[e2e] Is a non-TCP solution dead?

[Ping Pan]

David,

Very interesting story. Actually, I have been wondering something similar
lately. (Sorry about out of topic.) For years, we have been building bigger
and bigger IP backbones with all sort of goodies. The latest one is MPLS.
But when we look at the deployment speed, the installation base, and the
people who have the money to build networks, I often wonder: why do they
want all these goodies in the first place? when will they be able to fully
understand the usefulness of these goodies beyond the marketing hype?

Maybe we all need some reality checks once in a while. :-)

2 cents,

- Ping



> -----Original Message-----
> From: end2end-interest-admin at postel.org 
> [mailto:end2end-interest-admin at postel.org] On Behalf Of David P. Reed
> Sent: Thursday, April 24, 2003 6:16 PM
> To: cannara at attglobal.net; end2end-interest at postel.org
> Subject: Re: [e2e] Is a non-TCP solution dead?
> 
> 
> Apropos of not very much ... the following actually happened 
> in 1992.   I 
> was talking to a NYNEX marketing/planning exec, arguing that 
> they ought to 
> roll out ISDN more aggressively here in the Boston area 
> because so many 
> hightech and university people were using remote terminals, 
> and that they 
> would be able to test the market.   He said that their marketing data 
> didn't show any interest or use of data at all in residences, and 
> challenged me to prove that there was a ready market.    I 
> asked how he 
> could tell, and he said that they knew exactly how many data 
> lines there 
> were in homes in the Boston area.  I said, oh wow, how do you 
> know when 
> people are using modems on residential phone lines?   Do your 
> switches 
> recognize modem signals?   His response (classic!) was "what 
> do you mean? 
> It's illegal to send data on residential lines." (implying 
> that you had to 
> have a special line to send data)   (I guess he missed the 
> Carterphone 
> decision).
> 
> Lots of (perhaps most) multimedia streaming ends up flowing on TCP 
> connections to get past NAT boxes...  but still, the real 
> question is how 
> much of that is there, compared with non-streaming media 
> (like Napster).
> 
> At 02:10 PM 4/24/2003 -0700, Cannara wrote:
> >John, I don't know enough about Inet2 to argue, but what you 
> say makes 
> >sense from what my friends at Stanford, who manage parts of 
> the SU net, 
> >have to say.  I believe, as the other emails have pointed out, that 
> >unless one actually looks at pkt contents, one can't really get good 
> >stats, due to the mimicking of TCP to get through filters.  This is 
> >likely why tools used by CAIDA, Sprint, etc. would have to 
> be examined 
> >to see what they're actually looking at, if anything, other 
> than simple 
> >port #s.
> >
> >Alex
> >
> >John Kristoff wrote:
> > >
> > > On Thu, 24 Apr 2003 10:33:37 -0700
> > > Cannara <cannara at attglobal.net> wrote:
> > >
> > > > John, the various non-TCP flows that have been 
> increasing, and are 
> > > > planned to be used even more in the future, center on media 
> > > > communications, such as audio& video streaming, as well as IP 
> > > > phone, conferencing, etc.  I'm not sure that Inet2 is 
> > > > representative for what's going on in the bulk of the net.
> > >
> > > I've heard that before and I'm not sure I can buy it 
> without seeing 
> > > some additional data showing that to be the case.
> > >
> > > My intuition tells me that I2 is probably more representative of 
> > > non-TCP than one might first imagine.  There are plenty of audio, 
> > > video, conferencing and VoIP projects taking place over 
> I2, perhaps 
> > > in many cases more so than other parts of the net.  The 
> reason I'm 
> > > lead to believe that it would be fairly representative of non-TCP 
> > > protocols is because generally the I2 community networks are 
> > > relatively open.  Its users are probably much more likely 
> to be able 
> > > to use non-TCP protocols. Whereas in many other places, 
> protocols, 
> > > ports and applications are often heavily filtered.  In 
> fact, I might 
> > > be tempted to think that many of those non-TCP 
> applications you're 
> > > referring to may often be running over TCP port 80 to get past 
> > > filters.
> > >
> > > John
>