[ih] Fate of Alohanet
mbaer at cs.tu-berlin.de
Fri Apr 2 21:20:39 PDT 2010
>From Richard Binder, 03/13/2010 03:08 AM; just for the record; didn't
make it to the list despite the list being cc-ed.
re the arpanet connection: I can't recall our actual experiences with it
when it became operational, but I think the 1974 final report was just
saying that the software etc work wasn't completed till then.
Also, I think the terms "NCP" and "Telnet" may have been used loosely in
that report to refer to the Menehune software involved in arpanet
communication - since a TIP was involved and the aim was to let alohanet
users connect to arpanet hosts using their 'dumb' terminals (and not the
reverse), I think the menehune just set up temporary connections between
users and TIP ports and provided the necessary Telnet-prescribed flow
control (but I could be mistaken).
As for the manual retransmission required after the three automatic
attempts, I think that almost never happened. The channel traffic
loading was such that only an occasional single automatic retransmission
should have been necessary.
If you don't already have it, the paper "Aloha packet broadcasting -- a
retrospect", written near the end of the project in late 1974, is
probably the most definitive description of Alohanet and its operation
still around. It was given at the May 1975 NCC and is available for a
fee from acm at http://portal.acm.org/citation.cfm?id=1499985 .
btw, why the interest in alohanet?
Matthias Bärwolff wrote:
> Mind you, may I add two more questions to the soup:
> Did the Arpanet connection actually ever work fine? The final report
> indicates that only by 1974 did NCP and Telnet work well enough on the
> Alohanet side so as to allow terminal connections over to the Arpanet.
> And, did people from the Arpanet connect to Alohanet terminals, too?
> Second, what about the experiences with the retransmission scheme which
> in the initial design required users to reinitiate retransmission after
> three failed transmission attempts for a packet. How did that work out.
> I'd suspect it must have happened quite often that such user initiated
> retransmissions were required.
> Richard Binder wrote:
>> Matthias ,
>> I believe the arpanet connection was via a TIP at UH, but can't recall
>> the connection details from the Menehune side. I left Hawaii in early
>> 1975, but a good person to ask about the project's fate is Frank Kuo
>> (ffkuo at mindspring.com).
>>> from Bob Kahn:
>>>>> The Alohanet funding began (I think, because it was before my
>>>>> time) with funds from AFOSR. Then, around 1969, DARPA got into the
>>>>> picture and augmented the AF funding (via AFOSR as agent). Most
>>>>> likely the last funding was in FY 1974 or 75.
>>> If you Google "richard binder alohanet" you should turn up the final
>>> report of the Alohanet project that is dated late in 1974 so Bob's
>>> guess as to funding appears to be correct. As far as landlines, I
>>> don't think they got any better. The project itself successfully
>>> demonstrated the feasibility of the stochastic method for sharing
>>> capacity and by mid-1973, Bob Metcalfe, stimulated by his exposure to
>>> the Alohanet project, had invented and demonstrated Ethernet at Xerox
>>> PARC. The Internetting project, initiated by bob kahn at ARPA had
>>> already started in 1973 and was well underway in 1974 at Stanford.
>>> Packet Radio and Packet Satellite were also well underway and these
>>> also used some of the Alohanet ideas. In some sense, these other
>>> projects instantiated the Alohanet notions in more powerful, higher
>>> speed forms and it might have been thought that the Hawaiian project
>>> would not yield more beneficial results.
>>> Dick Binder is copied and may have more precision to offer.
>>> On Feb 25, 2010, at 4:40 AM, Matthias Bärwolff wrote:
>>>> Dear all,
>>>> I gather from the literature (largely Abramson and Kuo) that Alohanet
>>>> got connected to the Arpanet by means of an IMP at the Hawaii
>>>> in late 1972; then, by 1974 they had NCP and Telnet sufficiently up in
>>>> the Menehune to allow terminal connections to the Arpanet; and,
>>>> in 1976 the whole project died for lack of further funding. No further
>>>> information was provided for the latter point.
>>>> Does anyone know a specific reason why they discontinued Alohanet? Did
>>>> the landlines get better, and thus the raison d'etre vanished? What
>>>> happened to the IMP, did it stay connected to the Arpanet?
More information about the internet-history