BlueJeans: https://bluejeans.com/426716450

attending:

Topics for discussion:

  • Algorithms Workshop closeout (Eric Bellm ):
    • Thanks to everyone for your hard work! 
      • Presentations were a success
      • We got some some very useful tools prototyped
      • The implied performance on HiTS is... not bad?
      • There are some obvious places to look for improvement–will ticket in sprint planning
    • Audience questions
      • evaluating photometric accuracy (we have requirements here)
      • Features we can use beyond top-level "# of detections" to assess progress.
    • Ian on DCR:
      • thought presentation went well, was pleased to get the DECAm sub-band RGB images together, people reacted positively
      • not a lot of detailed feedback in the workshop–lots of DCR-interested folks weren't on the meeting.  Gordon Richards heard some, Ian has been following up
  • Gen 3 conversion (

    T Key Summary Assignee Reporter P Status Resolution Created Updated Due
    Refresh

    ):
    • first version of Image Difference Task in Gen 3 is merged
    • Decam Gen2→3 repository conversion worked, but we can't run Gen3 processCcd on it
    • bias and flats work but we're having to comment out most of the rest of the steps, which are HSC-specific even in Gen 3
    • decision is for Kryzsztof to close the conversion ticket, file new tickets for all stages of processCcd that don't work for DECam, and we'll discuss with the middleware folks Thursday.  John S. and Tim J will address any ticket duplication resulting
  • Precursor processing (Meredith Rawls ):
    • as part of starlink, made DECam calibs from raws (with some hacks) and run processCCD (all in Gen 2)–will write it up
    • some discussion of whether to put effort into Gen2 CP products–Eric, Meredith, & John to talk about in sprint planning
  • report from NCSA (Unknown User (emorganson) et al):
    • not much due to DES stuff; moving back onto live processing with condor
    • JDS points out that Spencer Nelson is starting at UW to work on alerts–we will coordinate with Eric to talk about more useful tests
  • ap_association/SDM/fakes (Unknown User (cmorrison)):
    • ongoing functor works
    • will talk to Eric tomorrow about production fake injection
  • DCR (Ian Sullivan ):
    • nothing beyond Alg. Workshop; will be ~50% time due to child care this week
  • Image differencing (Unknown User (gkovacs) ):
    • next is decorrelation afterburner cleanup
  • Review CI (https://chronograf-demo.lsst.codes/):
    • nothing major
  • AOB
    • none