Release History¶

v2.3.1 (2020-12-22)¶

Fix issue where events=-1 was a rejected argument rather than a valid configuration for infinite running.

Add hutch_name argument to Daq class to allow bypassing of the external get_hutch_name script.

Change bykik code to False by default in set_filter as per Silke’s instructions
Give up on the run_number getter on the first failure per session, it is merely cosmetic and should not slow down the scan.
Decrease the get run number timeout to 1s.
Copy the good docstring for set_filter over to the daq method
Increase timeouts for get_hutch and get_ami_proxy, then cache the result. The result does not change.
Allow controls arg to accept pseudopositioners
Allow “tri-state” configuration args (True, False, None), and allow the Daq class to know whether the user has passed in None versus not passed in anything at all
Allow record=None to mean “Use the record option selected in the DAQ GUI”
Increase the begin timeout to 15s
Add more pointed error messages to the wait timeouts

Fix issue with ami proxy script changes, making the library compatible with both the old and the new output.

Compatibility for ophyd=1.5.0

Fix issue with the begin timeout where the clock started ticking too early.

Fix issue with pcdsdaq_lib_setup that broke hutch environments.

Fix issue where daq would fail to load for det and tst hutches.

Fix issue where the daq could rapidly cycle under specific conditions
Fix issue where the daq class wasn’t correctly reset after a disconnect
Fix issue where deprecated platform argument was mistakenly left in at 2.0.0
Fix issue where module was broken on newest bluesky because None is no longer a valid data shape

Add AmiDet interface for interacting with pyami
Clean up and update pcdsdaq_lib_setup to work with pyami
Allow daq to be configured to run “forever, until everything else is done triggering” by setting events=0
Add configurable sleep time to account for the difference between the daq process claiming to be ready and actually being ready (daq.configure(begin_sleep=0.5))

Allow ctrl+c during a begin call with wait=True to stop the run.
Add sourcable pcdsdaq_lib_setup script that will get pydaq and pycdb ready for your python environment.
The connect method will provide more helpful error messages when it fails.
Allow the Daq class to be used as a bluesky readable device. Once staged, runs will end on run stop documents. A calibcycle will be run when the Daq is triggered, and triggering will be reported as done when the Daq has stopped. This means it is viable to use the Daq inside normal plans like scan and count.
Add an argument to Daq.begin: end_run=True will end the run once the daq stops running, rather than leaving the run open.
Add Daq.begin_infinite
Add Daq.config_info
Restore daq state after a bluesky plan, e.g. disconnect if we were disconnected, run if we were running, etc.
Add support for scan PVs via the ScanVars class. This class attaches itself to a RunEngine and knows when to update each PV, provided that the plan has reasonable metadata.

calib_cycle and related plans module has been removed, as using the Daq as a readable device is more intuitive and it’s still early enough to break my API.
daq_wrapper and daq_decorator have been move to the preprocessors submodule, as a parallel to the bluesky structure. They have been renamed to daq_during_wrapper and daq_during_decorator as a parallel to the built-in fly_during_wrapper. These are now simple preprocessors to run the daq at the same time as a daq-agnostic plan.
complete no longer ends the run. This makes it more in line with the bluesky API.

The daq no longer needs to be passed a platform argument. This argument will be removed in a future release, and will log a warning if you pass it.

Add the record option to the begin method. This allows a user running interactively to concisely activate recording for single runs.

Add daq.record property to schedule that the next run sould be configured with record=True