Both FR5 interface offers exclude the request-clearing decision

Both agree to attend; the PLC author thought the application latched the pulse, while the application author assumed the request remained present until receipt, which explains the contradictory notes.

Give them a state table with an owner, setting event and clearing event for each signal, plus the retained attempt identity. Work through a late receipt and either side restarting. The ordinary successful exchange will not expose the assumption that caught you.

Who changes the operator wording after they agree?

Elliot's restart case is worth keeping in the scope too. Otherwise it can emerge later as a separately chargeable surprise nobody says they were asked about.

Amara, the application author owns the screen and operator notes; Robin, purchasing has included those exception cases and the joint verification work in the scope request.

Make timeout a named unresolved exchange, not a fresh request by default. Its late receipt still belongs to the earlier attempt.

Does the normal screen show which attempt is waiting?

The draft now does, with no receipt separate from waiting for an inspection result; the two authors have also agreed who retains and clears the request and receipt for that same attempt.

Joint walkthrough found a late receipt could still clear the next request in the proposed logic; that rule is corrected in the design, and installed verification remains explicitly outstanding in both offers.

Useful catch. Has the verification owner accepted the revised cases and cost, rather than only the authors agreeing the design?