Re: Re[2]: Proper Webs

[dski -at- CAMEONET -dot- CAMEO -dot- COM -dot- TW]

Steve Fouts (stefou -at- eskimo -dot- com) wrote --

> I didn't catch the original message in this thread, but "the web" is not
> globally unique as the Internet is. The Internet was not, originally, the
> only net. There was Bitnet (the "Because It's Time" network), ARPAnet,
> Usenet, lots of nets. One of each mind you, and one Internet. The Internet
> eventually assimilated many of the other nets. There are still thousands
> of networks but only one Internet.

The Internet is the only Net, but it has never been the only net --
neither the only network nor the only internetwork.

Bitnet still exists, as far as I know, and still uses its own complete
protocol suite, as do others. Bitnet (maybe all of it, maybe only part)
and some other global internetworks are "gatewayed" to the Internet (that
is, made accessible on the Internet via machines that translate between
different protocols), but they have not been "assimilated" -- that would
mean they could be accessed _only_ through Internet connections.

Usenet and the World Wide Web were designed as Internet applications; you
might say they are "nets within the Net"; they have their own protocols,
but those protocols are incomplete without a transport mechanism such as
the Internet's TCP/IP (I'll bet some purist will come along and say that
"transport" is not the right term, but I'm using it loosely).

The only thing unique about the Internet is its combination of age,
openness, complete protocol suite, and number of connections. Any company
can set up a globe-girdling corporate internetwork using TCP/IP, and it
seems safe to assume that many have (though they may all have converted
to some other protocol suite by now for security reasons).

> Similarly, there are many webs but only one World Wide Web. That one is
> on the Internet. I've got a web on my intranet. Couple HTTP servers, an
> FTP server or two. Poof, instant web. I might work for a global company
> has an intranet that is truly global. I can call that web a world-wide
> web but I can't call it the World Wide Web.

No argument there (except that I wish "intranet" hadn't caught on as the
standard term for a corporate internetwork). Your hypothetical company's
HQ might have a LAN on which some machines use TCP/IP and some use
Novell's IPX; of those using IPX, some might use frame type A, and some
frame type B, C, or D. That's five logical networks, each completely
inaccessible to the others without special gateways, on one physical
network. The global internetwork is sort of like that. The largest TCP/IP
part of the global internetwork is what we call the Internet.

> Lastly, as to the question of why the Internet is capitalized and the
> telephone system is not. Convention. By convention, the web that exists
> on the Internet is called the World Wide Web. It is globally unique. It
> is not synonymous with the Internet. By convention, the globally TCP/IP
> network over which we are now conversing is called the Internet.

Again, right on the button. But if Bitnet is still operating the way it
did not so long ago, some folks are getting this via gateways that, if
suddenly shut down, would make techwr-l inaccessible to many but not all
of us (or maybe it would leave us with two or more baby techwr-ls).

Dan Strychalski   dski -at- cameonet -dot- cameo -dot- com -dot- tw