F Engine communication about Fringe Boundaries

[AaronParsons]

Fengine → Xengine communication about fringe boundaries
Currently, the start of observation is determined by the mcount on which GPU integration
begins (set by hera_ctl.py ). There is no communication between SNAPs and Xengines
on where the fringe stopping boundary is. The easiest way to implement this might be to
ensure that the first fringe stopping boundary is at mcount = 0 on the SNAPs. Then the
GPU mcount can be rounded to the nearest %(2**19) number.

[AaronParsons]

See https://eoranalysis.slack.com/files/U49NK7G5C/F01ATJFL4G7/bda_todo_list.pdf

[AaronParsons]

Despite having created this issue, I am not sure what this is about. Probably need a conversation between @david-macmahon and @jkocz . Assigning this issue to both of them to move this forward.