11:00 Eastern; 08:00 Project/Pacific Time. bluejeans.com/426716450

Attendees

Apologies

Notes

Andrés Alejandro Plazas Malagón

  • Gave Chris comments on ... a ticket.
  • Working with Adam Snyder to run his CTI code at NCSA (DM-26352).
    • Able to reproduce his measurements with the files that he gave it.
    • There is a paper in prep by Snyder; Andrés has a copy, but cannot share.
    • Can reproduce plots.
    • Agreed that this ticket is effectively done.
  • Aim to get a concrete, prioritised list on DM-26350 of the next several pieces of functionality that need to come over.
  • Agreed to focus on prioritising and transferring code to NCSA and getting it running there, before working on converting anything to the stack framework.
  • Craig Lage has been experimenting with the PTC.
    • Robert is concerned about getting gains from the PTC; he has a proposal which will avoid the variability introduced by the different PTC models.
    • Will be discussing this with both Craig and Robert this afternoon.
    • We'll discuss this meeting next week, and make sure that any actions are turned into specific tickets.

Christopher Waters

  • DM-26316 last week, at Robert's request.
    • Discussion about conflicting config options; out of scope for this meeting.
  • Working on ci_cpp.
    • Last week working on tests, cleaning them up.
    • This week, implementing a non-SCons-based running method.
      • This is not as trivial as one might hope.
      • The issue seems to be getting the bulk of processing done before running the tests.
      • Agreed to not do this just for aesthetic reasons, but Chris doesn't think this is the issue for the time being.
  • Next working on linearity getting converted to isr-calib.
    • Most of the code is already done last week (DM-24703).

Robert Lupton

  • Looking forward to being able to process BOT data.
    • This means a Gen3 repo for BOT data at NCSA.
    • RHL will be pushing on this, trying to persuade somebody at NCSA to do it.
    • Might be ComCam data rather than BOT data; no reason to prefer one or the other for now.
  • DM-26414 filed the other day; asking Chris to take a look.
    • And he will!
  • Also DM-26394, but this is an Architecture problem.
  • Benchmarking data access at NCSA; 1s per image access to retrieve an image from the Butler.
    • Think this is a Butler problem rather than an NCSA problem.
    • Hoping this is resolved in G3; not yet clear.
    • Long term, this level of overhead is unacceptable.