Closed csullivan closed 8 years ago
I am not sure if this surfaces elsewhere, but it is an issue in TMultiRawFile::GetEvent(TRawEvent& outevent)
, when output.next_event
is copied into outevent
[Link]. This then prevents the TBuildingLoop from correctly time correlating the events since outevent.fTimestamp
is uninitialized.
Edit: In reviewing a bit more, I realized that for GEB and NSCL data the timestamp (if uninitialized) is read from the respective child classes of TRawEvent. Thus it's not quite a bug. However, it does require the user to write a child class of TRawEvent to handle the request when fTimestamp is uninitialized (-1).
This showed up as a bug for me, where I have added a new data source but do not need a unique event type that inherits from TRawEvent.
thanks Chris, Do to arcane methods (see Eric) I don't think this really effects anything already there but it is always good to be consistent.
Fixed bug in TRawEvent &TRawEvent::operator=(const TRawEvent &rhs) in which not all members of TRawEvent were being copied (namely fTimestamp).