[e2e] Lost Layer?

Detlef Bosau detlef.bosau at web.de
Sat Jan 18 07:08:38 PST 2014


Am 14.01.2014 16:29, schrieb Hamed Haddadi:
> you may also like to have a look at this book chapter that covers some of
> the trade-offs very well with respect to multihoming and IPv6
>


Oh yeah. I don't know the situation abroad.

But besides Jews, Illuminati and Freemasons, Germany's top conspiracy
theories are 9/11 and IPv6. This even tops the landing in the desert in
Nevada, which was sold as the landing on the moon..... ;-)

Hamed, I want to talk about flow control and not about addressing. And
my claim is, that - please wait for the second half of the sentence
before putting me in your killfile - congestion control is questionable
because (second half of the sentence) congestion control works around a
missing flow control.

And precisely this flow control layer is really missing in TCP.

Perhaps, I make it into anybody's kill file here, however I considered
this claim quite some years and quite carefully, and I would appreciate
discussing it.

And no, I don't want to discuss addressing issues in this context,
because once a connection is established, the flow's endpoints are
associated and the flow is in a state which is far beyond any addressing
issues.

Detlef



More information about the end2end-interest mailing list