[e2e] Re: crippled Internet

Fred Baker fred at cisco.com
Wed Apr 25 14:52:53 PDT 2001

At 09:49 AM 4/25/2001 -0700, Randy Bush wrote:
> > When users measure Internet behavior, they say that delay is highly
> > variable and loss is excessive.
>to provide a solid baseline for this discussion and so we don't repeat the
>snafu we had on this one last time around, could you reference some actual
>scientific quality measurements?

I can only point you to the ones made by folks like the internet weather 
report folks (various firms), whom neither of us would call "scientific". I 
could take a bunch of traceroutes and perform a bunch of file transfers to 
measure delay between various points that I notice, and post the data file. 
I don't know what it would prove, specifically, about where the delay and 
loss would be, as they would not be scientific (which is to say "an 
experiment run in both a test environment and on a controlled environment, 
to prove or disprove a hypothesis").

However, I will suggest that you take a walk through town and tell people 
that your favorite ISP implements copious bandwidth throughout and to other 
major service providers - something which, for the record and the benefit 
of the audience listening, happens to be true. Then go on to say that you 
therefore believe that the internet imposes no delay other than propagation 
delay, no measurable jitter, and no measurable loss.

As you have been known to tell me, your marketplace begs to differ.

If you're not willing to make that assertion, then we have something to 
talk about.

More information about the end2end-interest mailing list