Turing Award Won by Co-Inventor of Ethernet Technology (nytimes.com) 32
In the 1970s, Bob Metcalfe helped develop the primary technology that lets you send email or connect with a printer over an office network. From a report: In June 1972, Bob Metcalfe, a 26-year-old engineer fresh out of graduate school, joined a new research lab in Palo Alto, Calif., as it set out to build something that few people could even imagine: a personal computer. After another engineer gave up the job, Dr. Metcalfe was asked to build a technology that could connect the desktop machines across an office and send information between them. The result was Ethernet, a computer networking technology that would one day become an industry standard. For decades, it has connected PCs to servers, printers and the internet in corporate offices and homes across the globe.
For his work on Ethernet, the Association for Computing Machinery, the world's largest society of computing professionals, announced on Wednesday that Dr. Metcalfe, 76, would receive this year's Turing Award. Given since 1966 and often called the Nobel Prize of computing, the Turing Award comes with a $1 million prize. When Dr. Metcalfe arrived at the Palo Alto Research Center -- a division of Xerox nicknamed PARC -- the first thing he did was connect the lab to the Arpanet, the wide-area network that later morphed into the modern internet. The Arpanet transmitted information among about 20 academic and corporate labs across the country. But as PARC researchers designed their personal computer, called the Alto, they realized they needed a network technology that could connect personal computers and other devices within an office, not over long distances. Further reading: Ethernet Creator Makes the Inventors Hall of Fame (2007).
For his work on Ethernet, the Association for Computing Machinery, the world's largest society of computing professionals, announced on Wednesday that Dr. Metcalfe, 76, would receive this year's Turing Award. Given since 1966 and often called the Nobel Prize of computing, the Turing Award comes with a $1 million prize. When Dr. Metcalfe arrived at the Palo Alto Research Center -- a division of Xerox nicknamed PARC -- the first thing he did was connect the lab to the Arpanet, the wide-area network that later morphed into the modern internet. The Arpanet transmitted information among about 20 academic and corporate labs across the country. But as PARC researchers designed their personal computer, called the Alto, they realized they needed a network technology that could connect personal computers and other devices within an office, not over long distances. Further reading: Ethernet Creator Makes the Inventors Hall of Fame (2007).
Supernova (Score:4, Informative)
In the issue of Infoworld dated December 4, 1995; Metcalfe wrote that the Internet “will soon go spectacularly supernova and in 1996 will collapse.” He identified a number of factors would bring about the Internet’s collapse, including security breaches, capacity overloads, and demand for video online.
Whatever happened to the Internets?
Re: (Score:2)
Re:Supernova (Score:5, Informative)
His pundit career was notable for its failures... the Internet didn't suffer a gigalapse, WiMax didn't destroy wifi, Open Source is still around, ...
Re: (Score:2)
The internet actually did collapse. This collapse of triggered a cascade effect deep in the planet core triggering a lot of Higgs Boson particles to shrug and just give up, thus collapsing the entire universe. Immediately a near duplicate universe sprang into existence due to mathematical models still pending, with the only difference being a new internet that was deemed too big to collapse. In fact the sheer amount of useless content on the internet, comprised mostly of dark matter, is what prevent the
Re: (Score:2, Interesting)
I had an office down the hall from Metcalfe for a brief time at another post after he left PARC and before 3Com. There was quite a lot of staff slosh between Xerox and my organization and we were aware of Alto and Ethernet, though had direct experience with neither. We were having a new building constructed, and thinking forward a little from our then mainframe terminals centered work, I passed a memo up the line suggesting that empty conduit and blank wall plates should be added to all offices for networ
Re: (Score:2)
Metcalfe went on to fame, and I went on to contribute to security breaches, capacity overloads, and demand for video online.
Anonymous Coward, we salute you!
Re: (Score:1)
Re: (Score:1)
You must not be keeping up with the cyber attack news, something close is in progress.
Comment removed (Score:5, Interesting)
Re: (Score:2)
There was a network connection I saw once described as "ethernet" which were two very very long wires stretching through a tunnel between buildings (said tunnel for transporting secure documents where nosy above ground dwellers couldn't spy). They were placed one foot apart with spacers the whole distance. It might not have been ethernet, someone may have just used that term as a generic word they overheard.
We had normal vampire coax cables too. They were a bit clumsy but not impractical. Much of the head
Re: (Score:2)
Coax was a nightmare.
Re: (Score:2)
The first wi-fi, ALOHAnet [hawaii.edu] from 1971, used a shared medium and faced the same problems. I wonder if Mr Metcalfe was aware of their work?
Re: (Score:1)
To be fair, having a single snake-like cable connecting all the PCs was a pretty common idea. To us today it seems obviously easier to have individual connections to a central hub or switch, but at the time a common cable was accepted, perhaps to save on the total amount of wiring used. Token Ring also used a single wire, though joined into a loop, and I think there was also Token Bus.
You are right that "transmit, and retry on collision" can never scale. But it is simple and hard to get wrong, as long as
Re: (Score:2)
Cancer (Score:1)
Ethernet certainly has seen fantastic wild growth, supplanting just about every other networking technology. It even stifles innovation because nobody can come up with a "better ethernet" and thrive. Thus, we're stuck with it, and its 1500 bytes fit for cheap 2kB chips of forty years ago.
Re: (Score:2)
There are tons of non-Ethernet layer 2 protocols running on any number of different PHY layers. Just because you don't get your pr0n through them doesn't mean they're not thriving.
Re:Cancer (Score:5, Informative)
Re: (Score:2)
Before the RJ-45 based Ethernet cable was a thing, there was thin Ethernet which used a co-axial cable to inter-connect nodes such as PCs. Thin Ethernet used CSMA/CD (Carrier Sense Multiple Access / Collision Detection). This means that only 1 Ethernet frame can be in transit at a time which means collision mitigation is needed.
For thin Ethernet, the maximum Ethernet frame length is constrained by the worst case delay time to be able to transmit and the transmission time. In other words, the performance is
Re: (Score:3)
Proper jumbo frames are difficult for IP because the TCP and UDP checksums are terribly weak. The practical limit is in ~10000 bytes. However, if you e.g. run everything encrypted, a lot of equipment supports ~16000 bytes. The theoretical maximum is ~65000 bytes for IPv4 and ridiculously large for IPv6.
Re: (Score:2)
Re: (Score:2)
ACM book on this (Score:2)
I've recently been working my way through an excellent book published by the ACM following the companies and products which resulted in the modern networking stack we take for granted today: Circuits, Packets and Protocols: Entrepreneurs and Computer Communications, 1968-1988 [amazon.com]. It's an amazing history of dozens of companies, products and people most of which have been largely forgotten or (at best) relegated to footnotes.
RS-232C (Score:1)
Re: (Score:1)
I believe that the inventor of RS-232C adapter (serial adapter, known from the times of modems) and the Kermit file transfer tool should also get the Turing award. The more the merrier.
Kermit was pretty decent, but most of us used Zmodem back then.
The nice thing about ethernet (Score:5, Insightful)
Good for Bob! (Score:2)
MIT boy (Score:2)
MIT boy makes good.
Why didn't ChatGPT get it? (Score:1)
Didn't Turing predict ChatGPT in 1950?
ALOHAnet (Score:3)
My picture of things is that Ethernet was basically ALOHAnet adapted to use the shared medium of wires rather than the shared medium of radio spectrum. I think the hard work went into creating that, and the easier work was reading the papers and seeing how it could use a different medium.