TI timestamp mismatch tolerance

[baltzell]

Sergey suggests setting it to 5 tics. Currently it's 1(?).

[whit2333]

What is the value of the RUN::config timestamp expected to be?

[baltzell]

I think it's a "global" TI timestamp, which is maybe only well-defined when all TI timestamps are equal. I think it's only used for (TDC??) jitter corrections.

[baltzell]

Partially closed by #482, but later realized ...

We must add an additional requirement that the i0 reference (which is what goes in RUN::config.timestamp if "all" ROCs have "equal" TI timestamps) is not from ATOF (same as is done currently for the tiMaster, which is defined as the one that has the "HEAD" bank 0xe10f/57615)

Then, #482 would work as intended for all data sets until we choose a better approach.