default save history for Database revision

[pavlis]

This is a minor issue for the new Database implementation, but worth preserving for the record. As you know I've been cleaning up duplicate code in the Database class. In the process as noted elsewhere I've been pushing to deprecate the functions that explicitly tag ensembles and simply reads and writes to two core method of Database: save_data and read_data.

With that background a simple question comes up. The current version of Database will always save any object-level history stored in a TimeSeries or Seismogram within the ProcessingHistory container. The user has no choice. If the container has data it will be saved. I made a choice to add a new boolean argument to save_data with the obvious name save_history. The function is backward compatible with the current version of save_data when I set the default to True. Several tests failed when I initially set the default False. From the above that would be expected, but I didn't remember that feature of the current implementation and the errors were baffling for a while.

So there are two points I want to preserve here for the record:

1. Do you, @wangyinz , concur that adding this save_history options makes sense?
2. If so, should the default remain True? Since we advocate history as something to use only for preserving a final workflow to promote reproducible science, maybe the default should be False? If so the several tests that broke with that selection will need some repair.

[wangyinz]

yes, we should provide that option, which will mirror the one in read_data:

mspass/python/mspasspy/db/database.py

Line 337 in 23898f4

load_history=False,

I think keeping the default as true is probably better. It wouldn't hurt to save the history even if users don't want it, but it will hurt in the opposite.