The log's lifecycle follow-up reads a lane snapshot nobody can search as no snapshot, and can say "another act got there first" when no other act ran (follow-up to #309) #317

From lane 4's #309 design (6350724542, its R6) and lane 3's review of it (#309 6352329119, ruling R6). It is filed on Brett Heap's approval, "File a v1.2.0 issue now (Recommended)", given in lane 3's session on the morning of 2026-10-23.

#309 fixes this reading in set-lane-state, set-lane-tree and rename-lane. This issue is the same reading in the follow-up that moves a lane's snapshot after a lifecycle line. #309 leaves it out because act 2 rewrites the functions it lives in.

What is wrong (at 33ce26d)

These facts come from reading the code, and lane 3's reviewer confirmed them by reading. Nobody has run them.

• lane_state_fingerprint (lanes-edit.sh:13629–13635) answers none/0/none, which means "this lane has no snapshot", for any file that [ -e ] cannot see. A lane-state.yaml under a directory nobody can search is one of those.
• lane_state_preimage (:13773–13778) also answers none/0/none when lane_control_root fails, whatever the reason.
• The two reads. write_event takes the pre-image before its line lands (:6131), and retire_rows takes one for each lane it sweeps (:12901). lane_state_follow (:13817–13867) then reads the snapshot again under the mutex, and compares the two answers (:13847–13848):
  • when both reads are blocked, the answers agree. The follow-up tries to write, the write fails, and its note says so (:13865);
  • when only one read is blocked, the answers differ. The note then says "Another act got there first — a recovery, or a second handoff" (:13850), which is false. Nothing is written.
• lane_state_follow returns without a word when lane_control_root fails (:13830). So a blocked lookup reads as a lane with no control root, which is the ordinary case before the cutover.
• The three functions are byte-identical at 3e97ae5, A lane control root that cannot be searched reads as "no trees", so an inventory sync removes its rows (from #289's review) #291's branch, where they start at :13695, :13843 and :13887.

Direction

• Tell "cannot be seen" from "not there" in all three functions, with A lane control root that cannot be searched reads as "no trees", so an inventory sync removes its rows (from #289's review) #291's lookup_blocked. A blocked read is neither none/0/none nor a match.
  • The follow-up then writes nothing, and names the blocked directory as A control root nobody can search reads as an empty one in two writers: rename-lane leaves the lifecycle records behind without a word, and set-lane-state's fence says 7 for a race that never happened (from #291) #309's notes do.
  • It says the snapshot is one act behind until the next transition.
  • It never says that another act got there first.
• Keep the follow-up's contract: it never fails the event. log's exit status does not change, and the new words are notes on stderr. The callers this reaches are:
  • lane-start:4178, its log call;
  • lane-end:2487, its log call;
  • retire-rows.
• Order. It lands after act 2's slice 1a (Every register write goes through its lane's own checkout <W>-lanes/<lane>/, where lane-start and lane-end read the lane's row; a LANES_REPO naming W refuses; locks name their holder's container; refresh-workspace fast-forwards W (Amendment 23) (#198) #282), which rewrites lane_state_follow, and slice 1b (Amendment 23 slice 1b: every writer fetches and fast-forwards under its lock or refuses with nothing written, pushes with no pull and reads origin back, a refused push replays the one edit, and commit with nothing staged is the publish act (#198) #307), which rewrites write_event. It also lands after A lane control root that cannot be searched reads as "no trees", so an inventory sync removes its rows (from #289's review) #291 and A control root nobody can search reads as an empty one in two writers: rename-lane leaves the lifecycle records behind without a word, and set-lane-state's fence says 7 for a race that never happened (from #291) #309, whose helper and wording it reuses.
• Review. lanes-edit.sh is lane 3's tooling, so lane 3 reviews the design before any code. Lane 3's refined lock-lines term (Triad migration (feature 004-migrate-to-triad): execution tracking under Brett Heap's word of 2026-10-07 #186 6352336244) applies.

Not started: v1.2.0 work waits for the v1.1.0 tag and Brett's word.