Status of reading the trajectory samples in cmsExp.¶
Used to re-use a member of the G4PrimaryParticle, to be able to expand tried to use a derived class.
It was failing after just 2 events. This was due to the way operator new was implemented in/for G4PrimaryParticle which required that the operator new and operator delete to be overloaded in the derived class.
Status on the safety error and zero step length issues when running with no Physics.¶
We used to have a double call to the Transportation process ; this leads to confused the system about either the safety or the step length.
The dummy physics now only have 'discrete process' rather than only 'transportation'. We can configure it by updating our dummyProcessMessenger.
So Philippe will retest and see if the infinite loop (step size set to 0) with dummy physics is gone.
Event display has been enabled in cmsExp.
Discuss the next step (What do you prototype first).¶
Marc propose that we focus of implementing a GPU integrating the equation of motion.
cmsExpMagneticField::GetFieldValue needs to be updated to avoid the string lookup.
Marc: what is the accuracy of the CMS magnetic map:
Soon: xy they have 900 points (for 9m) times * 3200.
This is the content of G4ClassicalRK4::DumbStepper.