Thoughts on timestamps in CoincidenceEvent.yml and TOF

[fedor-goncharov]

This issue serves just to recall that one should (re-)design TimeBlock and CoincidenceEvent keeping in mind TOF.

If you add TOF "first" detector can detect a photon within block n and "second" may arrive block n+1.
It's up to the manufacturer to decide which of the 2 timeblocks they put it in, but practically it shouldn't matter.

At least special fields should be considered for TOF data (for memory reasons possibly it is worth not to put it completely
in the main datafile and have a separate protocol for TOF and other extra measurements) - assumes one wants reconstruct without TOF, then it is no reason to read GBs of memory of TOF information.