Closed klauer closed 2 years ago
I think this looks correct to me, and should work for both the pyepics and caproto backends.
Do you remember exactly which device was throwing this error so we can test it? If not, I can start opening all the screens again.
I don't recall the device, but it had a PytmcSignalRO
if I recall correctly, so that should narrow it down.
That was enough to find it: the devices are rtdsl0
and rtdsk0
, whose subdevices have a state
setpoint that is a PytmcSignalRO
... so they don't function as movable state devices. I think they need an additional fix in pcdsdevices
but this probably removes the terminal spam error (I'll check)
The aforementioned screens are very happy with this PR. We can check if they would actually move their respective devices over the PAMM but I think this should be merged.
Nice - OK! 👍 This was totally untested, so thanks for the test run.
ah right I'll just make sure the filters move on the screen after lunch and then press merge
OK the device on the pcdsdevices side definitely needs a looking at but this is pretty good for all normal devices and it removes the error
Description
_write_pv
and_read_pv
are generally good assumptions forEpicsSignalBase
and its subclasses. However, typhos should be lenient if/when access to these attributes fails.pcdsdevices may also require adjustments to
PytmcSignalRO
, but that shouldn't affect this PR.Motivation and Context
Closes #474 Closes #473
How Has This Been Tested?
Test suite and interactively with problematic
PytmcSignalRO
signal.Where Has This Been Documented?
Issue and PR text.