[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[ih] When did "32" bits for IP register as "not enough"?

Brian E Carpenter <brian.e.carpenter at gmail.com> writes:

> On 2019-02-16 08:23, Clem Cole wrote:
>> On Thu, Feb 14, 2019 at 2:58 PM Craig Partridge <craig at tereschau.net> wrote:
>>> Dennis' decision to adopt TCP/IP for NSFNET was critical.
>> Hmmm.  It certainly did not hurt and I agree was cleafrly an amplifier and
>> important.   But I'm not sure that was as important as the basic economics
>> of the time -> Moore's Law and cheap cycles (i.e. I don't think the NFSNET
>> choice, while helped the effect, was not the high order bit on the success
>> function).
> I took the original question to be about when the success of TCP/IP became
> formally recognized. You are undoubtedly right that its market success was
> a result of availability in BSD, with a usable API, just when Unix was taking
> off.

My original question was kind of overly vague - we'd found a netnews
conversation in 89 between jon postel, vint and others that seemingly
led to both CIDR and ipv6 ("toaster-net"), as "32 bits is not enough")

>From a formal perspective the class B exhaustion documents mentioned on
these threads, at two ietfs, kicked restructuring the internet to handle
more addresses into higher gear, and it was realized that OSI had no
future benefit once that was addressed.


The writing on the wall moment for IP over OSI seems to be from around
the time of the "kobe revolt" from what I've been reading here. All the
events in 1992-94 - not just the rework of address spaces, but the
opening up of the internet, ISPs getting off the ground, cheap modems,
al gore, etc, etc....


going off topic yet again:

I do think people overstate the early impact of the web nowadays.

It was being able to talk to more people you had something in common
with than ever before.

When I started my ISP in 94, it was netnews and email that were the
primary draws, and the web thing was... well... a "browser". Yes, it
turned out folk had a huge appetite for cat videos, but huge numbers of
folk in college had been exposed to "real email" and netnews and wanted
that at home and it was within reach of the average geek, finally.

Everybody already had modems to contact compuserve and BBSes, (didn't
the first laptops also have integral modems?) so when slip started
working, BBSes started to die. There were very few Unix based BBSes (I
had a hand in one in 86, and this one in 1990, which amazingly still
exists: https://cellar.org/faq.php ) and their draw became "being

we had real apps also (client/server databases and so on), that helped
drive corporate adoption of IP vs ipx. Dialup access to big databases
for remote offices was done not via browser but a huge host of tools
(now departed) that created a custom UI element over that sql database.

I've been sitting here trying to remember some of those. Borland had
some, there was a really popular one (power-something) - hundreds and
hundreds of dollars per seat that corps were willing to spend to get
their remote offices online.

It was a heady time! 


OK, here's a story that I rarely tell and is OT:

in late 1993 I'd been having a bad time on every front and hadn't slept
well for a few months, and my network was under attack, and I was
essentially seeing snakes under the bed. I had a discussion with a
shrink, and at some point I said something like "my machines talk to
millions of people every day". (I was trying to explain netnews). She
wrote (and I read, upside down):

"The patient is grandiose and delusional".

(Thankfully I escaped from that particular interview relatively unscathed
and later found a shrink with a physics background that I could actually
talk to which straightened me out a lot)

A year later we'd found funding for ICANECT and a few months after that
we had 16,000 people online doing the same things that clueless shrink
had dismissed as fantasy.

Me, grandiose? *always*

Delusional? Sometimes I worry, but not very often, anymore.