[e2e] Open the floodgate

[Cannara]

And, Bob, this summarizes the impenetrability we've seen over the years for
genuine TCP/IP improvements -- "...there was...", "...we did...", "..VJ
did...".  That's why we're where we are these many years later.  Everything
done that left TCP/IP as it is today was fine, so 'go away'.  What better
example than:  

"The High Priest of XCP came to one of our meetings to describe his    
religion.  The discussion was polite, but we were not convinced."?

Is the "we" the royal one?  Why the references: "High Priest", or "his
religion"?  Sounds like religion on both sides maintaining suboptimal network
design for the masses.  Which is, of course, what has happened.

Just as lack of engineered Internet security has allowed spam etcetera to
assume dominance in capacity and user-time consumption, lack of engineered
Internet protocols has left us with the kludge of TCP being responsible for
"congestion control" of the network layer, while being forcibly denied
important network-layer info -- use of control theory?  Even getting ECN ideas
moving has been more like pulling a NY garbage barge off Staten Island shoals
with a Jet Ski.

In the Internet (you know, that fantastic enabler of progress and modern life)
we have accepted a lack of professional responsibility for protocol design and
implementation, combined with a lack of version and source control.  Yet this
public utility has always depended on public funding.  Why is such an
important utility treated so casually?  Or, why do various committees still
junket all around the world to eat, drink, play and exhange papers that go
nowhere as far as real improvements to The Internet?  This is an especially
relevant question considering all the ongoing noises made, for instance, about
TCP's problems in real life.

Seems to me these are the questions that are continually avoided.  The
Internet, as it is, is an existence proof that came out of DARPA and followup
research.  That research did not generate a system for management and self
improvement.  Rather, it has encouraged a grad-student + advisor +
paper-publishing + powerless-committee form of bureaucracy.  We all experience
the intransigency of bureaucracies.  They often mature to being self involved.

So, nowadays, the lack of systems-management discipline in The Internet leaves
us with few options.  We can build networks and even protocols that go around
it, if we have the money, or we can lump it!  Hey, that sounds like how things
were when corporations and government had Netware, DECnet, Vines, etc. and
simply had a gateway (in the true sense) to The Internet.  :]

Alex

Bob Braden wrote:
> 
>   *>
>   *> but in the community that were the True Defenders of the One Faith of
>   *> TCP, it became short hand for "Bad Idea" - i think i may be being a bit
>   *> of a revisionist, but I dont think One True Fair, or Bad Idea type
>   *> discussions help us with moving along. there are pieces of XTP worth
>   *> understanding....
>   *>
> 
> Well, Jon, while we are talking about history...
> 
> I guess I have to plead guilty to having been a Defender (but probably
> not "True") of the TCP/IP faith, but the fact is that the XCP folks
> were on a jihad against TCP/IP, and it was not possible to hold
> meaningful discussions with them.  They, quite a lot more than for the
> Internetters, thought they worshiped the One True Faith.
> 
> It is my impression that some of the disappointed supporters of XCP
> came back later to boost ATM as the One True Faith, the One Ring to
> Rule Them All.
> 
>   *> and there are other protocols worth reading up on (TP4, delta-t, VMTP,
>   *> etc) which also have things to offer. and we need to stay abrest of the
> 
> BTW, there WAS an End-to-End (task force) at the time, which discussed
> all of these transport issues to some depth.  This was before the IETF
> got into gear, so the E2E TF looked for important new E2E (and a few
> not-quite-E2E, like IP multicast) technologies to bring into reality in
> the Internet.  The High Priest of XCP came to one of our meetings to
> describe his religion.  The discussion was polite, but we were not
> convinced.
> 
> The E2E WG also danced several dances with Dave Cheriton's VMTP.  Yes,
> it had a number of interesting ideas, but it too suffered from too many
> ideas, interwoven in a somewhat impenetrable manner.  There were also
> worries about VMTP's possible impact on TCP congestion control (and
> collapse).
> 
> Delta-T was of course seminal, and hopefully every CS graduate
> student has encountered it in studying transport protocols.
> 
>   *> new techniques (control theory for rate and window management,
>   *> information theory for transmssion on wireless) that mean we can do
>   *> things that were NOT done in the past, or if done, only understood
>   *> through a glass, darkly
> 
> VJ's work was based on control theory. In fact, Van came to the Internet
> community (and the E2E TF) from designing control systems for particle
> accelerators.  So control theory is not a new idea.
> 
> Bob Braden