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

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

On 2019-02-17 02:03, Michael Kj?rling wrote:
> On 15 Feb 2019 16:29 -0500, from vint at google.com (Vint Cerf):
>> 6. MOSAIC hits about 1993 followed by Netscape Communications and its IPO.
> I'm curious about this point of yours. Why do you say that this was an
> important deciding factor in the choice between OSI vs TCP/IP, or in
> the commercialization of TCP/IP over OSI?
> I'm not saying that the Web was unimportant -- it absolutely was, and
> is (just look at how many people think that the Internet _is_ the
> World Wide Web) -- but is there something about the protocols which
> makes it much easier to run HTTP over TCP/IP than over OSI?

At the relevant time (1991, when the first line-mode web browser was
developed, along with HTTP itself) the only general-purpose substrate
that Tim Berners-Lee, Robert Cailliau and Nicola Pellow found at their
disposal was TCP/IP. So most simply put, they had no choice. Running
HTTP over OSI was never an option in the real world.

fwiw, it was by then two years after the rant** which marked my own
epiphany. Tim, Robert and Nicola were just users...


** B.E. Carpenter, Is OSI Too Late?, RARE Networkshop 1989, Computer Networks and ISDN Systems, 17 (1989) 284-286, DOI 10.1016/0169-7552(89)90040-8

> Or was it
> just the fact that by the time the Web came about, the world was
> somewhat firmly established in the TCP/IP camp and so it became a
> natural choice to focus on running HTTP over TCP/IP?
> I'll readily admit that my knowledge of the OSI stack is limited at
> best.