Skip to content

check: a stop shorter than 0.4 s is not reported #23

Closed

Description

@umiyosh

Problem

motion-render check does not report a mistake that lasts less than 0.4 s.

Seen while rebuilding the WIF film: a token that passes the attribute condition stopped on top of the condition box (#cond) for 0.3 s and covered both lines of the condition. check reported layout: ok. With the same pose held for 1.1 s, check reports it (div "assertion.repository == " is hidden under #good, and the ✓ partly overlapping the condition lines).

Why it is missed

checkLayout samples every 0.2 s, and findLayoutIssues reports a finding only when a still box shows it in 2 samples in a row. A box counts as still only once it has not moved since the previous sample. A 0.3 s stop is therefore still in a single sample at most, so it is never reported. The 2-sample rule is there so that a passing pose is not reported, but at 0.2 s it also hides short stops that a viewer does see.

Proposal

Sample every 0.1 s, keeping the 2-sample rule. A stop of 0.2 s or longer is then reported, and a chip in flight is still skipped. The check takes about twice as long (about 3 s → 6 s for a 57 s film).

Done when

  • A fixture with a chip that rests 0.3 s on a card's text is reported
  • The WIF and population films still pass

Activity

  1. changed the title [-]check: a chip resting on a card's text is not reported[/-] [+]check: a stop shorter than 0.4 s is not reported[/+] on Oct 25, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions