Task 6: Questionable test parameters

[miri64]

When testing for the last release I wondered why UDP over 6LoWPAN has less error margin (<5%) than ICMPv6 over 6LoWPAN (<10%). Especially since Task 6 asks for way harder to deliver 1kB, while some of Task 5's specs only ask for 0B. Is there a specific reason for these 5%?

The other questionable thing more from a timing constraint is: Why does the delay need to be 1s? For testing it is really a long amount of time to send 1000 packets (~17min), while the tests would be as significant if the interval between them would be 50ms or 100ms.

[jia200x]

The other questionable thing more from a timing constraint is: Why does the delay need to be 1s? For testing it is really a long amount of time to send 1000 packets (~17min), while the tests would be as significant if the interval between them would be 50ms or 100ms.

+1

[jia200x]

Also, isn't 5% too much for packet losses?

I'm losing 1-2 packets out of 1000 with this test (0.2%). Maybe we can lower this number? (e.g 1-2.5%)

[miri64]

Well, it's a lossy medium with fragmented packets, so for the larger payload, I do see 5% as realistic.