On 2006-07-31 22:28:52 +0100, "Doctor Drivel" said:


"Andy Hall" wrote in message ...
On 2006-07-31 16:34:27 +0100, "Doctor Drivel" said:


"Andy Hall" wrote in message ...

One version split it into the 7 layer OSI stack, when it looked as if
OSI was going to be the way forward, which it was and should have been.

Another blunder. You seem not to know the difference between the OSI
*model* and the ISO *stack*.

Matt, you haven't a clue. I have had this with you a long time ago and
it was clear you didn't have a clue about OSI.

Actually, you just demonstrated

Matt, it is clear you haven't a clue, or half at best. OSI was
primarily a "connectionless" protocol, which you didn't know.

More rubbish. There are connection based and connectionless network layers.



Its selling point was that it was connectionless. No handshake, which
takes time and resources, you just send. It anticipated reliable fast
infrastructure. Although at the last minute they brought out a
connection oriented protocol for situations where a handshake was
essential.

This is complete tosh.

The situation was exactly the opposite way around. Most of the OSI
devotees came from telco backgrounds and the use of X.25. They
naturally gravitated towards a connection oriented protocol and this is
why 4 out of the 5 TP transport classes require a connection oriented
network layer.
Performance was poor and so TP4 is able to use a connectionless or
connection oriented network layer. Unsurprisingly, TP4 is based on
TCP.




very clearly that the boot is on the other foot.

Lord Hall's or Matt's foot?

The ISO protocol stack never really stood a realistic chance of broad
adoption. That became apparent as early as the early to mid 80s.

Balls Matt. It was in vogue and was heavily funded until the w.w.w. came in.

No it wasn't.

Matt, it was and even after.

There was never any realistic likelihood of widespread ISO protocol deployment.

Matt, there was as all the governments and the EU were pushing it.
Even in the US NISK were involved.

That would certainly have killed it if the committees hadn't.



Then an inferior TCP/IP was adopted which didn't have enough scope for
all the addresses, as it was a cobbled together improvise in the first
place.

That is also rubbish.

Nonsense Matt. Read Tenambaum, well the earlier versions. All sorts of
clever IP address jiggery pokery was formulated to keep the crock
going. The only people who pushed TCP/IP were private companies who had
a vested interest in keeping OSI out.

Like the U.S. Department of Defense for example. When did Uncle Sam
outsource that to private enterprise?



TCP/IP was put together in Snowbird near Salt Lake City. I've been to
the hotel where a bunch of students zipped up this inadequate 5 layer
stack on backs of envelopes. OSI was deemed to be carrying too much
baggage in the headers ay the time.

.... and so it does. This is why it is so little used. Some
telephone switch equipment still uses it, but it's unusual to find it
other than that.



Today with high speed networks this is not a problem. It was stated
that it would be fine when infrastructure caught up. You could also
have null layers if you liked to speed it up.

Th rapid spread of the Internet and the w.w.w., which had not adopted
OSI as it was still being implemeted in various government departments
and had not quite reached the rest, killed OSI. Nothing else. It was
too late to turn back the TCP/IP protocol. If the www had been two
years later it probably would have had an OSI protocol stack.

No it wouldn't.