Store or return observations in Trace.tokenize()?

[e-cal]

Right now, Trace.tokenize() uses the given Token (Observation class) to get observation tokens for each step, and appends them to a list attribute.

With TraceList.tokenize() we are going to return an ObservationList object, not store anything on an internal attribute.

So should Trace.tokenize() just return the list of observations instead of storing them? Or is there a good reason to store the list of observations on a Trace?

[haz]

Ya, I think it's safe to keep them separate for now. So return the ObservationList, and only let the acquisition techniques have access to that.