After looking into the script I don't see an issue in the calculation. The numbers assume that we have 3 Publishes with a 2000 ms difference. The numbers do already put into account, that we are looking at 3 writes as 3*2000 = 6000 ms, while the lower limit 5700 provides 300 ms grace and the upper limit 7400 allows 1300 ms grace.

If we would divide the mdifference by NUMWRITES, we would need to do the same for the upper and lower limit what again would mean the provided publishing rate of 7100 ms is out of range.

Hi,

do you think that users actually understand "The result (see attachement) shows, that the average time of 7100 is in range but the received value is 21300."

where 21300 actually means 3 *7100?

Do users really need to study each error text in source code in order to understand what is going wrong?

We can clarify the first statement to say:
"Expected total duration over 3 publishing intervale of 6000 ms (average publishing interval 2000) but actual total was 21300 (publish interval average was 7100)"
second statement is just a log of the numbers that were being checked

this is just a clarification (not considered a change)

That would be sufficient.

Thanks for taking it in concideratation.

The error message has been updated to a more accurate text that should clearly state the problem.
In addition, the test script also aborts if the revised PublishingInterval does not match the requested one.

reviewed code in call , agreed to updates and closed issue