"First" is a dangerous word, isn't it? -- Will

From: Liam Proven Sent: Tuesday, October 29, 2019 10:01 AM Internet Protocol (IP) was developed in the very late 1970s, with the cutover of the ARPANET taking place 1/1/83. Prior to that, the underlying protocol was the one developed by Kleinrock et al. for the BBN IMP hardware. New to this list, are you? Rich

On Tue, Oct 29, 2019, 4:38 PM Rich Alderson via cctalk < cctalk at classiccmp.org> wrote: NCP was the immediate predecessor of TCP/IP. Was that the datagram format, more or less, 50 years ago? Warner

Until recently, a computer was a person that did calculations. I suspect that if they we smart enough to do the calculations they were smart enough to talk to each other. Dwight ________________________________ From: cctalk <cctalk-bounces at classiccmp.org> on behalf of ED SHARPE via cctalk <cctalk at classiccmp.org> Sent: Tuesday, October 29, 2019 12:11 PM To: cctalk at classiccmp.org <cctalk at classiccmp.org> Subject: Re: 50 yrs. ago today I will have to see if our old next boots! -Ed# In a message dated 10/29/2019 10:01:21 AM US Mountain Standard Time, cctalk at classiccmp.org writes: Tim Berners-Lee says it's the 50th anniversary of the first internet packets. I believe him more than pretty much anyone. It's also in multiple computer news stories today. The historic event was comms between heterogenous computers over a standardised protocol (IP, I think

surprising was intra-net. On Tue, 29 Oct 2019, Liam Proven via cctalk wrote: The first "internet" packet was certainly a significant event. But, calling it "The first inter-computer communication" is comparable to saying that Columbus was the first to think that the world was round and discovered America, or that Ford invented the automobile, or that Bill Gates invented software or at least operating systems, or that Steve Jobs invented computers. Some may beg to differ, or point out that those are not accurate descriptions of the events. "First"s are usually expanded into things that they aren't, and almost always fail to acknowledge those less "famous" who were already doing it. OK, I claim to be the first to say "first" is a bogus way to describe any historical event. It is how non-historians fail to comprehend historians. -- Grumpy Ol' Fred cisin at xenosoft.com

On Tue, 29 Oct 2019, Paul Koning wrote: not so sure, . . . I think that there had been others. BUT, first with the IP protocol, . . . so, . . . "first" means earliest that WE are aware of, . . . "first" means the one that our schoolbooks talk about, . . . Being in the history books may mean "most important", but not "first"; the textbooks in the schools are astonishingly inaccurate. And, yes, some of them are going to say that Steve Jobs invented computers. There were some Asians? quite a bit earlier. The residents did not evolve here.

On 10/29/19 5:50 PM, Jecel Assumpcao Jr. via cctalk wrote: Doesn't AUTODIN precede ARPAnet by a few years? --Chuck

On 10/29/19 8:50 PM, Jecel Assumpcao Jr. via cctalk wrote: The whole story of what was going on was far more complex and interesting. Funny thing was DECnet was in 1983 the largest world wide network period. By then is was well over 300 nodes and climbing fast. And none of it used IP or NCP though it did transport P packets encapsulated using DECnet. A lot of DEC hardware was involved in the DARPA/Arpanet. The network wars were warming up about then (1982ish) and it would take till the late 80s early 90s for IP to win that war. The big explosion was WWW. Other names or routable networks, Banyan vines, and IPX come to mind besides DECNET Phase III and IV. Allison

I am not so sure. Initially IBM's internal VNET network (and BITENT) didn't use SNA they used bisync RSCS (for mail and files) and bisync 3270 lines handled by PVM (Passthrough Virtual machine) Native SNA didn't come to VM until very late in the day.... Dave

And before DECnet learned to route (which came in Phase III) there was routing in Typeset-11 clusters. Those used the distance vector routing algorithm too, but it had nothing in common with DECnet other than the fact it used DMC-11s to communicate. As far as I know none of this has been preserved, unfortunately. paul

On 10/30/19 10:18 AM, Liam Proven via cctalk wrote: I may be a but US centric but hey I lived her all my life... its what I know. Clearly if it was happening here it was world wide. I was fortunate to be an early Compuserve user and BBS user and I also remember packet radio (amateur). That that network was interconnected to Arpanet at that time (81 as I remember and likely earlier) is significant. I really liked that I could email to Valbone or Galway plants or just as easily the east coast and Japan facilities. And still be present on the bitnet, news, and other services and places as well. It was one of the few large backbone networks that had mixed traffic. at that time the interconnect and transport was mostly private but included portals. Not so, the whole of Arpanet was my computer calls your computer a few times a day and exchanges info/news/email as The "NETWORK" if any was telco or lease lines. That was [all of the above and more] before there was and significant backbone to "hook up to". The backbone is the internet as we know it. True but the point was WWW made the need for the network connectivity "The Internet" became a commodity thing not unlike the interstate highways. WWW is not the internet it is a good reason for one and maybe the worst utilization for one (opine). However it forced more connectivity which is what the internet is. Hence my usage... it can be found ON the internet, not in. It was around far longer before that as my first contact was in the late 70s. It is but access for the nonacademic or small company was much later in the 80s. Back then I investigate getting connected and was basically you needed to know someone at a Uni and get UUCP going. It was around then the SOURCE and Compuserve became visible and BBS started to be interconnected likely by dialup. True as the first WAN was telco! Much of what came later either relied on telco tech (ESS and T1 lines any one remember them?) or built on it. For many here in the use the Carterphone decision made interconnecting easier and competitive for products and other services. Back before the 80s I used to provide UHF repeater service and I remember using leased lines ("copper pairs" without equalization) for interconnect (analog) and the cost was not trivial and the available interconnect length was limited (miles). Those Arpanet and the like users had the finances to have leased lines or even privately installed lines to connect points in their reach. it was a driver to have a common network for traffic and of course rules for how the traffic was formed and routed. Why I mentioned them. There were other some very limited like (M$)LANman, ad a bit of PTP stuff. I left out IBM by omission but they were also one of the big technology contributors of the day for backbone hardware and software. We should as they were existent and ideas grew from them and for periods even depended on them. As a result a more open set of specifications emerged. That was a contrast to the Telcos(PTT) and private companies. Like I said far more complex and very interesting. Its why "first" has to be specific as its at most a snapshot of one place and time. Its also a result of something or many somethings that came before. Without the context it is just a nice but largely incomplete picture that invites the question, there must be more? With the bigger picture the wow factor is not diluted but enhanced. After all like most technologies how we got to that point is often a good story to tell. and where it leads to after that as well. To think, it all grew from people on hills with flags, mirrors, or fires trying to send messages faster and further than physical means like runners and horses. Allison

ons 2019-10-30 klockan 13:17 -0400 skrev Paul Koning via cctalk: I remember stickers on modems and telephones (ie not televerket provided equipment) which said that this equipment is certified to be directly connected to televerket's telephone lines. But computer network equipment owned by the organization and used on the organization's premises ?? That i don't remember. PS Televerket : Sweden's state owned telephone monopoly, today known by the public as Telia company. Ellemtel the development organization was co-owned by Ericsson/LME/Three-bars and Televerket. DS PPS LME still exist in name basically as a holding company for Ericsson. DSS

On 10/29/19 8:21 PM, Fred Cisin via cctalk wrote: Maybe so, likely UUCP first. IP networks were a thing that sorta grew as standard developed. Either way it was well before the late 70s. Also Aloha-net was when? I thought that was the first wide area network that also used radios for links. Generally "first" works if your specific enough. But in the old Arpanet days if you said that on a list and were incomplete or wrong you got your head handed to you upside down. Sometimes gently, maybe. ;-) Allison

On 10/29/19 11:53 AM, William Sudbrink via cctalk wrote: I wonder if computer-to-computer connection dates back 70 years. (5->7) Same number digits, just one of them changed. --Chuck

On 10/29/19 4:19 PM, Fred Cisin via cctalk wrote: I suspect that the first inter-computer communication occurred during the 1930s or thereabouts when one Langley employee asked another if she'd like to have lunch. --Chuck