Extractions of the same data run under Cygwin and Linux differ in the number of data points reported. In one case, there are 39 additional rows of data at the end of the extraction. What is most likely happening is that the extraction is terminating before receive all the data from bfr, perhaps because bfr terminates before sending it. Either way, we are losing data.
This particular issue is not a very high priority simply because we usually don't care about the last few seconds or data. That said, 39 seconds is actually quite a bit, and we might well be curious about what was happening just at the end of the a run if something else was going wrong.
This issue of course relates to several other issues, so is something of a duplicate, but this is another aspect and the first clear evidence I have that this particular problem with bfr actually occurs.
One more example of effects of sloppy shutdowns:
Extractions of the same data run under Cygwin and Linux differ in the number of data points reported. In one case, there are 39 additional rows of data at the end of the extraction. What is most likely happening is that the extraction is terminating before receive all the data from bfr, perhaps because bfr terminates before sending it. Either way, we are losing data.
This particular issue is not a very high priority simply because we usually don't care about the last few seconds or data. That said, 39 seconds is actually quite a bit, and we might well be curious about what was happening just at the end of the a run if something else was going wrong.
This issue of course relates to several other issues, so is something of a duplicate, but this is another aspect and the first clear evidence I have that this particular problem with bfr actually occurs.