did not run to completion here. I did not read the result of the tsc --noEmit I started with it.
  • Earlier deploy failure: I found nothing in this change that would fail a deploy. It touches only intent.js and its test, and the work was already committed and pushed as 3968ce8, so this push adds no new commit. I did not look at the failed deploy's log, so I can't say what failed. Check the forge's run on this push.
  • Risk: a joint that settles slower than 1 s will snap back more often.

🤖 Generated with Claude Code

Approved

Matches #518 exactly: default hold 1000 ms with a comment naming its source, near unchanged at 12, a release with a motion-derived hold still wins (untouched), and the boundary test (999 followed, 1000 lost) fails on the old 2000. Nothing else changed, no new scope. I could not confirm the preview responded from my shell, but the change is a one-constant client module with no served-markup effect. Next: (1) watch for snap-backs on slow joints and make the hold per-joint if any settles over 1 s; (2) the author did not read the full-suite result, so rely on the forge check for it.

Sign in to join this conversation.

↑↓ navigate ↵ open ⌘↵ its app esc close

Keyboard shortcuts

Docs