Chunking was removed from the loggers, since long term experiments are not the norm in ucl-open, and short experiments can be broken up into multiple files per experiment. However, on reflection and discussion, it has been pointed out that if we build the API and data readers so they are able to take care of this at analysis time (as in aeon_api e.g.), then chunking data by time allows us to record very short or very long experiments with no change to the workflows or nodes.
Chunking was removed from the loggers, since long term experiments are not the norm in ucl-open, and short experiments can be broken up into multiple files per experiment. However, on reflection and discussion, it has been pointed out that if we build the API and data readers so they are able to take care of this at analysis time (as in aeon_api e.g.), then chunking data by time allows us to record very short or very long experiments with no change to the workflows or nodes.