From owner-ietf@ISI.EDU  Wed Jun 10 07:48:36 1992
Received: from mermaid.micro.umn.edu by mudhoney.micro.umn.edu (5.65c/Tony Tiger 1.34)
	id AA17734; Thu, 11 Jun 1992 00:31:43 -0500
Received: from venera.isi.edu by mermaid.micro.umn.edu (5.65c/Tony Tiger 1.34)
	id AA06361; Thu, 11 Jun 1992 00:31:41 -0500
Received: by venera.isi.edu (5.65c/5.65+local-6)
	id <AA11591>; Wed, 10 Jun 1992 14:50:32 -0700
Received: from alpha.Xerox.COM by venera.isi.edu (5.65c/5.65+local-6)
	id <AA11568>; Wed, 10 Jun 1992 14:50:10 -0700
Received: from skylark.parc.xerox.com ([13.2.116.7]) by alpha.xerox.com with SMTP id <12005>; Wed, 10 Jun 1992 14:48:48 PDT
Received: from localhost ([127.0.0.1]) by skylark.parc.xerox.com with SMTP id <22277>; Wed, 10 Jun 1992 14:48:43 -0700
To: ietf@ISI.EDU
Cc: dartnet@ISI.EDU, rem-conf@es.net, steig@pescadero.stanford.edu
Subject: seeking help with multicast topology for another IETF audiocast
Date: 	Wed, 10 Jun 1992 14:48:36 PDT
From: Steve Deering <deering@parc.xerox.com>
Message-Id: <92Jun10.144843pdt.22277@skylark.parc.xerox.com>

Steve Casner and I, with much help from the good folks at CNRI, MIT and
NEARnet, are organizing another Internet audiocast experiment (and, we hope,
videocast and imagecast) for the upcoming IETF meeting in Cambridge.  In
order to reach a larger number of participants than last time, we are looking
for volunteers to provide IP multicast forwarding machines and to help with
the (mostly manual, for now) configuration task.  We'd particularly like to
get help from some of the regional and backbone network providers, because
they are able to position the multicast relaying functions in the best
locations, relative to their customers and to high(er) capacity links.
It might also give them a chance to learn more about providing multicast
service, before their customers start clamoring for it. :-)

We'd also be delighted to get help from people other than network providers
who are:

	o  well-positioned, connectivity-wise (especially, in or to locations
	   outside the U.S.),

	o who are willing to use one or more of their own machines as
	  multicast forwarders (Sun sparcstation, DEC pmax, or Proteon
	  router, running multicast routing software) for the weeks up to
	  and including IETF week, and

	o who already have experience with the "mrouted" multicast routing
	  demon or the new Proteon MOSPF code, and have no qualms about
	  installing kernel patchs or new code releases on short notice.

(The people who helped us out with the San Diego audiocast all qualify, of
course!)

It should be an exciting experiment.  We are hoping to demonstrate compressed
audio (to reach less bandwidth-rich parts of the Internet), low-data-rate
video, real-time distribution of presentation slides, and new multicast
routing protocols.  But we need your help to make it happen.  If you can
help, please send me email; I will provide more details about required
hardware and software and about the operational aspects of configuring and
running the multicast Internet.  (I'll respond after June 22, when I get back
from travel.)

For those of you who are interested in participating in the audiocast as end
sites (i.e., non-forwarders), take a look in the vmtp-ip directory on host
gregorio.stanford.edu for software to modify your BSD, SunOS, or Ultrix kernel
to support IP multicast, and stay tuned to this list for information about
getting hooked into the multicast topology and about programs to receive and
generate audio/video/images.

Steve Deering



From owner-ietf@ISI.EDU  Wed Jun 10 18:13:09 1992
Received: from mermaid.micro.umn.edu by mudhoney.micro.umn.edu (5.65c/Tony Tiger 1.34)
	id AA18573; Thu, 11 Jun 1992 01:28:47 -0500
Received: from venera.isi.edu by mermaid.micro.umn.edu (5.65c/Tony Tiger 1.34)
	id AA06510; Thu, 11 Jun 1992 01:28:49 -0500
Received: by venera.isi.edu (5.65c/5.65+local-6)
	id <AA24108>; Wed, 10 Jun 1992 19:14:37 -0700
Received: from cs.umb.edu by venera.isi.edu (5.65c/5.65+local-6)
	id <AA24104>; Wed, 10 Jun 1992 19:14:34 -0700
Received: by cs.umb.edu (5.65c/1.34)
	id AA29279; Wed, 10 Jun 1992 22:13:10 -0400
Date: Wed, 10 Jun 1992 22:13:09 -0400 (EDT)
From: "Alan J. Zall" <azall@cs.umb.edu>
Subject: Unsubscribe!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
To: IESG Secretary <iesg-secretary@NRI.Reston.VA.US>
Cc: ietf@ISI.EDU, ietf-osi@cs.wisc.edu
In-Reply-To: <9206101629.aa05319@ietf.NRI.Reston.VA.US>
Message-Id: <Pine.2.4.55.9206102240.A29139@cs.umb.edu>




From owner-ietf@ISI.EDU  Thu Jun 11 02:30:50 1992
Received: from mermaid.micro.umn.edu by mudhoney.micro.umn.edu (5.65c/Tony Tiger 1.34)
	id AA19163; Thu, 11 Jun 1992 04:18:47 -0500
Received: from venera.isi.edu by mermaid.micro.umn.edu (5.65c/Tony Tiger 1.34)
	id AA07027; Thu, 11 Jun 1992 04:18:39 -0500
Received: by venera.isi.edu (5.65c/5.65+local-6)
	id <AA24650>; Wed, 10 Jun 1992 19:32:26 -0700
Received: from kona.CC.McGill.CA by venera.isi.edu (5.65c/5.65+local-6)
	id <AA24644>; Wed, 10 Jun 1992 19:32:21 -0700
Received: from expresso.CC.McGill.CA by kona.cc.mcgill.ca with SMTP (5.65a/IDA-1.4.2b/CC-Guru-2b)
        id AA28155  (mail destined for ietf@isi.edu) on Wed, 10 Jun 92 22:31:12 -0400
Received: by expresso.cc.mcgill.ca (NeXT-1.0 (From Sendmail 5.52)/NeXT-1.0)
	id AA16938; Wed, 10 Jun 92 22:30:51 EDT
Message-Id: <9206110230.AA16938@expresso.cc.mcgill.ca>
In-Reply-To: William Manning's message as of Jun 10,  8:15
From: peterd@expresso.cc.mcgill.ca (Peter Deutsch)
Date: Thu, 11 Jun 92 02:30:50 GMT-0:02
In-Reply-To: William Manning's message as of Jun 10,  8:15
X-Mailer: Mail User's Shell (6.5.6 6/30/89)
To: William Manning <bmanning@is.rice.edu>
Subject: Paying for the ride...
Cc: ietf@ISI.EDU


[ William Manning writes: ]
> 
> Peter Deutsch

[* paragraph deleted *]

> > I think we're getting to the point where we have
> > to look at pay as you go. 
> > 
> > This goes beyond getting access to paper copies of RFCs
> > and right at how we are going to fund Internet information
> > services. Right now, for example, anonymous FTP archives
> > are offered as a labour of love by some 1,000 sites. Can
> > we expect that to continue? I have my doubts, but I may be
> > proved wrong. If not, how are we going to pay for things
> > if not pay as you go?
> > 
> > 				- peterd
> > 
> This seems like a good morning for off-the-cuff remarks, SO...

Hey, this is the Internet. It's _always_ a good morning
for off-the-cuff remarks! :-)
.  .  .
> 2. Your second paragraph.  Why? or perhaps Why now? . .  .

Why now? A couple of reasons. For one, I happen to believe
that we've passed some magic point on the exponential
growth curve and the nature of the net is changing (yeh,
yeh, imminent death of the net predicted, film at 11..)

As we continue to double in size each year, it is my
belief that we are now seeing a new type of user come
on-line. The ranks of new-comers now include a growing
number of non-gurus, people who are not interested in
computer science, engineering or routing protocols. I
believe that what these people expect from the net, and
what they are willing to put into it, is different from
what we expected in the past.

In the process of accomodating these people we need some
mechanism for allocation scarce resources to satisfy their
apparently infinite demand. Since I happen to believe that
the free market is a pretty good way to allocate such
resources I'm not unhappy with the idea that we allow free
market forces to determine which Internet goods and
services get developed.

A separate argument would be that although we could
reasonably expect universities, research centres and other
volunteers to operate FTP archives, archies and other
services for free as part of their "mission" of promoting
research and teaching when most people on the net were at
other research and teaching institutions, I'm not sure it
is really reasonable to expect such people to operate
other such Internet services (for example databases of
commercial clip art and sound files, on-line libraries,
commercial software distribution and support services,
interactive tutorial and consultation services etc) for
commercial consumers.  Heck, the U.S.  doesn't even allow
these people to pass their packets on the NSFnet backbone!

Is there any reason to _not_ want such services? I don't
see any. Is it feasible to expect demand for such services
to be satisfied by volunteers? I personally don't think
so.

And why not simply allow volunteers to continue offering
such their goods and services? I'm not arguing that we
shouldn't. Far from it. It's just that it seems to me that
somewhere along the way up the growth curve the nature of
the environment changes. Where once people could offer
such services as a marginal cost increase on existing
machines with existing bodies, more and more we must look
to funding machines and people specifically to offer such
services to "the world". Where once we could hide such
costs, I believe we are approaching the point where such
costs must be brought out into the open and justified to
the bean-counters.

This may not be true for an FTP archive that contains
little original material, but there are surely _new_
services, requiring significant resources, that are only
going to be feasible when people are willing to invest
money in providing them?

As a simple example from personal experience, to offer an
archie server you really need to devote to the task a
machine with a lot of memory and a fair bit of disk. Yes, a
number of sites do this without charge, and I applaud them
for their generosity, but this does not provide any method
for covering the R&D costs of creating and improving the
service. I think we have to address this part of the
problem if we are going to see a lot of new archie-like
services.

Because Alan Emtage and I are pig-headed, and because
there are a number of good samaritans on the net, there
are now a fair number of archie servers running. But I
have to ask myself how many _other_ spiffy services we
would have by now if it wasn't so damned hard to get things
going. IMHO, a bit of money in the loop to encourage
experimentation and risk-taking would do wonders to expand
the range of full-time services...


> .  .  .Most folks either
> pay for a network connection outright, (eg phone bill, Internet membership,
> etc.) or have a subsidy. (eg NSFnet grants,  ANS/PSI/MCI/ATT internal nets).

Yes, but paying for a network connection is not the same as
paying for any specific Internet services.  Unfortunately,
there is currently no mechanism for flowing through some
of that money you pay <your-net-provider-here> to fund the
development of a NIC, or an archie, or anybody's library
card catalogue. I see tremendous potential in such
services, but I think we need to decide how we are going
to pay for them. I just don't think volunteers and a
barter economy are going to cut it when we get 20 million
machines on the net...

To me this is the heart of this problem. I can imagine a
whole range of spiffy new information services but given
the amount of grief I've had to go through to keep archie
alive, I personally am wary about striking out again and
trying to set up more such services unless it is clear to
me that they will be funded _somehow_. Certainly my
institution didn't want me doing it again unless I can
cover costs. To say nothing of the concern that was
expressed at what we did to the load on our U.S. link...

Now, I've been told by a couple of people that if we
simply release archie source to the world, archie would be
adopted and improved and kept going. This is probably
true, but then how would you have me pay my mortgage? I
could look around for a university that doesn't look quite
so closely at what I am doing and try it again, but it
seems to me that the barter economy model isn't much of a
mechanism for encouraging people to turn a good idea into
a widely used service. Remember, there's a big difference
between a nifty project and a service and it has a lot to
do with how much of your work week you can spend on it.

STANDARD DISCLAIMER: Yes, my partner and I have decided
that the way to make this work is to trust in the free
market, and yes, we've formed a company to try our ideas.
Obviously, you are hearing the opinions of a free
marketeer who will be selling software to the net. This
will probably colour the way you interpret my remarks.

Those who believe that the net can fund itself with
volunteers are free to continue to give away their
software and deploy their volunteer services. Personally,
there are times I'm willing to pay somebody else to make
it more reliable, and to have a phone number to call when
it doesn't work.


> The real problem (small) is to provide a way to get all those FAX machines
> on the net to take FTP downloads.  (Whats the NETFAX wg upto these days?)

Well, there are certainly still interesting technical
problems to be solved, but I think that the financial
considerations are just as important in the long run.
Boring, maybe, but important.

> 3. Labor of Love?  Not really. Greedy self interest. The faster I can
> get "our" standards into the hands of interested end users, the more
> accepted the standards become. This plays into the idea that we should
> "subsidize" education, to provide ourselves with qualified, talented
> pool of potential empolyees so that we can gain market share, or that
> we have a market predisposed to our standards.

This assumes your management will allow you to provide
goods and services to the net on their dime. Some will do
this and some are doing it without realizing it. I suspect
that not all of them would be willing allocate the true
cost of such services if we had to be spelled out to them
what they are paying to do it.


> Breath... Climb down off the soapbox...

Suggesting that we should pay for Internet goods and
services _does_ seem to have this affect on people. Rest
assured, I'm not out to attack the excellent work that has
been done over the years by volunteers in building the
net. As one of the guys who pushed archie on an
unsuspecting public I have tried to do my share in this
area. From my own experience I am just not convinced that
such a model is appropriate for offering all the products
that people will want now we are starting to get this
technology into a significant number of new hands.

The way I see it, the free market is a nice way to allow
end users to vote for the products they want, without
requiring everyone else to pay for it. There is nothing to
stop volunteers from working alongside commercial
providers, where appropriate. I just don't want to rely
entirely on this model to fund all the nify new things I
want to do, as I've experience its limitations for the
past two years.

Sorry for the length of this. I contemplated taking the
response to email but after some consideration I think this
topic is relevant to this list. Now we have a working net,
it's time to decide what we will be doing with it. Coming
up with funding mechanisms is part of that.

Now, where did I put that asbestos suit??? :-)


				- peterd

-- 


From owner-ietf@ISI.EDU  Thu Jun 11 01:27:53 1992
Received: from mermaid.micro.umn.edu by mudhoney.micro.umn.edu (5.65c/Tony Tiger 1.34)
	id AA20234; Thu, 11 Jun 1992 08:19:21 -0500
Received: from venera.isi.edu by mermaid.micro.umn.edu (5.65c/Tony Tiger 1.34)
	id AA08202; Thu, 11 Jun 1992 08:19:15 -0500
Received: by venera.isi.edu (5.65c/5.65+local-6)
	id <AA06287>; Thu, 11 Jun 1992 02:33:06 -0700
Received: from cerc.wvu.wvnet.edu (cathedral.cerc.wvu.wvnet.edu) by venera.isi.edu (5.65c/5.65+local-6)
	id <AA06243>; Thu, 11 Jun 1992 02:29:06 -0700
Received: by cerc.wvu.wvnet.edu (4.1/SMI-4.0:RAL-041790)
	id AA08672; Thu, 11 Jun 92 05:27:54 EDT
From: kannan@cerc.wvu.wvnet.edu (R. Kannan)
Message-Id: <9206110927.AA08672@cerc.wvu.wvnet.edu>
Subject: Re: seeking help with multicast topology for another IETF audiocast
To: deering@parc.xerox.com (Steve Deering)
Date: Thu, 11 Jun 92 5:27:53 EDT
Cc: ietf@ISI.EDU, dartnet@ISI.EDU, rem-conf@es.net,
        steig@pescadero.stanford.edu
In-Reply-To: <92Jun10.144843pdt.22277@skylark.parc.xerox.com>; from "Steve Deering" at Jun 10, 92 2:48 pm
X-Mailer: ELM [version 2.3 PL6]

CERC would be more than happy to be part of this.

We have the following 

	sparc
	dec5000
	HP 9000
	SGI

all valid internet sites.


-- 

--
thank you very much

R. Kannan,
Concurrent Engineering Research Center
Drawer 2000, West Virginia University,
Morgantown, WV 26505
Ph:(304)293-7536, FAX:(304)293-7541 
email:kannan@cerc.wvu.wvnet.edu


