question: should epoch-ids be enumerated instead of random stings of numbers? #307
Description
Right now, epoch ids for generic devices are ndi.ido random strings. Should they be numbered in order of creation? (Which may or may not reflect chronological order.)
Along with this idea, should we make it easier to set the epoch_id to a filename or some other feature without creating a class?
Advantages:
- It is cognitively easier to deal with sequential number than a set of random numbers (even if the numbers are sortable).
- Even easier if the epoch matches some name that the user is familiar with (like a file name, as long as the file name increments).
Potential issues:
- One would have to make sure that the epoch_id is unique within a session. If we allow people to specify a filename, and it isn't unique, then that's a problem. We'd have to check (probably in
ndi.file.navigator, where most of the epoch_ids get assigned.)