> Sure, one experiment would be what the actual performance of an IP over > MIME tunnel was. sounds good. why not specify a methodology so that the test is repeatable and the results from various experiments are comperable, and ask people to report their results to you, for you to summarize via a web page or other means?