Go to the Indico Home Page Oldest event Older event Up to category Newer event Newest event Printable version
  • Compact style
  • Indico style
  • Indico style - inline minutes
  • Indico style - numbered
  • Indico style - numbered + minutes
  • Indico Weeks View
    • Switch to the management area of this event
    Login

    Developer's meeting

    → Europe/Zurich
    Hide

    Compression (Milo):

    • see slides attached to the agenda
    • Milo described the (C++) code he has been working on to emulate (an approximation of) the WIB-to-Felix data format (see DocDB-1701), and the compression results obtained using this preliminary format
      • not all data format details clear (but should not have a major effect)
      • not attempting a wire-by-wire compression for now
      • for now, noise simulated as uniform but with varying number of bits, and study compression as a function of the number of bits
        • compression tested how just running gzip on files; will also study doing this within the code (as well as studying other algorithms)
      • compression results: as high as 30 if the data have no noise at all
    • Milo will share the code on github until we have a more permanent place
    • Kevin will attempt to find out whether there is some more information about noise levels and plausible channel-to-channel pedestal variations (e.g. from microBooNE)
      • or alternatively provide us with some actual waveforms from microBooNE

    Planning (Frank):

    • briefly went through topics of the planning file (attached to the agenda), which is now more detailed (especially BoardReader development)
    • details on monitoring still lacking (but not the most urgent thing)
      • Roland: this ought to be fairly straightforward, as a plan exists. Can show something in a future discussion
    • Enrico: volunteer to work on the circular buffer (have something very similar available already)
      • offer accepted gratefully

    Other:

    • Roland and Enrico have started with a software repository at CERN, and will see what can be done to get a gitlab repository available
    • they have also managed to get the artdaq demo code to run (under CCentOS7?) and have been studying the code
      • "it looks fairly straightforward"

    We will meet next at CERN, in person, during the DUNE CM. However, we should certainly start making progress before that (e-mail communication should go some way)

    • Frank will make a specific proposal for discussion of functionality
    There are minutes attached to this event. Show them.
      • 17:00 → 17:15
        Introductions 15m
      • 17:15 → 17:35
        Compression 20m
        Speaker: Mr Milo Vermeulen (Nikhef National institute for subatomic physics (NL))
        FrameGen.pdf
      • 17:35 → 18:00
        Planning 25m
        FELIX-20170111.pdf