[ih] OSI and alternate reality

[Clem Cole]

On Fri, Mar 15, 2024 at 9:06 AM Andrew G. Malis via Internet-history <
internet-history at elists.isoc.org> wrote:

> Back in the day, while I was at BBN, I worked on the GOSIP specs in
> addition to my work on the ARPAnet and TCP/IP, so I have some familiarity
> with the topic.

I was running Datacom for Masscomp.  We were trying to sell realtime
systems to Aerospace, Automotive and of course DoD.
So, I, too, took a long and hard familiarity with GOSIP and MAP.  And
because of the EU  (can you Airbus in particular) I learned to loath
X.* family (25, 3, 28, 21  and the directory cruft 400, 500).

We delivered an X.25 stack and X.3 pad for our European customers, and we
also provided some X.400 support for our mailers.
It was a sales check box.   Funny thing, little of it was actually sold or
used, best I could tell.



> INMO, I can imagine an alternate reality where if TCP/IP
> hadn't taken off as it did for whatever reason, then OSI would have taken
> its place, but things would have just been delayed.
>
We differ in our experience/view on this. I see essentially two issues,
both related to control/MetCalfe's Law.

To me, what made the Internet a success (and a vast improvement over
previous networks like ARPAnet) was based on what I like to call Dave
Clark's revelation [the storing being in the mid-1970s we did something
silly at CMU that caused Clark and company at MIT to have to reset
everything].  Dave is believed to have said something on the order of why
we are being affected by those numskulls in Pittsburgh.

My simple point is that IP was a* distribution network of networks, with
ease of communication, cooperation, and sharing as the north star but
designed so that what was on my network could not(should) negatively affect
what was on your network *[Morris worm discounted].

ISO was being pushed by the PTT and firms like DEC, who saw *central
control and being able to sell more of my services in my 'walled garden' as
the high order bit—not cooperation and sharing*. Manufacturing firms with a
closed factory floor saw no issue with this thinking - which, of course, is
why MAP was fine with them. There is no way Boeing's factory is going to be
connected to anything else.  But ... the walled garden thinking is in
direct violation of Metacalf's observation of value, which relies on
sharing and desire to communicate/share.

The question you can ask is, would the pressures of Metacalf's law have
started to bring some of those walls down? We look at the current market
to find our answer by examining what occurred with the cell and cable TV
industries.   We still have the same behavior -- be on my network, not
the other guys.

IMO: what would have happened in N different networks like we have today
with cell, each trying to compete with who had the better services and
offers?  Large enough forms would have connected to multiple of them.  You
would have ended up with content aggregators like we do with TV.

My few cents ... it would have been bleak (and confusing).
Clem