> I'm having a little trouble with this logic. You made two > speculations, > with no evidence, and now you wish to shift the burden of proof for > those speculations to someone else? It depends what you consider the hypothesis to be. The one I thought we were considering was the statements made by the measurement sites. I was pointing out that they can be contaminated by a number of sources. That is why I said that "There may also have been measurement problems..." If I had intended to claim that the effects hypothesised definitively happened I would have said "There were also". Try reading before you flame. The measurements from the measurement sites plus your own observations show a discrepancy. The measurement sites I saw had something like 7 sites offline. You say you have data showing responses from 10 of the servers. Thus there is a discrepancy of 4 servers to be accounted for and an indication that the effect was dependent on the observation point. The precise cause of the result is irrelevant, the only conclusions we can draw are that the measurements from the sites were not sound and that we should build measurement systems that are not subject to the possible sources of contamination described. Phill
Attachment:
smime.p7s
Description: application/pkcs7-signature