Instruments: handle local changes to calibration files

[FreddieAkeroyd]

Currently we have an instrument specific branch for configurations, but a separate repository for calibration files with all instruments on master. How do we handle local instrument changes e.g. locally created ramp files etc.

• do we just leave repo as "dirty"
• do we try and push these, but we may get conflicts
• do we put instruments onto their own calib branch and push, we then need to merge master rather than pull master on an update
• do we try and keep local changes in a separate area/naming scheme (e.g. local_ prefix) so we can gitignore them

Acceptance Criteria

• A meeting is held to decide on our process
• Tickets are created as appropriate

Time in Planning

00:49:49 02/05/24

[FreddieAkeroyd]

If we go for a local pushed instrument branch then we can track changes better and so potentially let other groups manage the area a bit themselves?

[rerpha]

the scientists already have a local calib directory which is their space really. perhaps we should make it more clear that if they want to use a custom file they need to use that directory?

[FreddieAkeroyd]

yes i think we need to make their local calib area clearer, then we need to discuss the scenario of a new temp sensor is needed in a hurry:

• do they drop it in the local calib directory, in which case we need to manage the fact a duplicate may appear in the real calib directory later, jenkins warning?
• do we provide a mechanism for SE team to update master and pull new version?

[KathrynBaker]

The main calib directory is kind of read only, the original idea was that the repo would be updated on instruments as part of a deploy, but new sensors could be added to the repo and pulled down as needed between times. The SSH script allows us to do these as a specific deploy. I think we were hoping to pass on the updating of the repo to SE at some point.
Local variations or calibrations should be in the directory that Jack referred to above in that world.