duan_jandClaude Fable 5.1
5c1e06376e
CI / lint (push) Skipped
CI / test (3.12) (push) Skipped
CI / test (3.13) (push) Skipped
CI / test-with-beamline-plugins (pxi_bec) (push) Skipped
CI / test-with-beamline-plugins (pxii_bec) (push) Skipped
CI / test-with-beamline-plugins (pxiii_bec) (push) Skipped
CI / lint (pull_request) Successful in 1m17s
CI / test-with-beamline-plugins (pxii_bec) (pull_request) Successful in 1m36s
CI / test (3.13) (pull_request) Successful in 1m41s
CI / test (3.14) (pull_request) Successful in 1m42s
CI / test (3.12) (pull_request) Successful in 1m42s
CI / test-with-beamline-plugins (pxiii_bec) (pull_request) Successful in 1m44s
CI / test-with-beamline-plugins (pxi_bec) (pull_request) Successful in 1m45s
CI / test-with-coverage (pull_request) Successful in 1m52s
CI / coverage-analysis (pull_request) Successful in 3s
The "apply after mounting" checkbox clicked Auto Centering as soon as the status poll showed a new sample id. /sample/mount still holds the hw lock at that point, /alc/center_loop takes the same lock non-blocking, and the resulting BeamlineBusyException is a quiet error the GUI swallows. So the automatic click never did anything; only a later manual click worked. Arm on the sample-id change and click on the first status tick with busy=False, which is exactly what the user did by hand. Skip arming while queue automation runs: it mounts and centers on its own and frees the lock between /scan/auto calls, so a GUI centering in that gap would steal the lock from the next scan. Pause during a scan's mount phase can still arm that one sample, but Pause ends in an unmount, so the window is a ms-wide lock gap sampled at 2 Hz; accepted, noted in the code. Supersedes the three-commit approach on fix/auto-center-after-mount-wait-idle (PR 243). Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>