solidjs / solid / 40756949756

78%

main: 89%

push

github

docs(signals): F-6 — the parent waits for a lane's flights; the lane does not wait for the parent

The 2026-10-09 restatement wrote the lane hold as "separate from the
parent hold on both sides". Only one side is sourced: the lane reveals at
its own landing (#3334, #3460; verdict-contract, lane-outside-view). The
other side — the parent landing while the lane's rendered derivation is in
flight — has no ruling behind it and three pins against it: the 2026-10-01
"a blocked lane blocks its parent" ruling (lane-contract.test.ts), the
pre-L2 "pending source does not leak" pin (createOptimistic.test.ts), and
today's nestedBlocked for verdict lanes. Landing the write beside a
derivation of it still in flight is not a whole frame (F-1), and latest()
cannot show a value behind the screen (D-22).

F-6 now says the parent waits for a lane's flights under the same F-2/F-3
test as for any flight; a boundary catching it (F-4, D-21) or its reader
leaving the display releases it. Closes the reader-seat design's R2:
I20 stands as is.

Co-authored-by: Claude <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>

788 of 1081 branches covered (72.9%)

Branch coverage included in aggregate %.

1124 of 1354 relevant lines covered (83.01%)

29.93 hits per line