Create simple Bluesky acquisition plan independent from DM calls

[qzhang234]

8-ID-I would like to have a simple Bluesky acquisition plan that does only the following:

1. Start the cam module to acquire;
2. Start the HDF plugin to save images.

We would prefer that this Bluesky plan DOES NOT READ/WRITE ANY PV unless specifically asked. This includes staging/unstaging. The beamline scientists will be responsible for setting the correct detector configuration, which includes providing the file name and file path to the HDF plugin. The beamline scientists will also be responsible for any sample motion/temperature control that is separate from the detector acquisition.

In addition, given that the x-ray measurement should not wait on the completion of DM to proceed, we would like to completely separate the DM part from the Bluesky acquisition. This will reduce the complexity and improve the stability of the code, and also makes it easier to debug and maintain.

Please let me know what your thoughts are on this.

[prjemian]

The code for that plan exists within the current full plan: https://github.com/aps-8id-dys/bluesky/blob/d42eb4d72fbb659d184df8406408c0ab0de53d1f/src/instrument/plans/bdp_demo.py#L356-L380

[qzhang234]

That looks very nice. I have two questions regarding that function:

1. Is it absolutely necessary to return the uuid of bp.count?
2. Can bp.count work with an empty stage_sigs?

[prjemian]

@qzhang234 Should this plan accept arguments for the acquisition parameters? (Lines 367-373).

1. Is it absolutely necessary to return the uuid of bp.count?

Yes.

[prjemian]

2. Can bp.count work with an empty stage_sigs?

Should be fine. Various detector features (that would have been configured by staging) should be setup before this plan is called. Such as enable the HDF plugin and prepare it to record an image to a named file.

[qzhang234]

@qzhang234 Should this plan accept arguments for the acquisition parameters? (Lines 367-373).

1. Is it absolutely necessary to return the uuid of bp.count?

Yes.

@prjemian I think just the bp.count with HDF plugin should be fine for now. I'll wrap it with another function where I pass all the variables manually. Would this be possible?

[prjemian]

Looking like this in my editor.

[prjemian]

What about the nxwriter here?

[prjemian]

If the nxwriter is dropped, the plan reduces to bp.count([det], md=md)