Instrument Scripts: Test on LOQ [timebox: 1 day]

[DominicOram]

As a SANS instrument scientists I would like the sans script in InstrumentScirpts to work to my satisfaction. To make sure everyone understand what is meant by this a selection of instrument scientists and IBEX developers will spend some time on LOQ on the week beginning 9th November to discuss the requirements and fix the script so that they work.

Acceptance Criteria

• The sans scripts work to the satisfaction of the instrument scientist
• Information on the requirements of the sans scripts and what they are expected to do is written up

[kjwoodsISIS]

• The sans scripts work to the satisfaction of the instrument scientist

This is not a good acceptance criterion. The "satisfaction of the instrument scientist" is not a measurable objective quantity. As it stands the acceptance criterion is completely open-ended and cannot be estimated. This ticket feels like a "spike" - it's about finding something out/researching something we don't yet understand.

I think it would be better to do the following:

• set a timebox
• work with the scientists to identify and document the shortcomings of the scripts
  • this should include getting a clear statement of what each script is supposed to do (i.e. the second acceptance criterion)
• fix any that can be fixed within the duration of the timebox
• create tickets for the remainder

[DominicOram]

I agree @kjwoodsISIS. I will discuss with scientists how much time we'll have on the beamline and timebox it based on that.

[FreddieAkeroyd]

2 people for 1 day plus writeup

[DominicOram]

@ThomasLohnert, @RaiBishal and I attended the meeting. The crux of the issue on LOQ was that it wasn’t up to date with the latest scans library. We’re going to do the following:

• Add a step so that when we update IBEX in general we also make sure to update the scans library
• Correct the documentation for the scan command (on LOQ it didn't have the correct arguments from Add kwargs to scan for pixel_range and min/max_pixel. rename detector… ISISNeutronMuon/InstrumentScripts#77 e.g. det)

Given the meeting was significantly less than the 1 day timebox we can do these things in this ticket

[John-Holt-Tessella]

People from other instrument can update the shared instrument scripts did we really want to update it, or is this just on LOQ? It is possible for one instrument to break another. As long as everyone knows the situation I am happy

[DominicOram]

I think we need to more actively manage it. We can have a discussion if you like?

[DominicOram]

In discussion with @John-Holt-Tessella, @Tom-Willemsen, @ThomasLohnert and @RaiBishal the following was decided:

1. When doing a release we will tag a release of InstrumentScripts and release it to all instruments, whether they currently use the library or not
2. We will write an automated service to tell us if instruments have local changes (or if not on the release tag any more)
3. If instruments do have local changes we will discuss with them getting the changes into master for the next release
4. We will block pushing to master, instead requiring people to make PRs which we will review and check for unexpected side effects to the best of our ability. Other instrument scientists are welcome to contribute to this review process

The acceptance criteria of this ticket is now to:

• Confirm the above process is ok within the Experiment Controls group
• Confirm it is ok within the SANS/reflectometry groups (Emailed to instrument scientists, they asked for clarification on release numbering and how they would revert mistakes. Otherwise had no major reservations)
• Document the process