[ih] Internet addressing history section

Noel Chiappa jnc at mercury.lcs.mit.edu
Thu Feb 14 08:33:44 PST 2019


    > From: Paul Ruizendaal

    > why did tcp/ip win out over ipx/spx. There is probably a multitude of
    > reasons, economic, social and technical. ... my hypothesis is that
    > tcp/ip was the only networking stack in the late 80's/early 90's that
    > did well in both WAN and LAN contexts.

Not so much WAN/LAN (the PUP network had both LANs and WANs), although
inclusion of WANs does bring in some issues, e.g. adaptive timeouts.

I think a bigger factor was that TCP/IP was at least attempting to build a
_large_ network, and also starting to think about the issues involved in
building a 'multi-owner' network (which of course are not just technical, but
also managerial). Anyone can build a small internet, but scaling it up is
really hard.

And that feeds in to what I still think is the killer - Metcalfe's Law, and
the user base size. I had a saying I used to people at Proteon (many of whom
were still focused on LANs) a lot: 'The neatest thing about the phone on
your desk is not that it can call someone else in the company, but that it
can call anyone anywhere.'

	Noel

PS: I would be great if this list could focus on _one_ topic at a time.  This
ping-ponging from 'total silence for weeks' -> '4 active threads at once' is
not optimal if the goal is carefully thought out replies. You (non-specific)
have a topic you want to discuss? Stick it in your pocket until the list goes
quiet for a day...



More information about the Internet-history mailing list