Twinned from https://github.com/cfsh/ninefold/issues/448 by tools/beads/import_github.py's reconcile pass.
#442 put the NAME THE CHANNEL ring in front of the ＋ new pill, so tools/probes/row-label-wrap.json could no longer reach an auto-named channel. It was repointed at the ring and now asserts BASS on row 4 — which keeps the probe honest about row-label RENDERING, but drops the thing the file is named for.
Every ring option is a short pool word (bass, lead, pad, arp…), so no ring path produces a name long enough to wrap onto two lines. The wrap is still reachable in the product via the ring's ⌨ type name wedge; it just has no probe.
What is still covered: PennantInkMathTests covers the wrap MATHS. What is not: that a long name actually draws on two lines in a row label, which is what #366 filed and what the SILKEN/CHESTNUT assertion pinned.
Options:
- Drive the virtual keyboard in the probe (~30 steps of focus+accept per character; deterministic but tedious, and letter keys are single characters so focus matching by text is fragile).
- Add a probe step that sets a name directly, the way expect_prop generalised assertions.
- Seed a demo song containing one deliberately long channel name and assert against that — cheapest, and it also gives every other probe a wrap case for free.
Filed unclaimed. Raised by ninefold-reviewer on PR #447 (finding F1's tail) and by the #442 design note that predicted exactly this: *"If the pill path no longer auto-names, that probe's premise changes."*
None.
No comments.