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

[ih] Fwd: [IP] OSI: The Internet That Wasn't

On 31/07/2013 09:13, Bill Nowicki wrote:
>  Yes, many of us probably have anecdotes that confirm what Noel and others are saying. The standards process might play a role, but working software also does. Here is my story. 
> I was the lone TCP/IP person at Sun Microsystems in the 1980s. I would maybe buy Van Jacobson an espresso, and then integrate his improvements into Sun's OS along with other software from the BSD community as I found the time. Meanwhile there was a large (for Sun) dedicated group developing the ISO OSI stack. Marketing was convinced that only a few researchers might use TCP, while businesses would use OSI, so they wanted 20 times the people working on OSI.
> We published the NFS protocol as an "informational" RFC since it had not (yet) gone through any standards process. However, the implementations of NFS worked with each other. In particular, I was called in on a support issue for one customer: the Corporation for Open Systems. They of course used NFS over TCP/IP to get their work done, which was ironically promoting OSI. It was I think Milo Medin who realized the letter of the GOSIP mandate was that federal agencies needed to have the OSI protocol "available" for any equipment they bought, but did not need to actually buy it, let alone use it.

I was told by somebody, probably somebody in ESNET or at Fermilab, that they
had to order GOSIP-compliant code and pay for it, but they didn't have to open
the box (presumably a box of floppy disks in those days).