You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
The idea is that you would register a schema with a startSampler call. This call would return an ID.
Then you could ask for samples with getSample. An alternative could be getRealTimeSample which would delay the next sample until enough time has passed that a specific field has a time in the past. You would pass in the ID you got from startSampler. Each sampler should have a buffer so multiple clients can request samples in parallel and each would wait for the next sample in turn (for real-time samples). In any case, thread safety constraints should be observed ... some samplers might be run in many threads.
Eventually, you would call stopStream to free up resources.
An alternative API would have one call startStream where you would give a schema and any realtime constraint and results would be streamed back. This is safer in terms of deallocating resources automatically but slightly more complex.
The text was updated successfully, but these errors were encountered:
The idea is that you would register a schema with a
startSampler
call. This call would return an ID.Then you could ask for samples with
getSample
. An alternative could begetRealTimeSample
which would delay the next sample until enough time has passed that a specific field has a time in the past. You would pass in the ID you got fromstartSampler
. Each sampler should have a buffer so multiple clients can request samples in parallel and each would wait for the next sample in turn (for real-time samples). In any case, thread safety constraints should be observed ... some samplers might be run in many threads.Eventually, you would call
stopStream
to free up resources.An alternative API would have one call
startStream
where you would give a schema and any realtime constraint and results would be streamed back. This is safer in terms of deallocating resources automatically but slightly more complex.The text was updated successfully, but these errors were encountered: