14th February 2014

• opsim review and scheduler development

• • External review - committee Roc Cutri, Phil Pinto, Eric Neilsen, Michael Droettboom 

• • Question was status of opsim and its suitability for developing the scheduler (docs in collection 3107)

• • Verbal debrief - code mature, needs to go from development to integrated in the development of the scheduler. Needs prioritized objectives, need to address failed observations (in a probabilistic manner rather than pass/fail), need to understand ICD with scheduler (including possible data from DM), increase the speed to hours not days for a run,  need to better understand the metrics in terms

• • Tasks ahead - rolling cadence for workshop in August, srini and francisco on vacation (3 weeks), will move to git. Require a detailed understanding of the ICD with the scheduler and to create a design for this
    
    
  • Transition to git - End of next week for completion of any cvs clean up

August meeting - LSST Project and Community Workshop: August 11 - 15, 2014

• transitioning from slalib 
  • simple transition to palpy (slight disagreement eg prenut uses different standards)
  • palpy doesnt accept nd-arrays - not vectorized. This costs about a factor of 3 for simple speed up 
  • speed not an issue currently for opsim and catalogs but may be an issue going forward
  • Monet - will ping groups about update schedule of sofa and comparison of accuracy of slalib and pypal routines
  • Scott will talk with Francisco
• brief update on software standardization study

• Chris Stubb's and Chuck's sky camera and photodiode 
  • confluence page 
• Debbie (meeting at noon for DESC)
  
  • series of phosim and catsim examples and uses and recipes (python scripts)
  • release date end of March (github repo is accessible currently)
  • Simon - Adrian Pope will visit in a couple of weeks (UW) and will create some recipes during his visit
    
    • Chris - the documented steps that go with the end-to-end walkthrough (movie) would be useful
• Richard 
  • NERSC (DESC) 4 million CPU hr allocation
  • requests currently in place for abut 2 million hrs
  • working on imsim on demand facility (specify how to run imsim and then launch on grid). Goal is June to have great it running
  • Chris - not really a centralized installation