Open prjemian opened 3 years ago
MyProcessPlugin
can be replaced by https://github.com/bluesky/ophyd/blob/03585f0b02752d5975da393a889e5c5cecd06fd7/ophyd/areadetector/plugins.py#L1504-L1507
That removes pool_max_buffers
via https://github.com/bluesky/ophyd/blob/03585f0b02752d5975da393a889e5c5cecd06fd7/ophyd/areadetector/plugins.py#L1025-L1028
EpicsPathSignal
may need some updating to fix this? https://github.com/bluesky/ophyd/blob/03585f0b02752d5975da393a889e5c5cecd06fd7/ophyd/areadetector/paths.py#L87
On slack.com, it was suggested that #749 might address this directly with a resolution.
A problem occurred when staging an area detector HDF file writing plugin and the IOC is running on Windows.
One feature of any of the AD file writers is the ability to control the directory where the file will be written. When the
file_path
attribute is written, the EPICS IOC will append a/
directory separator to the given file path if it does not have one already. The choice of/
even happens for Windows IOCs!The problem in ophyd is two-fold:
.set()
to this signal waits at this line (with no timeout): https://github.com/bluesky/ophyd/blob/1b9e91be7483d94636ae0bfa14bf13702e07b969/ophyd/areadetector/filestore_mixins.py#L432 since the return value will be different than the expected.Here is the code that fixes the problem for one detector, by making a custom HDF5 Plugin for a Windows IOC (the other support code is added):