Column key "info" defaulted in load_from_file isn't included by default when calling snapshot

[mmore500]

This causes files serialized with Snapshot using default params to not be loadable, which is probably confusing to end users.

Option 1: change default taxon column key to "id" which is guaranteed to be present in alife standard data
Option 2: serialize taxa to "info" column by default, like

phylotrackpy/test/test_systematics.py

Line 97 in b70b3dd

sys.add_snapshot_fun(systematics.encode_taxon, "info")

Option 3: leave as-is and document

[emilydolson]

This behavior is the way it currently is because on the C++ side we can't count on ORG_INFO being easily serializable. In Python, though, serialization is a lot easier, so I'd lean towards Option 2. The big downside of that is that if the user has some weird ORG_INFO type that break encode_taxon they won't be able to output their data at all, which I don't love.

Maybe there's a way to make it configurable?