Present: Vito Di Benedetto, Lynn Garren, Krzysztof Genser, Herbert Greenlee, Thomas Junk, Gianluca Petrillo, Erica Snider, Hans Wenzel
Remote: David Adams, Christoph Alt, Robert Hatcher, Katherine Lato, Saba Sehrish, William Seligman, Gleb Sinev, Dorota Stefan, Robert Sulej, Tingjun Yang
Project Status Report [Erica Snider]
- breaking change:
SingleGen
module has configuration validation
- art 2.9.2 will be on next week (not this release)
- proposing to move to
VIGILANT
diagnostic level (catching more "questionable practices")
MemoryPeakReporter
service removed because obsolete
- [Lynn Garren]
BackTracker
factorisation may be ready now
- [Erica Snider] it will be integrated next week, after the holiday
Conclusions
- we seek input on feature request for making run and subrun data products
- also seek input about diagnostic level increase
LArG4 Refactoring and other changes [William Seligman]
- project lead by Hans Wenzel and William Seligman
- partition of active volume in "voxels" was driven 10 years ago by concerns that turned out not to apply in the end
- voxelisation is being replaced by step limiting (a Geant4 feature)
- parallel world also going to disappear, as their main motivation was to host voxelisation
- when moving to physics lists from Geant4, scintillation and ionisation will be correlated
- ionisation and scintillation calculation can't be factorised with the current version of NEST, which depends on Geant4
- this refers to NEST code in LArSoft; there may be some newer version to be checked in "the outside world"
- considering the introduction of a new data product
sim::SimEnergyDeposit
to store energy deposits (has some drawbacks)
- [Erica Snider] does the photon library generation require a different physics list in the new model? [Hans Wenzel] No, just the enabling or disabling of Geant4 "stacking" feature
- [Erica Snider] changes to
sim::SimChannels
are not on the table now, but they might be considered as a future step
- [A] aiming for this work to be usable by the next MicroBooNE Monte Carlo production; such a redesign step should be informed by deep understanding of the needs and about how data is used now
- [Erica Snider] anything here precluding the implementation of simulation of induction on multiple channels? [A] not necessarily; that effect is simulated after
LArG4
(in the SimWire
class of modules)
- [Christoph Alt] is electron drifting going to be adaptable to dual phase simulation? [A] the target here is to deliver the same functionality as currently present [Erica Snider] are we nevertheless thinking in these terms of abstraction? [A] it will still be easier to implement this changing the small module than from the middle of
LArG4
like now [Erica Snider] pixel readout is another use case where this abstraction will be useful
- [Gianluca Petrillo] can physics lists be customised? [Hans Wenzel] physics processes can be freely added, or your own physics list can be plugged in
Conclusions
- we should survey whether a newer version of NEST exists that has loose or no dependency from Geant4
- further discussion will happen on some details of the choices to be made