Vendor AggregateSignal from pcdsdevices to remove pcdsdevices build dependency
Motivation and Context
We should remove circular dependencies
Issues encountered while playing test-suite-whackamole:
ophyd cleanup via ophyd.cl.dispatcher.stop() seemed to break later tests, which do not restore this cl. There was error spam that led to its initial inclusion, but removing the no-arg test fixes that (I think)
the no-arg test would try to pull a HAPPI_CFG, which doesn't exist in CI and is functionality possibly better tested in happi than lightpath
locally, many intermittent failures were fixed by the inclusion of time.sleep calls after changing signals, presumably to allow callbacks to process and LightApp to update. These delays were arbitrary and the test suite passed here, so I haven't included them. (This also doesn't address the pcds-envs issues zach found, which I haven't been able to replicate)
EPICS got mad at the test suite trying to repeatedly change the EpicsSignal timeout, so I monkeypatched it in the test suite
Description
Vendor
AggregateSignal
frompcdsdevices
to remove pcdsdevices build dependencyMotivation and Context
We should remove circular dependencies
Issues encountered while playing test-suite-whackamole:
ophyd.cl.dispatcher.stop()
seemed to break later tests, which do not restore this cl. There was error spam that led to its initial inclusion, but removing the no-arg test fixes that (I think)time.sleep
calls after changing signals, presumably to allow callbacks to process andLightApp
to update. These delays were arbitrary and the test suite passed here, so I haven't included them. (This also doesn't address the pcds-envs issues zach found, which I haven't been able to replicate)How Has This Been Tested?
test suite
Where Has This Been Documented?
This PR