Bo knows networking

*Bo knows networking; cooperation and curiosity, not financial gain, motivate the InteropNet's creators. (Bo Pitsker, Worldwide Networking manager at ZD Expos)
PC Week May 2, 1994 v11 n17 pN17(3)
PC Week May 2, 1994 v11 n17 pN17(3)

Bo knows networking; cooperation and curiosity, not financial
gain, motivate the InteropNet's creators. (Bo Pitsker,
Worldwide Networking manager at ZD Expos)

by Blakeley, Michael

Abstract
Exhibitors participating in ZD Expos' Interop trade shows are required to
connect to the show's network, InteropNet. Bo Pitsker, manager for
Worldwide Networking at ZD Expos uses the cooperative tradition of the
Internet to develop relationships with contributors and volunteer
technicians to successfully install the InteropNet during trade shows.
The volunteer system lets users exchange their labor to help install the
network for real-world LAN installation knowledge. Technology has
evolved since the first Interop in 1987. The early focus was on TCP/IP,
Unix and proprietary connections. Ethernet was the primary backbone
until FDDI was proven reliable. Open standards, interoperability and
customer service replaced the proprietary standards of 1987.

Full Text
Bo Pitsker, manager for Worldwide Networking at ZD Expos in Foster
City, Calif., is something of a networking Renaissance man -- he's
designed and assembled the Interop LAN (called InteropNet) for the last
five years, trained technicians, been head troubleshooter for new
technologies, and implemented those technologies in a mixed-platform
environment for use at the trade show.

Netweek talked with Pitsker after his team assembled the
infrastructure of, or hot-staged, the 1994 InteropNet in a warehouse in
Sunnyvale, Calif., a few weeks ago.

NETWEEK: How did the InteropNet become part of the Interop show?
PITSKER: The beginning of Interop was an accident. Dan Lynch, when he
founded Interop Co. in 1985, basically wanted to go on the lecture
circuit, and he did, successfully. Dan was one of the original authors of
the Internet. He basically led the programming team that converted the
old NCP protocols from ARPAnet into TCP/IP.

That made him justifiably famous, and he wanted to go out and preach
the gospel according to Saint TCP/IP. So he left Stanford Research
Institute and went out on his own -- mortgaged the house, you know, the
standard Silicon Valley story.

After some time, a number of individuals said, "Listen, Dan, this is
really great stuff, and we sort of hear that manufacturers are doing this
stuff, but frankly, the idea of interoperability is not proven yet. IBM
wants to sell you SNA, DEC has DECnet, Wang has VS, and so on, but none
of these things necessarily play, and we don't know that TCP/IP really
works. What are you going to do to show us that it works, instead of
giving us lectures?"

Dan took this as a challenge, and sent out an invitation to 300-plus
manufacturers. It said, "Why don't you come to the Santa Clara
Convention Center, and throw your stuff onto the show floor, and we'll
run a piece of thicknet down the middle. You connect up, and we'll see
how many people can talk to each other." Fifty-two manufacturers took
him up on the challenge.

NETWEEK: So in the beginning, it was like a bake-off? PITSKER: Yes. They
did that in '87, and the manufacturers found lots of problems, but they
also had a lot of success.

Today we expect, maybe unjustifiably, interoperability and standards to
just work. Before, the expectation levels were quite low, but it was
demonstrated that, fundamentally, TCP/IP on multiple platforms did in
fact work. It was a major success.

The manufacturers went back to work to improve their problems. Then
they said, "Well, Dan, we'd like to do this again, but we'd also like to put
up a few stands and put our literature on the table, and promote
ourselves. Do you have a problem with that?" Dan said no, and managed
to back into the trade show business.

NETWEEK: So the InteropNet came before the Interop show? PITSKER:
That's right, and that's why our orientation is different than other
shows that "have networks." Now, of course, it's not the only reason for
the show, but it's still a significant factor. Which is why, from the days
of the bake-off up to the present, connection to the network is
mandatory.

As the show gets larger, this becomes more and more challenging,
because we're drawing in manufacturers and suppliers who are
somewhat peripheral to the core network business. When you think of
the early days, it was primarily TCP/IP, and there was a preponderance
of Unix-based solutions.

That's a historical artifact. There's absolutely no connection between
Unix and TCP/IP. It's simply an accident. In fact, there's no connection
between TCP/IP and Ethernet, and our show is heavily Ethernet-oriented.
As time goes on, they become less and less predominant in the show
itself.

It turns out that having this embedded in the show takes us from merely
demonstrating technology to shaping its direction. We are putting up an
environment this large, and demonstrating it in public.

It's sort of like going to a Paris fashion show -- you see all the
extremes. You can see all the extremes of networking here, and make
your own judgments about which are successful and which are still
struggling.

Take FDDI. We starting playing [with it] in 1990, and have successively
improved the performance record each year, and now it is the primary
backbone. I don't even use Ethernet as a secondary backbone. Three years
ago, I wouldn't have dared to do that.

I joined Interop in 1991. Prior to that I was managing a microcomputer
reseller. I had done that for a number of years, and had a feel for the
retail market, especially NetWare and departmental LANs.

Coming to Interop, it was a bit of a shock to find out that people thought
departmental LANs were small potatoes. Global computing, global
networking was the core.

I also spent a couple of years in an IS organization, at Pacific Bell,
trying to integrate microcomputers and minis. At that time, at the end
of the '80s, we were very dependent on such things as DECnet. Trying to
get DECnet to work on PCs and Macs was quite challenging. Today, it's
more humdrum.

NETWEEK: Did companies jump at the chance to participate in early
Interop shows? PITSKER: They did because the early adopters of this
technology saw that the old, monolithic computing structures were
breaking apart. They had to get out in public and demonstrate that they
could fill the cracks in between.

In fact, with the standards ever-evolving today, it may be a foregone
conclusion that 10BaseT and TCP/IP work, it's expected, but there's
always that new technology lurking in the corner.

The open standardization process itself has become a norm, which was
not the case in the '80s. The expectation then was that anyone who
promised interoperability went at it with "deep magic." You could buy, in
the '80s, products that would connect DEC machines to IBM machines,
but that just meant adding another proprietary solution to the mix.

Now, the customer expectation is that if you don't stand up and salute
the open systems concept, and standards that can be looked up
somewhere, you're doing something wrong. It's a fundamental shift in
the balance of power -- the consumer has it now, and didn't before.

NETWEEK: How did you build up the volunteer system and the Network
Operations Center team? PITSKER: The finely tuned program that we
have for volunteers and contributors -- and frequently they are one and
the same -- comes out of the tradition in the Internet of cooperative, as
opposed to strictly competitive, relationships.

The Internet isn't owned by anybody. It is, in effect, an agreement by a
large number of participants that they will aid and assist each other for
mutual benefit. Our early volunteers, almost exclusively, came out of
the Internet.

We have taken that model and applied it to other areas. Our volunteer
program now reaches out to universities, the government -- federal,
state, military, municipalities -- and to end-user communities.

The key idea here is that volunteers are trading labor for knowledge.
They are gaining knowledge in an environment that's about as real-world
as they can get.

They could pay a lot of money and go off to a three-day class in some
aspect of networking, listen to a lot of boring lectures, and maybe take
a machine apart and put it back together.

Here, they get to play in an extremely large and complex network. They
can work on stuff that they aren't necessarily good at, but are
interested in.

It's cooperative, too. They get to work with vendors and other
volunteers, exchanging views on the best way to do things. So, they're
getting real benefit out of it, and the success of the program
demonstrates that.

We're getting something out of it, too. There is something very appealing
about disinterested cooperation. We've tried to maintain that spirit,
even as the show grows larger, and I think we've been very successful at
it.

Type: Interview

Company: Ziff-Davis Exposition and Conference Co.

Topic: Interview
Network Architecture

Record# 15 350 740