Seth R. Johnson
Seth R. Johnson
@lebuller Did you see the comments above?
Thanks @amandalund , this one's completely slipped my mind (sincere apologies). It's likely another case of overengineering. And either way, it's a subtle physics bug fix rather than an important...
I think I'm missing something too 😔 but I trust your judgment here and let's close this and the other issue. We can probably vastly simplify the applicability code too....
I think this is superseded by @2077 and in either way has diverged way too much since this was first pushed ☹️
Without GPU on my laptop (clang 15 debug), on VecGeom ca34e35ea5fc822fcffd3c7c5d117e6bc671ce32 (reldeb), the following tests fail: ``` 54 - geocel/vg/Vecgeom:SimpleCmsTest.* (Failed) 55 - geocel/vg/Vecgeom:FourLevelsGeantTest.* (Failed) 56 - geocel/vg/Vecgeom:SolidsGeantTest.* (SEGFAULT) 57...
@mrguilima I think we can merge this after #1652 and https://github.com/spack/spack/pull/49313 are merged! I'll pull those in and update the CI when those are in. The one outstanding issue is...
I don't think this will help https://github.com/celeritas-project/celeritas/issues/1351 ; we can't determine from the graph alone what pairs of nodes will connect.
@amandalund this is what I was talking about over coffee
Add detectors to the `inp::Model`: - `Detector` struct - DetectorId moves to geocel/Types - GeantGeoParams defines `G4SD* id_to_geant(DetectorId)` and `geant_to_id` functions (use code from `GeantSd::setup_volumes` to map) - Add tests...