Closed jonnew closed 8 years ago
This is likely fixed because of the completely refactored schyronization scheme used by the new shmemdf library. However, I'm leaving it open until the code has all the assert statements required to confirm the problem is solved.
In the position recordings, samples can be duplicated. For example, look at this screenshot showing the recorded position information at increasing indices of the JSON array holding position data:
index
7679
and7680
hold the same position information (dict value) and timestamp (dict key). This points to a synchronization issue somewhere in the data processing chain. The most likely candidate is the recorder itself, but the cause must be found because if its not a simple issue with the recorder, this issue points to a potentially serious design flaw in the sample synchronization enforcement across processes.