This MR is now ready to go. With it I have removed all use of detdesc objects when the stack is built for dd4hep, apart from a small set of calls that require volume intersections
DetDesc intersectionTicks method to get the entry/exit points with the volume associated to an object
DetDesc isInside method to check if a given point is inside a given volume.
As the reconstruction now fully uses the dd4hep geometry I am running into an issue due to the fact I only have MC samples generated with DetDesc, and as this geometry does not match the dd4hep one, the PID performance is highly degraded. Essentially, I cannot fully validate the reconstruction w.r.t. final PID performance without MC samples generated with a geometry that matches the dd4hep one. @seaso@gcorti any timeline on when samples matching the dd4hep geometry might become available ?
Note, due to some small machine-precision level changes, ref. log updates are expected.
