history entries should have sequential IDs

[curusarn]

Desc

each history entry should be uniquely identifiable using: sessionID + seqID

Why

this will be useful e.g. when you delete a history record locally that is already synchronized to other devices. Deleting will clear the record and mark it as deleted - other replicas of the record will be cleared and marked during sync.

[curusarn]

sequential IDs have too many edge-cases; we are using random UUIDs instead