Minutes for LArSoft coordination meeting on May 22, 2018

Present

Jim Amundson, Vito Di Benedetto, Giuseppe Cerati, Lynn Garren, Chris Green | Herbert Greenlee, Tom Junk, Kyle Knoepfer, Erica Snider, Hans Wenzel

Remote

Katherine Lato, Paul Russo, Saba Sehrish

Project status report [Erica Snider]

  • Last week's release v06_76_00_01 (May 15)
  • Last week's release v06_77_00 (See release notes for changes)
  • This week's release will have several breaking changes:
  • ticket 17179: DetectorClocks moved to lardataalg
  • ticket 18137: Remove 2nd argument from geo::GeometryCore::WirePitch()
  • ticket 18136: Fix design faw in geo::ChannelMapAlg interface
  • ticket 19313: Input histogram fles to SingleGen:

Conclusions

  • Need experiment signoff before proceeding with v06_76_00_01, test release for art 2.11.02

Status of Spack Development / Migration [Chris Green]

  • Please see slides for detailed update
    • Replace UPS with an off-the-shelf build/packaging/deployment system based on Spack.
  • Produce an “MVP” – Minimum Viable Product: The purpose is to let experiments experiment with the software stack.
    • Currently external packages and art suite are built and all tests passing
    • No guarantee that everything will eventually work out
  • [Herbert Greenlee] Is ups going away entirely? Originally said that it would remain for user environment configuration
    • [Jim Amundson] Yes. There is no value added. But he took responsibility for those earlier comments.
  • [Tom Junk] : Equivalent to ups list, ups active, etc.
    • Acknowledged, and will plan on providing these tools.
  • More discussion
    • Spack was put together from the perspective of HPC admin and we have brought in a lot of new perspectives and provided feedback that have been received positively.
    • There was interest in Spack by multiple groups during early HSF packaging group discussions

Conclusions

  •  

Simulation of optical processes [Hans Wenzel]

  • optical Pyhsics very important in Geant4
  • Want to see photon propagation / visibility calculation moved into Geant4
  • N2 concentration in LAr is a strong determinant of absorption length. Published values are too short for our detectors. But those are for welding grade LAr. Our detectors have lower N2 spec, so have 30+ m absorption lengths
  • Rayleigh scattering is important. Changes arrival times and locations by a lot
  • [Tom Junk]: Spectrum of LAr scintillation line is 3 nm. This is significant given steepness of index of refraction at scintillation wavelength.
  • [Hans]: agree. The line is simulated as a Gaussian shape, he believes.

Conclusions