Abandoned realtime update method over all new data, replaced with recursive method for a single timespan at a time, oldest to newest.

Refactored starttime/endtime to default to None, so algorithm and outputFactory can use starttime/endtime from input source.

Added update_realitime flag to init.  run_as_update is completed up to finding gaps, and backing up to find 'All' new data.  Still need to call the algorithm,  and write out the data that's 'new'

Added routines to return the number of seconds in a given interval, and to detect if 'new' data exists between two data sets

moved create_gap_stream from controller.py to TimeseriesFactory.py, made some sense,  and it gave me access to starttime and endtime

broke out update vs. normal timeseries, added functions to help analyze gaps.  Still need to complete update

Iaga sampling rate

Fixed odds and ends,  and made convert_stream_to_masked, and _convert_trace_to_int non destrutive calls

Refactored to reflect the change to write out an entire channel before closing socket. Also divided out buffer creation, from sending buffer.

Added static variables of FORCEOUT and TAG to keep -1/-2 number of sample special cases from getting mixed up,  like I did.

Removed cwbhost/cwbport logic from code, will let EdgeFactory deal with that. Rewrote the get_tag routine to use strings, which only work if the string is spliced to exactly 12 characters.  Rewrote the get_tag routine to remove the sequence code.  Rewrote the sequence code for data packets to use class variables. Other cleanup, especially comments.

Refactored code to put one channel at a time,  instead of one trace at a time, to fix a problem with writing to edge. Added logic to use cwbhost/cwbport if data is older then 10 days, and the cwbhost/cwbport are provided. Added starttime/endtime to put_timeseries to match algorithm base class. Other cleanup as discovered.