When the recording to disk is started and stopped multiple times it is difficult to then find the boundaries between contiguous segments of frames after the fact. Therefore I think it would be much better if the program created a new SER file each time recording is resumed and finishes off the file each time recording is paused. The filenames could have a number suffix like [filename]_000.ser that increments for each segment.
Or I suppose I could write a tool that separates a single SER file into multiple files, using the frame timestamps to determine where the boundaries between segments are.
After pondering on this for a full two years, I no longer think creating multiple SER files is the right approach. I can write tools to deal with this after the fact and avoid adding complexity to this program.
When the recording to disk is started and stopped multiple times it is difficult to then find the boundaries between contiguous segments of frames after the fact. Therefore I think it would be much better if the program created a new SER file each time recording is resumed and finishes off the file each time recording is paused. The filenames could have a number suffix like [filename]_000.ser that increments for each segment.
Or I suppose I could write a tool that separates a single SER file into multiple files, using the frame timestamps to determine where the boundaries between segments are.