Our lathe suppliers need separate meanings in the bush-loading plan. Machine side says ready when it receives a request; integration side reads loading area available. Same signal direction, different event. How do I get a usable agreement?
Give each side the other definition in writing. Ask what their equipment does when the request is received but the loading area is unavailable. That case cannot be represented by a shared word without more conditions.
Both sides accept that receipt can happen while the lathe is occupied. The machine supplier proposes a separate availability indication. We haven't agreed how long either indication remains valid.
Lin, that belongs in the agreement. Retaining a signal is not the same as retaining enough context to use it. Carla's teams should describe restart on either side, not assume both restart together.
We added separate restart cases. The machine keeps a request context that the integration proposal currently loses, so the mismatch is more than naming. Both leads are revising the recovery entry together.
Useful finding. Keep that inconsistent starting condition in their test plan after they revise the interface; a clean-start demonstration will not show whether they fixed it.
Normal receipt and availability are agreed now. Restart reconciliation is still outstanding, with the retained-machine-context case explicitly attached. I won't label the whole interface agreed on the strength of the normal sequence.