[ih] Fwd: Re: Early Internet history

[Dave Crocker]

with Tom's permission.


-------- Forwarded Message --------
Subject: Re: Early Internet history
Date: Wed, 4 Jul 2018 09:41:02 -0400
From: Tom Van Vleck <***@multicians.org>
To: Dave Crocker <***@bbiw.net>


> On Jul 3, 2018, at 10:17 PM, Dave Crocker <***@bbiw.net> wrote:
>
> I'm trying to make a point of notifying anyone I mentioned during the
> interview, mostly in case I got something wrong. I've already had
> two corrections to notify the Collective folks about...

Hi Dave, very interesting!

(I am not a fan of podcasts.. too linear. My superpower is READING and
I can read faster than I can listen.)

There are a few things I would have said a little differently than you did.

MAIL

The late Noel Morris and I did the CTSS MAIL command together, and he
should not be forgotten.
We wrote MAIL because it had been proposed in a MIT Comp Center memo by
Louis Pouzin, Glenda Schroeder, and Pat Crisman, all of whom worked at
MIT on CTSS for F. J. Corbató. We wanted to use it. Nobody had time to
write MAIL though.
Noel and I were users of CTSS and asked if we could write MAIL and
contribute it to the system.

Your description of CTSS MAIL mixes implementation and function.
Functionally, a user wrote a small text file, and then typed MAIL MY
FILE M1416 786 to send it. (Files on CTSS had two-word names. User
addresses on CTSS had a "problem number" like M1416 and a "programmer
number" assigned by the Comp Center.) The recipient would be notified if
his MAIL BOX was non-empty, and could view it with the PRINT command.
Each user's MAIL BOX file had the file mode "private" meaning that only
the account owner could read or modify it -- except that
system-privileged programs could do so.

As we were implementing MAIL, we wrote the absolute minimum program that
would work.
There were many suggested features, options, etc. We left them out.
Idea being to get it working, and put features in later after community
discussion.

The essential features of MAIL were
- user to user, i.e. personal
- asynchronous
- file based, persistent
- secure

Messages were
- identified by sender and date
- limited in size to one disk record: 2592 BCD characters

We didn't send between computers because we had only one computer. We
didn't send graphics because nobody had a graphics terminal. We didn't
have SUBJECT any other mail headers; but conventions rapidly sprang up.
We limited messages in size because disk space was very scarce and
expensive.


Multics

As the editor of Multicians.org I am sometimes vexed by people's
theories about Unix. Ken and Dennis would probably not have "gone off on
their own" and written Unix if Bell Labs management had not dropped out
of the Multics project in 1969. The size and scope of Unix was a result
of the resources available to its creators. Ken and Dennis made useful
contributions to Multics and never questioned its scope or elaborateness
in my hearing while they worked on the system. Ken was one of the
smartest people I have ever met: his clarity of thought and writing were
an inspiration, and he was a great programmer and colleague.


regards, tom

[Vint Cerf]

what was the time frame for CSS Mail?

v


On Wed, Jul 4, 2018 at 2:43 PM, Dave Crocker <***@dcrocker.net> wrote:

> with Tom's permission.
>
>
> -------- Forwarded Message --------
> Subject: Re: Early Internet history
> Date: Wed, 4 Jul 2018 09:41:02 -0400
> From: Tom Van Vleck <***@multicians.org>
> To: Dave Crocker <***@bbiw.net>
>
>
> > On Jul 3, 2018, at 10:17 PM, Dave Crocker <***@bbiw.net> wrote:
> >
> > I'm trying to make a point of notifying anyone I mentioned during the
> > interview, mostly in case I got something wrong. I've already had
> > two corrections to notify the Collective folks about...
>
> Hi Dave, very interesting!
>
> (I am not a fan of podcasts.. too linear. My superpower is READING and
> I can read faster than I can listen.)
>
> There are a few things I would have said a little differently than you did.
>
> MAIL
>
> The late Noel Morris and I did the CTSS MAIL command together, and he
> should not be forgotten.
> We wrote MAIL because it had been proposed in a MIT Comp Center memo by
> Louis Pouzin, Glenda Schroeder, and Pat Crisman, all of whom worked at
> MIT on CTSS for F. J. Corbató. We wanted to use it. Nobody had time to
> write MAIL though.
> Noel and I were users of CTSS and asked if we could write MAIL and
> contribute it to the system.
>
> Your description of CTSS MAIL mixes implementation and function.
> Functionally, a user wrote a small text file, and then typed MAIL MY
> FILE M1416 786 to send it. (Files on CTSS had two-word names. User
> addresses on CTSS had a "problem number" like M1416 and a "programmer
> number" assigned by the Comp Center.) The recipient would be notified if
> his MAIL BOX was non-empty, and could view it with the PRINT command.
> Each user's MAIL BOX file had the file mode "private" meaning that only
> the account owner could read or modify it -- except that
> system-privileged programs could do so.
>
> As we were implementing MAIL, we wrote the absolute minimum program that
> would work.
> There were many suggested features, options, etc. We left them out.
> Idea being to get it working, and put features in later after community
> discussion.
>
> The essential features of MAIL were
> - user to user, i.e. personal
> - asynchronous
> - file based, persistent
> - secure
>
> Messages were
> - identified by sender and date
> - limited in size to one disk record: 2592 BCD characters
>
> We didn't send between computers because we had only one computer. We
> didn't send graphics because nobody had a graphics terminal. We didn't
> have SUBJECT any other mail headers; but conventions rapidly sprang up.
> We limited messages in size because disk space was very scarce and
> expensive.
>
>
> Multics
>
> As the editor of Multicians.org I am sometimes vexed by people's
> theories about Unix. Ken and Dennis would probably not have "gone off on
> their own" and written Unix if Bell Labs management had not dropped out
> of the Multics project in 1969. The size and scope of Unix was a result
> of the resources available to its creators. Ken and Dennis made useful
> contributions to Multics and never questioned its scope or elaborateness
> in my hearing while they worked on the system. Ken was one of the
> smartest people I have ever met: his clarity of thought and writing were
> an inspiration, and he was a great programmer and colleague.
>
>
> regards, tom
>
>
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>



--
New postal address:
Google
1875 Explorer Street, 10th Floor
Reston, VA 20190

[Jack Haverty]

On 07/05/2018 05:11 AM, Vint Cerf wrote:
> what was the time frame for CSS Mail?
>
> v

In 1968/9 I was a regular user of CTSS and remember using MAIL. It
wasn't new then, just another command in the system.

If "email" is considered broadly as electronic communication, I remember
using IBM JCL in 1967 to send messages from one human to another, via
punch cards - e.g., asking the operator to mount a tape.

Prior to that, of course there was the telegraph and telegrams. Perhaps
that would be "electric mail".

I was in Paris recently and spent several hours at the Musee des Arts et
Metiers, essentially a museum of technology. One section is devoted to
"Communications". I noticed one display cabinet containing a machine
that was somehow used to "allow several telegraph operators to share the
same wire" - so I guess Multiplexing has been around since the 19th century.

Apparently humans need to communicate, and as each new means of
transporting messages comes around, someone figures out a way to use it
to talk with others.

There's lots of interesting old stuff preserved in those glassed-in
display cases.

When you look into one of those cases and see a well-worn and carefully
preserved piece of history, and your immediate reaction is "Hey, I used
to use one of those!" -- that's when you know you're getting old.....

/Jack Haverty

[Łukasz Bromirski]

Jack,

Watch out, this self-promoting lunatic is going to US Senate and he may sue you at some point in time for misrepresenting his own version of history:

https://arstechnica.com/tech-policy/2018/06/inventor-of-email-appeals-ruling-that-tossed-his-libel-suit-against-techdirt/

;)

--
./

> On 5 Jul 2018, at 17:34, Jack Haverty <***@3kitty.org> wrote:
>
>> On 07/05/2018 05:11 AM, Vint Cerf wrote:
>> what was the time frame for CSS Mail?
>>
>> v
>
> In 1968/9 I was a regular user of CTSS and remember using MAIL. It
> wasn't new then, just another command in the system.
>
> If "email" is considered broadly as electronic communication, I remember
> using IBM JCL in 1967 to send messages from one human to another, via
> punch cards - e.g., asking the operator to mount a tape.
>
> Prior to that, of course there was the telegraph and telegrams. Perhaps
> that would be "electric mail".
>
> I was in Paris recently and spent several hours at the Musee des Arts et
> Metiers, essentially a museum of technology. One section is devoted to
> "Communications". I noticed one display cabinet containing a machine
> that was somehow used to "allow several telegraph operators to share the
> same wire" - so I guess Multiplexing has been around since the 19th century.
>
> Apparently humans need to communicate, and as each new means of
> transporting messages comes around, someone figures out a way to use it
> to talk with others.
>
> There's lots of interesting old stuff preserved in those glassed-in
> display cases.
>
> When you look into one of those cases and see a well-worn and carefully
> preserved piece of history, and your immediate reaction is "Hey, I used
> to use one of those!" -- that's when you know you're getting old.....
>
> /Jack Haverty
>
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

[Richard Bennett]

Nobody goes to the Senate without winning an election. This self-promoting lunatic is running as a Republican in Massachusetts against Elizabeth Warren. His chances of winning a lawsuit are better than winning that election. Dude has’t even won the primary, which will be held on Sept. 4.

> On Jul 5, 2018, at 10:02 AM, Łukasz Bromirski <***@bromirski.net> wrote:
>
> Jack,
>
> Watch out, this self-promoting lunatic is going to US Senate and he may sue you at some point in time for misrepresenting his own version of history:
>
> https://arstechnica.com/tech-policy/2018/06/inventor-of-email-appeals-ruling-that-tossed-his-libel-suit-against-techdirt/ <https://arstechnica.com/tech-policy/2018/06/inventor-of-email-appeals-ruling-that-tossed-his-libel-suit-against-techdirt/>
>
> ;)
>
> --
> ./
>
> On 5 Jul 2018, at 17:34, Jack Haverty <***@3kitty.org <mailto:***@3kitty.org>> wrote:
>
>> On 07/05/2018 05:11 AM, Vint Cerf wrote:
>>> what was the time frame for CSS Mail?
>>>
>>> v
>>
>> In 1968/9 I was a regular user of CTSS and remember using MAIL. It
>> wasn't new then, just another command in the system.
>>
>> If "email" is considered broadly as electronic communication, I remember
>> using IBM JCL in 1967 to send messages from one human to another, via
>> punch cards - e.g., asking the operator to mount a tape.
>>
>> Prior to that, of course there was the telegraph and telegrams. Perhaps
>> that would be "electric mail".
>>
>> I was in Paris recently and spent several hours at the Musee des Arts et
>> Metiers, essentially a museum of technology. One section is devoted to
>> "Communications". I noticed one display cabinet containing a machine
>> that was somehow used to "allow several telegraph operators to share the
>> same wire" - so I guess Multiplexing has been around since the 19th century.
>>
>> Apparently humans need to communicate, and as each new means of
>> transporting messages comes around, someone figures out a way to use it
>> to talk with others.
>>
>> There's lots of interesting old stuff preserved in those glassed-in
>> display cases.
>>
>> When you look into one of those cases and see a well-worn and carefully
>> preserved piece of history, and your immediate reaction is "Hey, I used
>> to use one of those!" -- that's when you know you're getting old.....
>>
>> /Jack Haverty
>>
>>
>>
>> _______
>> internet-history mailing list
>> internet-***@postel.org <mailto:internet-***@postel.org>
>> http://mailman.postel.org/mailman/listinfo/internet-history <http://mailman.postel.org/mailman/listinfo/internet-history>
>> Contact list-***@postel.org <mailto:list-***@postel.org> for assistance.
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

—
Richard Bennett
High Tech Forum <http://hightechforum.org/> Founder
Ethernet & Wi-Fi standards co-creator

Internet Policy Consultant

[John Day]

Most railroad telegraphs were ‘party lines,’ multi-access media, etc.

> On Jul 5, 2018, at 11:34, Jack Haverty <***@3kitty.org> wrote:
>
> On 07/05/2018 05:11 AM, Vint Cerf wrote:
>> what was the time frame for CSS Mail?
>>
>> v
>
> In 1968/9 I was a regular user of CTSS and remember using MAIL. It
> wasn't new then, just another command in the system.
>
> If "email" is considered broadly as electronic communication, I remember
> using IBM JCL in 1967 to send messages from one human to another, via
> punch cards - e.g., asking the operator to mount a tape.
>
> Prior to that, of course there was the telegraph and telegrams. Perhaps
> that would be "electric mail".
>
> I was in Paris recently and spent several hours at the Musee des Arts et
> Metiers, essentially a museum of technology. One section is devoted to
> "Communications". I noticed one display cabinet containing a machine
> that was somehow used to "allow several telegraph operators to share the
> same wire" - so I guess Multiplexing has been around since the 19th century.
>
> Apparently humans need to communicate, and as each new means of
> transporting messages comes around, someone figures out a way to use it
> to talk with others.
>
> There's lots of interesting old stuff preserved in those glassed-in
> display cases.
>
> When you look into one of those cases and see a well-worn and carefully
> preserved piece of history, and your immediate reaction is "Hey, I used
> to use one of those!" -- that's when you know you're getting old.....
>
> /Jack Haverty
>
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

[Brian E Carpenter]

On 06/07/2018 03:34, Jack Haverty wrote:
...
> I was in Paris recently and spent several hours at the Musee des Arts et
> Metiers, essentially a museum of technology. One section is devoted to
> "Communications". I noticed one display cabinet containing a machine
> that was somehow used to "allow several telegraph operators to share the
> same wire" - so I guess Multiplexing has been around since the 19th century.

It was perceived as a requirement very early. Wikipedia dates it to the 1870s.
But I think the first really successful version was the "Murray Multiplex" in
about 1909. My colleague Bob Doran has studied Donald Murray's works at some length:
https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/Murray/MurraySpielLR.pdf

Brian

[Vint Cerf]

alexander graham bell had been trying to find a way to multiplex several
concurrent telegraph signals on the same line when he ended up carrying
voice - of course there is also the Elisha Gray initiative that may
actually have preceded Bell's experiment.

v


On Thu, Jul 5, 2018 at 4:59 PM, Brian E Carpenter <
***@gmail.com> wrote:

> On 06/07/2018 03:34, Jack Haverty wrote:
> ...
> > I was in Paris recently and spent several hours at the Musee des Arts et
> > Metiers, essentially a museum of technology. One section is devoted to
> > "Communications". I noticed one display cabinet containing a machine
> > that was somehow used to "allow several telegraph operators to share the
> > same wire" - so I guess Multiplexing has been around since the 19th
> century.
>
> It was perceived as a requirement very early. Wikipedia dates it to the
> 1870s.
> But I think the first really successful version was the "Murray Multiplex"
> in
> about 1909. My colleague Bob Doran has studied Donald Murray's works at
> some length:
> https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/
> Murray/MurraySpielLR.pdf
>
> Brian
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>



--
New postal address:
Google
1875 Explorer Street, 10th Floor
Reston, VA 20190

[Richard Bennett]

So there’s really nothing novel about the Internet. Digital communication has been done with end-to-end control of multiplexed packets since the early 20th century, and the phone network was just a diversion.

Cool.

RB

> On Jul 5, 2018, at 2:59 PM, Brian E Carpenter <***@gmail.com> wrote:
>
> On 06/07/2018 03:34, Jack Haverty wrote:
> ...
>> I was in Paris recently and spent several hours at the Musee des Arts et
>> Metiers, essentially a museum of technology. One section is devoted to
>> "Communications". I noticed one display cabinet containing a machine
>> that was somehow used to "allow several telegraph operators to share the
>> same wire" - so I guess Multiplexing has been around since the 19th century.
>
> It was perceived as a requirement very early. Wikipedia dates it to the 1870s.
> But I think the first really successful version was the "Murray Multiplex" in
> about 1909. My colleague Bob Doran has studied Donald Murray's works at some length:
> https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/Murray/MurraySpielLR.pdf
>
> Brian
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

—
Richard Bennett
High Tech Forum <http://hightechforum.org/> Founder
Ethernet & Wi-Fi standards co-creator

Internet Policy Consultant

[Craig Partridge]

Earlier -- 1790s (the so-called Napoleonic telegraph -
https://www.bbc.com/news/magazine-22909590). It interleaved messages from
multiple spots, I believe.

Craig

On Thu, Jul 5, 2018 at 3:50 PM, Richard Bennett <***@bennett.com> wrote:

> So there’s really nothing novel about the Internet. Digital communication
> has been done with end-to-end control of multiplexed packets since the
> early 20th century, and the phone network was just a diversion.
>
> Cool.
>
> RB
>
>
> On Jul 5, 2018, at 2:59 PM, Brian E Carpenter <***@gmail.com>
> wrote:
>
> On 06/07/2018 03:34, Jack Haverty wrote:
> ...
>
> I was in Paris recently and spent several hours at the Musee des Arts et
> Metiers, essentially a museum of technology. One section is devoted to
> "Communications". I noticed one display cabinet containing a machine
> that was somehow used to "allow several telegraph operators to share the
> same wire" - so I guess Multiplexing has been around since the 19th
> century.
>
>
> It was perceived as a requirement very early. Wikipedia dates it to the
> 1870s.
> But I think the first really successful version was the "Murray Multiplex"
> in
> about 1909. My colleague Bob Doran has studied Donald Murray's works at
> some length:
> https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/
> Murray/MurraySpielLR.pdf
>
> Brian
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>
> —
> Richard Bennett
> High Tech Forum <http://hightechforum.org> Founder
> Ethernet & Wi-Fi standards co-creator
>
> Internet Policy Consultant
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>


--
*****
Craig Partridge's email account for professional society activities and
mailing lists.

[Larry Press]

Before the French semaphore network, the ancient Greeks had a system for sending digital messages:


http://cis471.blogspot.com/2018/07/telecommunication-in-ancient-greece.html

[Vint Cerf]

well, you might want to think about scale for a moment.

v


On Thu, Jul 5, 2018 at 5:50 PM, Richard Bennett <***@bennett.com> wrote:

> So there’s really nothing novel about the Internet. Digital communication
> has been done with end-to-end control of multiplexed packets since the
> early 20th century, and the phone network was just a diversion.
>
> Cool.
>
> RB
>
>
> On Jul 5, 2018, at 2:59 PM, Brian E Carpenter <***@gmail.com>
> wrote:
>
> On 06/07/2018 03:34, Jack Haverty wrote:
> ...
>
> I was in Paris recently and spent several hours at the Musee des Arts et
> Metiers, essentially a museum of technology. One section is devoted to
> "Communications". I noticed one display cabinet containing a machine
> that was somehow used to "allow several telegraph operators to share the
> same wire" - so I guess Multiplexing has been around since the 19th
> century.
>
>
> It was perceived as a requirement very early. Wikipedia dates it to the
> 1870s.
> But I think the first really successful version was the "Murray Multiplex"
> in
> about 1909. My colleague Bob Doran has studied Donald Murray's works at
> some length:
> https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/
> Murray/MurraySpielLR.pdf
>
> Brian
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>
> —
> Richard Bennett
> High Tech Forum <http://hightechforum.org> Founder
> Ethernet & Wi-Fi standards co-creator
>
> Internet Policy Consultant
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>


--
New postal address:
Google
1875 Explorer Street, 10th Floor
Reston, VA 20190

[Jack Haverty]

Interesting. Is the Internet now larger than the Phone network --
measured in terms of numbers of entities (people and organizations) who
can interact? I still know lots of people who have phones but "don't do
email".

One thing that sticks in my mind as being "novel" in the Internet world
is one of the techniques used to achieve today's large scale.

AFAIK, the Internet is the first implementation of a "network" (i.e., a
data communications system using lots of distributed processors to move
bits) that is done with the components being designed, manufactured,
installed, and operated independently but cooperatively by more than one
organization. Earlier networks of computers, e.g., the ARPANET, IBM's
multi-drop networks, et al, were all under control of one entity. Some
such networks got to be pretty large, but were always under single
management. The Internet changed that, and IMHO that facilitate its
growth. Somebody will correct me if I'm wrong...

Of course, the international Phone network was multi-organizational, at
least at the level of each country involved and their local PTT
responsible for their part. So in effect the Internet followed the lead
of the Phone system to enable the interoperation of disjoint pieces.

IMHO, one of the key pieces enabling this technique was the creation of
EGP. I remember sitting with Eric Rosen in my office at BBN to figure
out how to respond to Bob Kahn's directive to "make it possible for
people other than BBN to build gateways". We batted around ideas for a
few hours, with the knowledge that the PTTs seemed to have been somehow
able to do this. EGP was the result. Novel? Maybe...

There were probably many other such techniques that enabled scale. For
example, the notion of defining a protocol and formats "on the wire"
seems also important. That was part of the ARPANET, but was likely
preceded by similar techniques applied in places like memory and I/O
busses, or railroad tracks, etc.

The recent discussion of "invention" got me thinking. It seems that,
every once in a while, something truly new is created. I'd argue that
an example would be the Stored Program Computer of von Neumann's vision.
Then, for years or decades afterwards, that new element is
opportunistically applied to do things that were done before in a
different way. As computers got small and cheap enough, packet switches
became viable as replacements for human phone and telegraph operators or
mechanical components (like I saw in the Paris museum)

Which of all these things are "inventions" will be debated forever...

/Jack


On 07/05/2018 03:25 PM, Vint Cerf wrote:
> well, you might want to think about scale for a moment.
>
> v
>
>
> On Thu, Jul 5, 2018 at 5:50 PM, Richard Bennett <***@bennett.com
> <mailto:***@bennett.com>> wrote:
>
> So there’s really nothing novel about the Internet. Digital
> communication has been done with end-to-end control of multiplexed
> packets since the early 20th century, and the phone network was just
> a diversion. 
>
> Cool.
>
> RB
>
>
>> On Jul 5, 2018, at 2:59 PM, Brian E Carpenter
>> <***@gmail.com <mailto:***@gmail.com>>
>> wrote:
>>
>> On 06/07/2018 03:34, Jack Haverty wrote:
>> ...
>>> I was in Paris recently and spent several hours at the Musee des
>>> Arts et
>>> Metiers, essentially a museum of technology.  One section is
>>> devoted to
>>> "Communications".  I noticed one display cabinet containing a machine
>>> that was somehow used to "allow several telegraph operators to
>>> share the
>>> same wire" - so I guess Multiplexing has been around since the
>>> 19th century.
>>
>> It was perceived as a requirement very early. Wikipedia dates it
>> to the 1870s.
>> But I think the first really successful version was the "Murray
>> Multiplex" in
>> about 1909. My colleague Bob Doran has studied Donald Murray's
>> works at some length:
>> https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/Murray/MurraySpielLR.pdf
>> <https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/Murray/MurraySpielLR.pdf>
>>
>>   Brian
>>
>> _______
>> internet-history mailing list
>> internet-***@postel.org <mailto:internet-***@postel.org>
>> http://mailman.postel.org/mailman/listinfo/internet-history
>> <http://mailman.postel.org/mailman/listinfo/internet-history>
>> Contact list-***@postel.org <mailto:list-***@postel.org> for
>> assistance.
>
> —
> Richard Bennett
> High Tech Forum <http://hightechforum.org> Founder
> Ethernet & Wi-Fi standards co-creator
>
> Internet Policy Consultant
>
>
> _______
> internet-history mailing list
> internet-***@postel.org <mailto:internet-***@postel.org>
> http://mailman.postel.org/mailman/listinfo/internet-history
> <http://mailman.postel.org/mailman/listinfo/internet-history>
> Contact list-***@postel.org <mailto:list-***@postel.org> for
> assistance.
>
>
>
>
> --
> New postal address:
> Google
> 1875 Explorer Street, 10th Floor
> Reston, VA 20190
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>

[Richard Bennett]

Telegraph had an interesting way of bridging incompatible national networks. At national boundaries, PTTs had terminals for their own network and for their neighbors'. When messages came in, they were written down and then carried to operators of the desired terminal by office boys. The messages were then rekeyed and sent on their way.

> On Jul 5, 2018, at 5:36 PM, Jack Haverty <***@3kitty.org> wrote:
>
> Interesting. Is the Internet now larger than the Phone network --
> measured in terms of numbers of entities (people and organizations) who
> can interact? I still know lots of people who have phones but "don't do
> email".
>
> One thing that sticks in my mind as being "novel" in the Internet world
> is one of the techniques used to achieve today's large scale.
>
> AFAIK, the Internet is the first implementation of a "network" (i.e., a
> data communications system using lots of distributed processors to move
> bits) that is done with the components being designed, manufactured,
> installed, and operated independently but cooperatively by more than one
> organization. Earlier networks of computers, e.g., the ARPANET, IBM's
> multi-drop networks, et al, were all under control of one entity. Some
> such networks got to be pretty large, but were always under single
> management. The Internet changed that, and IMHO that facilitate its
> growth. Somebody will correct me if I'm wrong...
>
> Of course, the international Phone network was multi-organizational, at
> least at the level of each country involved and their local PTT
> responsible for their part. So in effect the Internet followed the lead
> of the Phone system to enable the interoperation of disjoint pieces.
>
> IMHO, one of the key pieces enabling this technique was the creation of
> EGP. I remember sitting with Eric Rosen in my office at BBN to figure
> out how to respond to Bob Kahn's directive to "make it possible for
> people other than BBN to build gateways". We batted around ideas for a
> few hours, with the knowledge that the PTTs seemed to have been somehow
> able to do this. EGP was the result. Novel? Maybe...
>
> There were probably many other such techniques that enabled scale. For
> example, the notion of defining a protocol and formats "on the wire"
> seems also important. That was part of the ARPANET, but was likely
> preceded by similar techniques applied in places like memory and I/O
> busses, or railroad tracks, etc.
>
> The recent discussion of "invention" got me thinking. It seems that,
> every once in a while, something truly new is created. I'd argue that
> an example would be the Stored Program Computer of von Neumann's vision.
> Then, for years or decades afterwards, that new element is
> opportunistically applied to do things that were done before in a
> different way. As computers got small and cheap enough, packet switches
> became viable as replacements for human phone and telegraph operators or
> mechanical components (like I saw in the Paris museum)
>
> Which of all these things are "inventions" will be debated forever...
>
> /Jack
>
>
> On 07/05/2018 03:25 PM, Vint Cerf wrote:
>> well, you might want to think about scale for a moment.
>>
>> v
>>
>>
>> On Thu, Jul 5, 2018 at 5:50 PM, Richard Bennett <***@bennett.com
>> <mailto:***@bennett.com>> wrote:
>>
>> So there’s really nothing novel about the Internet. Digital
>> communication has been done with end-to-end control of multiplexed
>> packets since the early 20th century, and the phone network was just
>> a diversion.
>>
>> Cool.
>>
>> RB
>>
>>
>>> On Jul 5, 2018, at 2:59 PM, Brian E Carpenter
>>> <***@gmail.com <mailto:***@gmail.com>>
>>> wrote:
>>>
>>> On 06/07/2018 03:34, Jack Haverty wrote:
>>> ...
>>>> I was in Paris recently and spent several hours at the Musee des
>>>> Arts et
>>>> Metiers, essentially a museum of technology. One section is
>>>> devoted to
>>>> "Communications". I noticed one display cabinet containing a machine
>>>> that was somehow used to "allow several telegraph operators to
>>>> share the
>>>> same wire" - so I guess Multiplexing has been around since the
>>>> 19th century.
>>>
>>> It was perceived as a requirement very early. Wikipedia dates it
>>> to the 1870s.
>>> But I think the first really successful version was the "Murray
>>> Multiplex" in
>>> about 1909. My colleague Bob Doran has studied Donald Murray's
>>> works at some length:
>>> https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/Murray/MurraySpielLR.pdf
>>> <https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/Murray/MurraySpielLR.pdf>
>>>
>>> Brian
>>>
>>> _______
>>> internet-history mailing list
>>> internet-***@postel.org <mailto:internet-***@postel.org>
>>> http://mailman.postel.org/mailman/listinfo/internet-history
>>> <http://mailman.postel.org/mailman/listinfo/internet-history>
>>> Contact list-***@postel.org <mailto:list-***@postel.org> for
>>> assistance.
>>
>> —
>> Richard Bennett
>> High Tech Forum <http://hightechforum.org> Founder
>> Ethernet & Wi-Fi standards co-creator
>>
>> Internet Policy Consultant
>>
>>
>> _______
>> internet-history mailing list
>> internet-***@postel.org <mailto:internet-***@postel.org>
>> http://mailman.postel.org/mailman/listinfo/internet-history
>> <http://mailman.postel.org/mailman/listinfo/internet-history>
>> Contact list-***@postel.org <mailto:list-***@postel.org> for
>> assistance.
>>
>>
>>
>>
>> --
>> New postal address:
>> Google
>> 1875 Explorer Street, 10th Floor
>> Reston, VA 20190
>>
>>
>> _______
>> internet-history mailing list
>> internet-***@postel.org
>> http://mailman.postel.org/mailman/listinfo/internet-history
>> Contact list-***@postel.org for assistance.
>>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

—
Richard Bennett
High Tech Forum <http://hightechforum.org/> Founder
Ethernet & Wi-Fi standards co-creator

Internet Policy Consultant

[Joly MacFie]

On Thu, Jul 5, 2018 at 8:11 PM, Richard Bennett <***@bennett.com> wrote:

> Telegraph had an interesting way of bridging incompatible national
> networks. At national boundaries, PTTs had terminals for their own network
> and for their neighbors'. When messages came in, they were written down and
> then carried to operators of the desired terminal by office boys. The
> messages were then rekeyed and sent on their way.
>
>
>
​Hunter Newby, formerly a manager at Telx at 60 Hudson tells this story. In
the 20s & 30s in NYC Western Union at 60 Hudson​, and AT&T at 32 Ave of A
had to exchange traffic, So they, as Richard says, would exchange paper -
not by office boys but by clay pneumatic tubes under the streets.

Come the 80s, MCI won a consent decree that they could access AT&T's long
distance. However, to do so they needed to colocate at 32 Ave of A. Sorry,
AT&T said, no room at the Inn. MCI were fortunate enough to find some old
lineman who revealed the existence of the old, and pretty much disused clay
tubes. Meanwhile, the lower floors of 60 Hudson were vacant as all tenants
wanted to be up in the light. MCI snapped them up, and pulled wire through
to AT&T's basement. "We're here." they said "Hook us up!" and AT&T couldn't
refuse, and thus it was those tubes that essentially broke open the phone
business.

Later, Telx used the same conduit to establish itself as an IXP.

In 2010 I shot vid of Hunter telling the story. It's about 10 mins in on
http://isoc-ny.org/1637

More recently he has managed to dig up a picture of the original telegraph
routers which, like computers in those days, were women.

That can be seen at around 4:25 on
https://livestream.com/internetsociety/de-cix/videos/138995706

j

--
---------------------------------------------------------------
Joly MacFie 218 565 9365 Skype:punkcast
--------------------------------------------------------------
-

[Steffen Nurpmeso]

Richard Bennett wrote in <9B0F604C-B099-482F-8212-***@bennett.com>:
|Telegraph had an interesting way of bridging incompatible national \
|networks. At national boundaries, PTTs had terminals for their own \
|network and for
|their neighbors'. When messages came in, they were written down and \
|then carried to operators of the desired terminal by office boys. The \
|messages were
|then rekeyed and sent on their way.

Last year the German (Culture) Radio Station Deutschlandfunk had
in their series "The Long Night Of ..." a play (it is art a.k.a.
sophisticated, unless it is about persons where it is more like an
interview) about the visual, then electromagnetical Telegraph
system, including technical problems (isolation, crossing the
ocean)[1]. The full text log is luckily at [2] (PDF around, too;
in German, but for German speaking interesting also as a read).
It contains an episode with text from 1867 about the
"Central-Telegraphenstation zu Berlin", for example

[1] https://www.deutschlandfunk.de/eine-lange-nacht-ueber-telegramme-jedes-wort-zaehlt.704.de.html?dram:article_id=387933
[2] https://www.deutschlandfunk.de/29-07-2017-lange-nacht-jedes-wort-zahlt-eine-lange-nacht.media.40d807840f38ebb2ee05efb6272d5353.txt

[electricity..]
die wunderbare Kraft [erzeugen], deren Wirkung die Menschheit
the wonderful power, whose virtue moved humanity
näher zusammengerückt hat, als die größten Eroberungszüge es
closer together, than the largest campaign of conquest ever did.
vermochten. Aeußerst interessant ist die Wahrnehmung, in wie
Most interesting is the perception
hohem Grade die Sinne der Beamten in der Ausführung ihres
of the high grade of awareness of the senses of the officials when
Berufes geschärft werden. Da bei dem Arbeiten der Apparate die
executing their profession.
Angabe der Zeichen durch den Apparat stets in einem gewissen
Rhythmus erfolgt, so vermögen einige der geübtesten
Telegraphisten eine ankommende Depesche nach dem Gehör zu lesen,
noch ehe die Zeichen auf das Papier gedrückt ihnen zukommen.
Allerdings ist dies nur in der dem Beamten geläufigen Sprache
möglich; fremde Sprachen telegraphirt der Nichtkenner mechanisch
nach, indem er die ihm zukommenden Zeichen nachdruckt und
..as has been said, input to paper, and the re-telegraphed with
new battery power.
wiedergiebt. Die Berliner Central-Telegraphenstation ist ein
The Berlin Central-Telegraphenstations is a
Stück Welt für sich, eine fast zauberische Welt, deren Eindrücke
world on its own, an almost magical world, whose impressions
auch dann nicht weichen, wenn wir aus dem Hauptthore des großen
do not vanish when we step through the main gate of the great
Gebäudes schreitend in den Lärm der Straßen gelangen und
building into the noise of the streets, and
zurückblicken auf jene Stätte, wo "Himmelskräfte auf und nieder
look back to the place where the "forces of heaven rise up and
steigen und sich die goldnen Eimer reichen."
down and hand each other the golden buckets."

Hmhm. There is also

..
"Nach diesen Vorbemerkungen begeben wir uns in den Apparatsaal
der Centralstation, die eigentliche Werkstätte und zugleich den
Sammelpunkt jener dienstbar gemachten ungeheuern Kraft, welche
wir gemeinhin die elektrische oder galvanische nennen. Eine jede
Station, größere oder kleinere, setzt sich zusammen aus 1) den
Batterien, 2) den Schlüsseln, 3) dem Schreibeapparat, 4) den
Hülfsapparaten. Die Central-Station correspondirt auf vier
The Central-Station corresponds on four
verschiedenen Linien. Auf den internationalen Linien gehen die
distinct lines. The international lines transport
Depeschen Tag und Nacht ohne Unterbrechung, denn im großen
telegrams Day and Night without discontinuity, because in the big
Weltverkehr giebt es keinen Tag und keine Nacht. Die internationalen
World traffic there is no Day and no Night. Moreover, the international
Linien verbinden außerdem nur direct die Hauptstädte und sie
lines connect directly only the capitals, and they
betragen ein Viertel des gesamten telegraphischen Verkehrs, der also
account for a quarter of the entire telegrafical traffic, which
ganz der Central-Station zufällt. Die zweite Linie verbindet die
also entirely falls to the Central-Station. The second line
größeren Städte des preußischen Telegraphennetzes, z. B. Berlin,
connects the greater Cities ...
Köln, Königsberg, Frankfurt etc. Auf diesen Linien wird z. B. mit
Königsberg und Frankfurt direct verhandelt und gesprochen. Diese
..These
Leitungen gehören zu den vorzüglichsten, denn ohne Einschaltung
lines rank to the most exquisite...
besonderer Kraft arbeiten die Apparate und Drähte und wirken
unmittelbar in die bedeutendsten Entfernungen mit großer Präcision.
Die dritte Art wird die große Omnibus-Linie genannt. Sie verbindet
The third is called the big Omnibus-Line. It connects
die mittleren Städte des Netzes. Die vierte Abtheilung oder die kleine
the midize Cities of the grid. The fourth division, or the small
Omnibuslinie vermittelt den Verkehr zwischen Punkten, die nur zwei
Omnibus-Line, mediates traffic in between points which lie apart
bis vier Meilen auseinander liegen. Sie schließen sich an größere
no further than two to four Kilometres. They connect to the greater
Linien an, um auf solche Weise in den allgemeinen, großen Verkehr
lines in order to become incorporated in the general big traffic
aufgenommen zu werden."
by this means.

--steffen
|
|Der Kragenbaer, The moon bear,
|der holt sich munter he cheerfully and one by one
|einen nach dem anderen runter wa.ks himself off
|(By Robert Gernhardt)

[Miles Fidelman]

On 7/5/18 4:36 PM, Jack Haverty wrote:

> AFAIK, the Internet is the first implementation of a "network" (i.e., a
> data communications system using lots of distributed processors to move
> bits) that is done with the components being designed, manufactured,
> installed, and operated independently but cooperatively by more than one
> organization. Earlier networks of computers, e.g., the ARPANET, IBM's
> multi-drop networks, et al, were all under control of one entity. Some
> such networks got to be pretty large, but were always under single
> management. The Internet changed that, and IMHO that facilitate its
> growth. Somebody will correct me if I'm wrong...
>
Kind of sounds like the international postal system.  Or shipping
packages internationally.  (For a very loose definition of "data
communications.")  I seem to recall folks referring to to packets as
"mailgrams" every once in a while.

And, I believe that both USENET (or at least UUCP) and BITNET predated
the TCP/IP cutover by a couple of years.  They were both decentralized
and "operated independently but cooperatively by more than one
organization."

Miles Fidelman








--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra

[Scott O. Bradner]

> On Jul 6, 2018, at 12:45 AM, Miles Fidelman <***@meetinghouse.net> wrote:
>
…

>
> And, I believe that both USENET (or at least UUCP) and BITNET predated
> the TCP/IP cutover by a couple of years. They were both decentralized
> and "operated independently but cooperatively by more than one
> organization."

TCP/IP - developed in mid 1970s
BITNET - founded 1981 by Ira Fuchs
USENET - established in 1980 by Tom Truscott
(UUCP was the file copy protocol not the system)
TCP/IP cutover - 1983

[Vint Cerf]

BITNET and USENET eventually linked to INTERNET by way of email relays.

v


On Fri, Jul 6, 2018 at 8:09 AM, Scott O. Bradner <***@sobco.com> wrote:

>
>
> > On Jul 6, 2018, at 12:45 AM, Miles Fidelman <***@meetinghouse.net>
> wrote:
> >
> 

>
> >
> > And, I believe that both USENET (or at least UUCP) and BITNET predated
> > the TCP/IP cutover by a couple of years. They were both decentralized
> > and "operated independently but cooperatively by more than one
> > organization."
>
> TCP/IP - developed in mid 1970s
> BITNET - founded 1981 by Ira Fuchs
> USENET - established in 1980 by Tom Truscott
> (UUCP was the file copy protocol not the system)
> TCP/IP cutover - 1983
>
>
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>



--
New postal address:
Google
1875 Explorer Street, 10th Floor
Reston, VA 20190

[Olivier MJ Crépin-Leblond]

Two quick comments:

1. one could argue that Internet eventually got linked to BITNET by way
of relays. :-)
No my first comment was that as BITNET was a store and forward network,
to my knowledge the only gateways there were, were email gateways. Or do
you remember of a way to interface from a BITNet real time session to
TCP/IP?

2. My understanding of USENET is that it was not a network as such, but
most people referred to it as the system of newsgroups using a clever
hierarchy and peer to peer for dissemination (NNTP). Again, it was a
store and forward network, this time using the UUCP suite and dial-up.
Ironically most USENET traffic as well as UUCP traffic eventually ended
up carried on TCP-IP. Still working these days.

Kindest regards,

Olivier

On 06/07/2018 14:27, Vint Cerf wrote:
> BITNET and USENET eventually linked to INTERNET by way of email relays.
>
> v
>
>
> On Fri, Jul 6, 2018 at 8:09 AM, Scott O. Bradner <***@sobco.com
> <mailto:***@sobco.com>> wrote:
>
>
>
> > On Jul 6, 2018, at 12:45 AM, Miles Fidelman
> <***@meetinghouse.net <mailto:***@meetinghouse.net>>
> wrote:
> >
> 

>
> >
> > And, I believe that both USENET (or at least UUCP) and BITNET
> predated
> > the TCP/IP cutover by a couple of years.  They were both
> decentralized
> > and "operated independently but cooperatively by more than one
> > organization."
>
> TCP/IP - developed in mid 1970s
> BITNET - founded 1981 by Ira Fuchs
> USENET - established in 1980 by Tom Truscott
>         (UUCP was the file copy protocol not the system)
> TCP/IP cutover - 1983
>
>
>
>
> _______
> internet-history mailing list
> internet-***@postel.org <mailto:internet-***@postel.org>
> http://mailman.postel.org/mailman/listinfo/internet-history
> <http://mailman.postel.org/mailman/listinfo/internet-history>
> Contact list-***@postel.org <mailto:list-***@postel.org> for
> assistance.
>
>
>
>
> --
> New postal address:
> Google
> 1875 Explorer Street, 10th Floor
> Reston, VA 20190
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

--
Olivier MJ Crépin-Leblond, PhD
http://www.gih.com/ocl.html

[Miles Fidelman]

On 7/6/18 10:33 AM, Olivier MJ Crépin-Leblond wrote:

>
>
> 2. My understanding of USENET is that it was not a network as such,
> but most people referred to it as the system of newsgroups using a
> clever hierarchy and peer to peer for dissemination (NNTP). Again, it
> was a store and forward network, this time using the UUCP suite and
> dial-up. Ironically most USENET traffic as well as UUCP traffic
> eventually ended up carried on TCP-IP. Still working these days.
>

That's a good point.  I too remember USENET as referring more to the
newsgroups than to uucp transport.

Do I recall correctly that "uunet" was the term used to denote the
network of machines that could be reached by bang path addressing?

Miles


--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra

[Paul Vixie]

Miles Fidelman wrote:
> ...
>
> Do I recall correctly that "uunet" was the term used to denote the
> network of machines that could be reached by bang path addressing?

no.

usenet really was just the newsgroups.

the uucp mapping project defined the set of bang-path-addressible nodes.

neither usenet nor the uucp network could have existed without the other.

uunet was a company.

--
P Vixie

[Clem Cole]

On Fri, Jul 6, 2018 at 2:48 PM, Paul Vixie <***@redbarn.org> wrote:

> usenet really was just the newsgroups.
>
​A squishy definition that could be argued - I'm not sure which I side I
pick and I lived the period and helped with some of the code at the time.​

​IIRC​
​ Mark (now Mary Ann) Horton used to ​distinguish between the set of UUCP
nodes (UUCPnet) and the Newsgroups nodes with the term 'newsnet'

>
> the uucp mapping project defined the set of bang-path-addressible nodes.
>
​Right...
​

>
> neither usenet nor the uucp network could have existed without the other.
>
I don't think that is really correct, because they can into being
differently.

I remember Truscott describing the his 3 or 4 site UUCP network at the
Boulder USENIX ​(Circa 1979 I think) -- 'The Black Hole'. I was
personally fancianted with his description of their home grown dialer using
a relay to short Tip and Ring on their POTS line, because they could not
accord an real one from Western Electric. When we return from the
conference, Horton at Berkeley, me at Teklabs and ber at Whippany added our
nodes to Truscotts network within a few weeks. aps added decvax within 6
months (I don't remember who was behind ihnp4 which was also an early
node).

They key point was that the uucp network took shape from the internal BTL
network that was basically private already and when Truscott announced what
he had duplicated something for a couple of Universities in research
triangle, and was offering let anyone else be part. We all did. BTL was
not on the Arpanet and they already had UUCP so it was not bug deal to join
UUCP for mail.

The newsgroups part was about 3 years later



>
> uunet was a company.
>
​And was to solve a big problem the UUCPnet had. Before USENIX forked off
the UUNET Experiment, it also forked off another ​using the cable TV system
who's exact name escapes me -- I want to say Stargate - that broadcast the
news traffic over cable TV. The problem was that it was discovered that
decvax had an over $600/yr phone bill. Somebody at Indian Hill had
researched their phone bill which was similar, but has realized for ever
every long distance call they placed it was generating 10-20 down stream
and most of the those were going to AT&T, so it was worth it. But we all
knew the system was on borrowed time and better solution was needed. Rick
Adam made a proposal to the USENIX BOD to try a at cost system that could
use commercial systems to try bring the cost down. UUNET was born. It was
successful and was eventually forked off as its own company. The Stargate
experiemtn was abandoned.

ᐧ

[Paul Vixie]

Clem Cole wrote:
> neither usenet nor the uucp network could have existed without the
> other.
>
> I don't think that is really correct, because they can into being
> differently.

i agree that they came into being differently. but without usenet, the
distribution of horton's maps would not have been possible, and without
uucpnet, replies to newsgroup postings would not have been possible. so
i ought to have said "neither could have achieved the scale it did
without the other". thanks for this correction.

> uunet was a company.
>
> ​... Rick Adam made a proposal to the USENIX BOD to
> try a at cost system that could use commercial systems to try bring the
> cost down. UUNET was born. It was successful and was eventually forked
> off as its own company. ...

incidentally, when uunet technologies (a non-profit) recast itself as
uunet communications services (a for-profit), the money left over was
used to launch internet systems consortium (a non-profit).

--
P Vixie

[Clem Cole]

On Fri, Jul 6, 2018 at 3:58 PM, Paul Vixie <***@redbarn.org> wrote:

>
>
> Clem Cole wrote:
>
>> neither usenet nor the uucp network could have existed without the
>> other.
>>
>> I don't think that is really correct, because they can into being
>> differently.
>>
>
> i agree that they came into being differently. but without usenet, the
> distribution of horton's maps would not have been possible, and without
> uucpnet, replies to newsgroup postings would not have been possible. so i
> ought to have said "neither could have achieved the scale it did without
> the other". thanks for this correction.
>
>> Interresting... Paul I think you are referring to on the issue UUCPnet
that should have been a sign for the Internet.

The UUCPnet grew incredibly fast because it was easy and reasonably cheap
to attach ... but quickly the 'routing problem' emerged. In traditional
UNIX style, UUCP had been design without worrying about some problems --
UUCP was thinking small scale so how mail (packets or whatever got there)
was not an issue. You did your own routing. But with many, thousands of
nodes, this was a huge problem.

Netnews gets layered on top of UUCP and because of the growth, and the wild
nature, this lack of support for routinr quickly becomes an issue. So as
Paul mentioned the 'mapping project' got started and there became an effort
to layer some concept of sanity of routing (at least for news messages).
But message/network sub-routing was never really well thought, much less a
solved.

Years later the actual same issue would show up, with networking routing
once it became cheap to add to the IP network. But, it just took maybe 10
years because the technology to attach many, many IP networks cheaply over
long distances had to drop in cost enough before IP hit the same wall.




> incidentally, when uunet technologies (a non-profit) recast itself as
> uunet communications services (a for-profit), the money left over was used
> to launch internet systems consortium (a non-profit).

+1 ​Indeed, as a one time President of USENIX, the whole UUNET/ISC, Open
Access for all documents legacy are some of a many cools things I'm very
proud of the organizations legacy.

Clem​

ᐧ

[Paul Ruizendaal]

> On 6 Jul 2018, at 22:34, Clem Cole <***@ccc.com> wrote:
>
> In traditional UNIX style, UUCP had been designed without worrying about some problems -- UUCP was thinking small scale so how mail (packets or whatever got there) was not an issue. You did your own routing. But with many, thousands of nodes, this was a huge problem.

In the 70’s the Murray Hill team was more oriented towards circuit switching than packet switching. Although uucp was an ad-hoc solution, maybe Mike Lesk’s thinking in ’76 was that uucp could eventually run over Spider/Datakit networks. In such a setup routing would be a thing that uucp delegated to the network.

[Clem Cole]

On Sat, Jul 7, 2018 at 3:52 AM, Paul Ruizendaal <***@planet.nl> wrote:

>
> In the 70’s the Murray Hill team was more oriented towards circuit
> switching than packet switching. Although uucp was an ad-hoc solution,
> maybe Mike Lesk’s thinking in ’76 was that uucp could eventually run over
> Spider/Datakit networks. In such a setup routing would be a thing that uucp
> delegated to the network.
>
>
​Well yes, the whole telephony idea was a full connectivity like the phone
system, so host!user by definition made sense. The problem was UUCP added
host1!host...n!user semantics which broke that model. I remember talking
to Chesson about it once​ who was the Datakit architect.

I admit, I do not 'know' but I >>suspect<< it was just not something anyone
considered less because of circuit switching and more because the scale was
small (local-ish) originally.

I wish Greg were still here to ask more questions ;-)

Clem
ᐧ

[Miles Fidelman]

On 7/6/18 11:48 AM, Paul Vixie wrote:

>
>
> Miles Fidelman wrote:
>> ...
>>
>> Do I recall correctly that "uunet" was the term used to denote the
>> network of machines that could be reached by bang path addressing?
>
> no.
>
> usenet really was just the newsgroups.
>
> the uucp mapping project defined the set of bang-path-addressible nodes.
>
> neither usenet nor the uucp network could have existed without the other.
>
Bringing this full circle...
Jack asked if there were any networks, before the Internet, that moved
bits around, across systems owned by multiple people.
I suggested BITNET & the UUCP network.
Somebody said - "that's called USENET."
The question was intended to clarify that USENET and the underlying UUCP
network were (are) separate beasts.
The ensuing discussion has made that abundantly clear - and dredged up
some additional interesting history.

Cheers,

Miles



--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra

[Lyndon Nerenberg]

> That's a good point.  I too remember USENET as referring more to the
> newsgroups than to uucp transport.

Yes, Usenet was the collection of newsgroups, originally shipped over
UUCP.

> Do I recall correctly that "uunet" was the term used to denote the
> network of machines that could be reached by bang path addressing?

No, 'uunet' was a commercial UUCP hub, started up by Rick Adams. It
handled a great volume of mail and news traffic, as well as an extensive
public file archive. It eventually morphed into Alternet, one of the
very first commercial Internet providers.

As for the collection of hosts relaying mail over UUCP, we always referred
to it as "the UUCP network." It was all very adhoc, with many private
links. When we talked about "the UUCP network" it generally referred to
the hosts exchanging mail as defined in the monthly UUCP maps, published
in the comp.mail.maps newsgroup.

--lyndon

[Johan Helsingius]

_______
internet-history mailing list
internet-***@postel.org
http://mailman.postel.org/mailman/listinfo/internet-history
Contact list-***@postel.org for assistance.

[Louis Mamakos]

On 6 Jul 2018, at 16:40, Johan Helsingius wrote:

> On 06/07/18 20:53, Lyndon Nerenberg wrote:
>
>> No, 'uunet' was a commercial UUCP hub, started up by Rick Adams.
>>   It handled a great volume of mail and news traffic, as well as an
>> extensive public file archive.  It eventually morphed into Alternet,
>> one of the very first commercial Internet providers.
>
> Technically "Alternet" was a product/service provided by UUNET,
> the company established by Rick (and funding by Usenix) when
> he transferred the USENET node from seismo (Center for Seismic
> Studies) where he was the admin.

More accurately, "UUNET Communication Services" was the USENIX
non-profit entity. This spawned UUNET Technologies, Inc, which
was a for-profit entity, which raised private equity and eventually
did an IPO. ALTERNET was the product name of the Internet
service, presumably selected to differentiate that service from
the dial-up UUCP/USENET offering. The ALTERNET / UUNET backbone
was at the time one of the largest global tier-1 ISPs.

UUNET Technologies was acquired by MFS, and later acquired by
WorldCom and some might recall how that later worked out.

For some period of time, UUNET Technologies paid USENIX a royalty
for the UUNET Communications UUCP/USENET business that was folded into
UUNET Technologies.

louie
(formerly, ***@uu.net)

[Brian E Carpenter]

On 07/07/2018 06:53, Lyndon Nerenberg wrote:
>> That's a good point.  I too remember USENET as referring more to the
>> newsgroups than to uucp transport.
>
> Yes, Usenet was the collection of newsgroups, originally shipped over
> UUCP.
>
>> Do I recall correctly that "uunet" was the term used to denote the
>> network of machines that could be reached by bang path addressing?
>
> No, 'uunet' was a commercial UUCP hub, started up by Rick Adams. It
> handled a great volume of mail and news traffic, as well as an extensive
> public file archive. It eventually morphed into Alternet, one of the
> very first commercial Internet providers.
>
> As for the collection of hosts relaying mail over UUCP, we always referred
> to it as "the UUCP network." It was all very adhoc, with many private
> links. When we talked about "the UUCP network" it generally referred to
> the hosts exchanging mail as defined in the monthly UUCP maps, published
> in the comp.mail.maps newsgroup.

EUNET in Europe was the UUCP setup, starting formally in 1982, which in
due time migrated from UUCP-over-dialup to UUCP-over-TCP/IP. Its HQ was
in Amsterdam and was a large part of why RIPE also grew up in Amsterdam.
The best source, I think, is a section in:
A History of International Research Networking: the People who Made it
Happen, Howard Davies & Beatrice Bressan (ed), ISBN: 978-3-527-32710-2,
Wiley, 2010.

Personally I was using UUCP email between Switzerland and New Zealand
in ~1984, well before SMTP was available on that path. We saw either
MCVAX (in Amsterdam) or UNIDO (in Dortmund) in most bang paths
from CERNVAX.

Brian

[Dr Eberhard W Lisse]

And,

UNDO and MCVAX were both on EARN/BITNET, though I don’t remember if they routed any UUCP traffic over BITNET.

EL

Sent from Dr Lisse’s iPad mini 4
On 6 Jul 2018, 23:24 +0100, Brian E Carpenter <***@gmail.com>, wrote:
> On 07/07/2018 06:53, Lyndon Nerenberg wrote:
> > > That's a good point.  I too remember USENET as referring more to the
> > > newsgroups than to uucp transport.
> >
> > Yes, Usenet was the collection of newsgroups, originally shipped over
> > UUCP.
> >
> > > Do I recall correctly that "uunet" was the term used to denote the
> > > network of machines that could be reached by bang path addressing?
> >
> > No, 'uunet' was a commercial UUCP hub, started up by Rick Adams. It
> > handled a great volume of mail and news traffic, as well as an extensive
> > public file archive. It eventually morphed into Alternet, one of the
> > very first commercial Internet providers.
> >
> > As for the collection of hosts relaying mail over UUCP, we always referred
> > to it as "the UUCP network." It was all very adhoc, with many private
> > links. When we talked about "the UUCP network" it generally referred to
> > the hosts exchanging mail as defined in the monthly UUCP maps, published
> > in the comp.mail.maps newsgroup.
>
> EUNET in Europe was the UUCP setup, starting formally in 1982, which in
> due time migrated from UUCP-over-dialup to UUCP-over-TCP/IP. Its HQ was
> in Amsterdam and was a large part of why RIPE also grew up in Amsterdam.
> The best source, I think, is a section in:
> A History of International Research Networking: the People who Made it
> Happen, Howard Davies & Beatrice Bressan (ed), ISBN: 978-3-527-32710-2,
> Wiley, 2010.
>
> Personally I was using UUCP email between Switzerland and New Zealand
> in ~1984, well before SMTP was available on that path. We saw either
> MCVAX (in Amsterdam) or UNIDO (in Dortmund) in most bang paths
> from CERNVAX.
>
> Brian
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

[Jaap Akkerhuis]

Dr Eberhard W Lisse writes:

> And,
>
> UNDO and MCVAX were both on EARN/BITNET, though I don't remember
> if they routed any UUCP traffic over BITNET.

The mcvax started out as the central eunet uucp node. When SARA
(the center which gave the Amsterdam Universities & MC (now CWI) the
bulk computer services) got connected to BITNET we managed to get
a leaf node connection (via a 9600 B/s serial line but that is
another story). So this "speed" is why we never routed uucp traffic
over BITNET. Too slow. It was probably not allowed to do anyway.
As far as I remember, it ran a unix version (Purdue?) of the software.
We already had fixed lines to various places which where.

As an experiment, Piet once had an uumap type routing database for
an the next day not only the serial line but also the poor VAX 750
was saturated. The load was 150 or so, so the experiment was ended
quickly. (After we mentioned to login, that was a challenge on
itself).

jaap (once ..|mcvax!jaap)

[Johan Helsingius]

On 07-07-18 08:29, Dr Eberhard W Lisse wrote:

> UNDO and MCVAX were both on EARN/BITNET, though I don’t remember if they
> routed any UUCP traffic over BITNET.

Mcvax definitely gatewayed stuff over to BITNET, but I don't
think it was ever used for transit/transport.

Julf

[Lyndon Nerenberg]

> UNDO and MCVAX were both on EARN/BITNET, though I don’t remember if they routed any UUCP traffic over
> BITNET.

I can't speak to that specific case, but I would guess this would be a
decision of last resort. BITNET spoke EBCDIC, and UNIX email often didn't
survive the ASCII->EBCDIC->ASCII translations unscathed.

I ran a BITNET node (AUCS) using UREP running on a Sun 3/160. Dealing
with character set translation was always a pain in the butt.

--lyndon

[Steven G. Huter]

On Fri, 6 Jul 2018, Lyndon Nerenberg wrote:

> As for the collection of hosts relaying mail over UUCP, we always referred to
> it as "the UUCP network." It was all very adhoc, with many private links.
> When we talked about "the UUCP network" it generally referred to the hosts
> exchanging mail as defined in the monthly UUCP maps, published in the
> comp.mail.maps newsgroup.

for a snapshot from November 1989, nearly 30 years ago, as South Africa
started its transition from FidoNet to a UUCP mail feed to Rhodes
University...

Subject: two hours, uucp, uunet, modems, the universe, and everything

https://nsrc.org/db/lookup/report.php?id=898128827821:488960662&fromISO=ZA

this eventually resulted in Randy Bush and colleagues at Rhodes
University facilitating connections to neighboring countries via UUCP,
and was the beginning of the stepping stones that led to the first TCP/IP
links in a number of countries in Africa.

Steve Huter

[Johan Helsingius]

On 06/07/18 19:33, Olivier MJ Crépin-Leblond wrote:

> 2. My understanding of USENET is that it was not a network as such,
> but most people referred to it as the system of newsgroups using a
> clever hierarchy and peer to peer for dissemination (NNTP). Again,
> it was a store and forward network, this time using the UUCP suite
> and dial-up.

I still remember running the USENET node for Finland over an
X.25 connection to mcvax in Amsterdam. Charged by the packet.
On a Zilog Z8000 machine, running their version of System III UNIX.
Just like many of us, the Zilog engineers noticed the discrepancy
between the documentation and code in the termios TTY interface.
Unlike everybody else, they thought the documentation was right
and code wrong, and "fixed" the code. As a result, if there was any
delay in the packet stream, the tty driver would time out and return
"0 characters read" - the UNIX convention for End of File. So UUCP
would hang up. Great, if you just had transferred 45 K of a 48K binary
posting - and were paying per packet. :(

    Julf

[Miles Fidelman]

On 7/6/18 5:09 AM, Scott O. Bradner wrote:
>
>> On Jul 6, 2018, at 12:45 AM, Miles Fidelman <***@meetinghouse.net> wrote:
>>
> …
>
>> And, I believe that both USENET (or at least UUCP) and BITNET predated
>> the TCP/IP cutover by a couple of years. They were both decentralized
>> and "operated independently but cooperatively by more than one
>> organization."
> TCP/IP - developed in mid 1970s
> BITNET - founded 1981 by Ira Fuchs
> USENET - established in 1980 by Tom Truscott
> (UUCP was the file copy protocol not the system)
> TCP/IP cutover - 1983
>
>
To Jack's question:  Does that count as "before the Internet?"
(Inquiring minds want to know!)

--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra

[Richard Bennett]

Hmm
something like scalable addresses?

(sorry, you made it too easy)

> On Jul 5, 2018, at 4:25 PM, Vint Cerf <***@google.com> wrote:
>
> well, you might want to think about scale for a moment.
>
> v
>
>
> On Thu, Jul 5, 2018 at 5:50 PM, Richard Bennett <***@bennett.com <mailto:***@bennett.com>> wrote:
> So there’s really nothing novel about the Internet. Digital communication has been done with end-to-end control of multiplexed packets since the early 20th century, and the phone network was just a diversion.
>
> Cool.
>
> RB
>
>
>> On Jul 5, 2018, at 2:59 PM, Brian E Carpenter <***@gmail.com <mailto:***@gmail.com>> wrote:
>>
>> On 06/07/2018 03:34, Jack Haverty wrote:
>> ...
>>> I was in Paris recently and spent several hours at the Musee des Arts et
>>> Metiers, essentially a museum of technology. One section is devoted to
>>> "Communications". I noticed one display cabinet containing a machine
>>> that was somehow used to "allow several telegraph operators to share the
>>> same wire" - so I guess Multiplexing has been around since the 19th century.
>>
>> It was perceived as a requirement very early. Wikipedia dates it to the 1870s.
>> But I think the first really successful version was the "Murray Multiplex" in
>> about 1909. My colleague Bob Doran has studied Donald Murray's works at some length:
>> https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/Murray/MurraySpielLR.pdf <https://www.cs.auckland.ac.nz/historydisplays/FifthFloor/Murray/MurraySpielLR.pdf>
>>
>> Brian
>>
>> _______
>> internet-history mailing list
>> internet-***@postel.org <mailto:internet-***@postel.org>
>> http://mailman.postel.org/mailman/listinfo/internet-history <http://mailman.postel.org/mailman/listinfo/internet-history>
>> Contact list-***@postel.org <mailto:list-***@postel.org> for assistance.
>
> —
> Richard Bennett
> High Tech Forum <http://hightechforum.org/> Founder
> Ethernet & Wi-Fi standards co-creator
>
> Internet Policy Consultant
>
>
> _______
> internet-history mailing list
> internet-***@postel.org <mailto:internet-***@postel.org>
> http://mailman.postel.org/mailman/listinfo/internet-history <http://mailman.postel.org/mailman/listinfo/internet-history>
> Contact list-***@postel.org <mailto:list-***@postel.org> for assistance.
>
>
>
>
> --
> New postal address:
> Google
> 1875 Explorer Street, 10th Floor
> Reston, VA 20190

—
Richard Bennett
High Tech Forum <http://hightechforum.org/> Founder
Ethernet & Wi-Fi standards co-creator

Internet Policy Consultant

[Dave Walden]

Should I be able to open postel.org in my browser?  I get a message
saying not in the DNS.  I am trying to find where someone goes to get
added to this list so I can tell the person that information.

[Dave Crocker]

-------- Forwarded Message --------
Subject: Re: [ih] Fwd: Re: Early Internet history
Date: Thu, 5 Jul 2018 10:21:19 -0400
From: Tom Van Vleck <***@multicians.org>
To: Dave Crocker <***@bbiw.net>

1965 is what I believe.

MAIL could not be made available to CTSS users until
after the "new CTSS file system" was installed on August 9,1965.
I am not sure of the exact date in 1965 that it was installed.. probably
soon after.

I may be able to find a MIT CC memo that has it.

regards, tom



> On Jul 5, 2018, at 9:54 AM, Dave Crocker <***@bbiw.net> wrote:
>
>
> Tom,
>
> I'm used to hearing 1965, but figure I should ask you for the actual date before responding to the list.
>
> d/
>
> -------- Forwarded Message --------
> Subject: Re: [ih] Fwd: Re: Early Internet history
> Date: Thu, 5 Jul 2018 08:11:29 -0400
> From: Vint Cerf <***@google.com>
> To: David Crocker <***@bbiw.net>
> CC: internet-***@postel.org <internet-***@postel.org>
>
>
>
> what was the time frame for CSS Mail?
>
> v




--
Dave Crocker
Brandenburg InternetWorking
bbiw.net

[Noel Chiappa]

> From: Craig Partridge

> Earlier -- 1790s (the so-called Napoleonic telegraph ... It interleaved
> messages from multiple spots, I believe.

If that's so, Kleinrock _did_ invent packet switching (since his network
"interleaved messages from multiple spots")... (The part he missed was
breaking a message up into chunks, and sending the chunks separately.)

Noel

[Vint Cerf]

Kleinrock's analysis was for message switching but the mathematics of
message switching and packet switching are essentially comparable
especially when you consider variable length messages.

v


On Thu, Jul 5, 2018 at 7:05 PM, Noel Chiappa <***@mercury.lcs.mit.edu>
wrote:

> > From: Craig Partridge
>
> > Earlier -- 1790s (the so-called Napoleonic telegraph ... It
> interleaved
> > messages from multiple spots, I believe.
>
> If that's so, Kleinrock _did_ invent packet switching (since his network
> "interleaved messages from multiple spots")... (The part he missed was
> breaking a message up into chunks, and sending the chunks separately.)
>
> Noel
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>



--
New postal address:
Google
1875 Explorer Street, 10th Floor
Reston, VA 20190

[Dave Crocker]

> Kleinrock's analysis was for message switching but the mathematics of
> message switching and packet switching are essentially comparable
> especially when you consider variable length messages.


Howdy.

A question to the group...

That distinction coincidentally surfaced in a discussion a couple of
months ago, after some decades of my not hearing it.

I know what it meant on the Arpanet. And I know what wikipedia and some
other entries say about it. But while the ability to handle smaller
chunks independently -- and even in an overlapping manner -- encourages
some useful performance improvements, I find myself generally thinking
of them as the same category of communications technology.

Namely: Discrete segments of data being handled through a network.
Certainly for some form of multiplexing and possibly with dynamic
routing. (These days, we'd take stat mux and dynamic routing as
inherent, but my recollection is that 45 years ago, those were
variations being played with.)

I'm not looking to re-start the religious wars on the distinction but am
curious whether, from the perspective of those 45 years and global
scaling, it is fair to have most discussions -- I emphasize most, not
all -- treat them as the same construct?

If not, why not?


d/
--
Dave Crocker
Brandenburg InternetWorking
bbiw.net

[Vint Cerf]

The basic methods of queueing theory apply to both message and packet
switching.
In the packet case, segmentation allows the message to be in flight
concurrently over multiple hops so that reduces the total delay since there
is, as you say, overlapping transmission along multiple hops of the path.Of
course a one hop path produces no advantage since all the segments have to
transit the one hop. TCP segmented the entire transmission partly based on
flow control signals. For a time, TCP Segments might have been broken into
smaller pieces in, e.g., the ARPANET. Later transit over Frame Relay and
ATM also potentially broke things up and permitted similar overlaps. I was
only making the point that the same mathematical models work for both cases.

v


On Sat, Jul 7, 2018 at 7:15 PM, Dave Crocker <***@dcrocker.net> wrote:

>
> > Kleinrock's analysis was for message switching but the mathematics of
> > message switching and packet switching are essentially comparable
> > especially when you consider variable length messages.
>
>
> Howdy.
>
> A question to the group...
>
> That distinction coincidentally surfaced in a discussion a couple of
> months ago, after some decades of my not hearing it.
>
> I know what it meant on the Arpanet. And I know what wikipedia and some
> other entries say about it. But while the ability to handle smaller
> chunks independently -- and even in an overlapping manner -- encourages
> some useful performance improvements, I find myself generally thinking
> of them as the same category of communications technology.
>
> Namely: Discrete segments of data being handled through a network.
> Certainly for some form of multiplexing and possibly with dynamic
> routing. (These days, we'd take stat mux and dynamic routing as
> inherent, but my recollection is that 45 years ago, those were
> variations being played with.)
>
> I'm not looking to re-start the religious wars on the distinction but am
> curious whether, from the perspective of those 45 years and global
> scaling, it is fair to have most discussions -- I emphasize most, not
> all -- treat them as the same construct?
>
> If not, why not?
>
>
> d/
> --
> Dave Crocker
> Brandenburg InternetWorking
> bbiw.net
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>



--
New postal address:
Google
1875 Explorer Street, 10th Floor
Reston, VA 20190

[Dave Crocker]

On 7/7/2018 5:05 PM, Vint Cerf wrote:
> I was only making the point that the same mathematical models work for
> both cases.

Ack. Tnx. And it added to my long-time assumption that from a macro-view
they can be seen as similar, although the engineering differences can
make a significant operational different. (theory, practice, ...)

d/

--
Dave Crocker
Brandenburg InternetWorking
bbiw.net

[Jack Haverty]

As The Judge recently pointed out, it's impossible to define colloquial
technical terms like "email". I suspect "message" and "packet" are
similarly imprecise.

Personally, I've always thought that the distinction is that a "message"
is a self-contained and complete unit of information, that makes sense
to both the endpoint sender and receiver. You send a whole "message" to
some destination, and the destination can only use it after the entire
message is received.

A "packet" is something that some communications mechanism uses to
actually transport data, driven by pragmatic and engineering concerns.
E.g., packet sizes might be set based on expected error rates due to noise.

So, messages get chopped up into packets; or maybe even several messages
get packaged into one packet. It's up to the communications machinery
(or its designers) to decide, based on local conditions.

Messages get sent between users, who decide what constitutes a complete
unit of useful information.

Of course, one player's packets might also actually be another player's
messages. SMTP messages get carved up into IP datagrams which were
carved up into pieces that were sent between endpoint IMPs, which were
carved up into other pieces that actually traversed wires between IMPs.
These chunks were packets or messages depending on your viewpoint.

/Jack



On 07/07/2018 04:15 PM, Dave Crocker wrote:
>
>> Kleinrock's analysis was for message switching but the mathematics of
>> message switching and packet switching are essentially comparable
>> especially when you consider variable length messages.
>
>
> Howdy.
>
> A question to the group...
>
> That distinction coincidentally surfaced in a discussion a couple of
> months ago, after some decades of my not hearing it.
>
> I know what it meant on the Arpanet. And I know what wikipedia and some
> other entries say about it. But while the ability to handle smaller
> chunks independently -- and even in an overlapping manner -- encourages
> some useful performance improvements, I find myself generally thinking
> of them as the same category of communications technology.
>
> Namely: Discrete segments of data being handled through a network.
> Certainly for some form of multiplexing and possibly with dynamic
> routing. (These days, we'd take stat mux and dynamic routing as
> inherent, but my recollection is that 45 years ago, those were
> variations being played with.)
>
> I'm not looking to re-start the religious wars on the distinction but am
> curious whether, from the perspective of those 45 years and global
> scaling, it is fair to have most discussions -- I emphasize most, not
> all -- treat them as the same construct?
>
> If not, why not?
>
>
> d/
>

[John Day]

In the ARPANET, a message could be up to 8 packets.

> On Jul 7, 2018, at 20:48, Jack Haverty <***@3kitty.org> wrote:
>
> As The Judge recently pointed out, it's impossible to define colloquial
> technical terms like "email". I suspect "message" and "packet" are
> similarly imprecise.
>
> Personally, I've always thought that the distinction is that a "message"
> is a self-contained and complete unit of information, that makes sense
> to both the endpoint sender and receiver. You send a whole "message" to
> some destination, and the destination can only use it after the entire
> message is received.
>
> A "packet" is something that some communications mechanism uses to
> actually transport data, driven by pragmatic and engineering concerns.
> E.g., packet sizes might be set based on expected error rates due to noise.
>
> So, messages get chopped up into packets; or maybe even several messages
> get packaged into one packet. It's up to the communications machinery
> (or its designers) to decide, based on local conditions.
>
> Messages get sent between users, who decide what constitutes a complete
> unit of useful information.
>
> Of course, one player's packets might also actually be another player's
> messages. SMTP messages get carved up into IP datagrams which were
> carved up into pieces that were sent between endpoint IMPs, which were
> carved up into other pieces that actually traversed wires between IMPs.
> These chunks were packets or messages depending on your viewpoint.
>
> /Jack
>
>
>
> On 07/07/2018 04:15 PM, Dave Crocker wrote:
>>
>>> Kleinrock's analysis was for message switching but the mathematics of
>>> message switching and packet switching are essentially comparable
>>> especially when you consider variable length messages.
>>
>>
>> Howdy.
>>
>> A question to the group...
>>
>> That distinction coincidentally surfaced in a discussion a couple of
>> months ago, after some decades of my not hearing it.
>>
>> I know what it meant on the Arpanet. And I know what wikipedia and some
>> other entries say about it. But while the ability to handle smaller
>> chunks independently -- and even in an overlapping manner -- encourages
>> some useful performance improvements, I find myself generally thinking
>> of them as the same category of communications technology.
>>
>> Namely: Discrete segments of data being handled through a network.
>> Certainly for some form of multiplexing and possibly with dynamic
>> routing. (These days, we'd take stat mux and dynamic routing as
>> inherent, but my recollection is that 45 years ago, those were
>> variations being played with.)
>>
>> I'm not looking to re-start the religious wars on the distinction but am
>> curious whether, from the perspective of those 45 years and global
>> scaling, it is fair to have most discussions -- I emphasize most, not
>> all -- treat them as the same construct?
>>
>> If not, why not?
>>
>>
>> d/
>>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

[Lawrence Stewart]

Both communications circuits and multiplexing were developed by telegraph systems, but what of routing?

In 1942, George O. Smith wrote a science fiction story called QRM Interplanetary, one of a collection of stories called Venus Equilateral about communications engineers on a relay station. I rather like the book, but YMMV.

In the prolog to the story, a message from Venus to Detroit is followed along:

"The punched tape from Operator No. 7's machine slid along the line until it entered a coupling machine.

The coupling machine worked furiously. It accepted the tapes from seventy operators as fast as they could set them. It selected die messages as they entered me machine, placing a mechanical preference upon whichever message happened to be ahead of the others on the moving tapes. The master tape moved continuously at eleven thousand words per minute, taking teletype messages from everywhere in the Northern Hemisphere of Venus to Terra and Mars."

(The combined signal goes by radio to the Venus Equilateral relay station)

“The signal was amplified and demodulated. It went into a decoupler machine where the messages were sorted mechanically and sent, each to the proper channel, into other coupler machines. Beams from Venus Equilateral were directed at Mars and at Terra.”

Excerpt From: George O. Smith. “Venus Equilateral.

Now this does not appear to be packet switching, but rather message switching, although perhaps the telegrams were limited in length.

I don’t know how systems like TELEX worked, for telegram routing. The Wikipedia article suggests it was circuit switched, over POTS, but it is not much of an article. It does claim TELEX as the forerunner of email :)

[John Levine]

In article <11fd50bf-d495-c4a5-ce77-***@meetinghouse.net>,
Miles Fidelman <***@meetinghouse.net> wrote:
>Kind of sounds like the international postal system.  Or shipping
>packages internationally. 

About 20 years ago, in Internet for Dummies, I analogized the
operation of the Internet to paper mail. The pre-TCP protocols were
sort of like registered mail, where each package has great value* and
is carefully logged in and out every time it is sorted or transported
to be sure it doesn't get lost.

TCP is more like certified mail, where the package itself is of no
value, only its message ("usually a letter from your insurance company
saying your policy has been cancelled.") Certified mail is only
logged when mailed and delivered. If it isn't delivered after a
while, you just send another copy.

I further tortured the analogy by saying you were mailing a copy of
the ten-pound manuscript of your novel, but the regulations limit each
package to one pound so you divide it into pieces and mark each one
PART 1, PART 2, and so forth. The packages arrive in whatever order
the post office delivers them, and the recipient puts them back in
order.

R's,
John

* - in the 1800s the government shipped gold bars by registered mail

[Jack Haverty]

While we were debating and evolving TCP2 into TCP4, circa 1979-80 or so,
I remember thinking about my four years of Latin classes in high school,
where I coincidentally learned quite a lot of Roman history and what
life was like in the Roman Empire.

It turned out that the Roman empire had a problem of scale. As it got
larger, it became increasingly difficult for outlying parts of the
Empire to reliably and securely communicate with Rome. Caesar wanted
quick, reliable, and secure communications with his Armies and
Governors. And he couldn't find a suitable ISP...

The commanders in the field used a variety of techniques to communicate.
Messages were written (Packetized?) on some convenient media by
Scribes. They were handed to Couriers. To improve reliability, several
Couriers would be sent with the same message, by diverse means and
routes (one on foot, one on a ship, etc.). Local knowledge (soldier in
charge of the local garrison) could tell the courier where best to head
next (Routing), based on current conditions. If no reply was received
in a reasonable time, more Couriers could be sent. More resources could
be allocated, i.e., more Couriers and Scribes (Multi-Channel?) for
really important messages. Slaves were cheaper than T1 circuits. For
really secret messages, the text could be split apart (fragmented?) and
pieces put into separate messages, so that the message could be
understood only when the pieces were re-united (reassembled?) at the
destination. Caesar had reliable, secure, and state-of-the-art fast
communications.

Several thousand years later, the Postal System adopted similar
techniques, adapted to the technology of the day. Circa 1980, so did
The Internet.

Caesar probably wasn't the first... Maybe the Assyrians?

/Jack


On 07/06/2018 09:49 AM, John Levine wrote:
> In article <11fd50bf-d495-c4a5-ce77-***@meetinghouse.net>,
> Miles Fidelman <***@meetinghouse.net> wrote:
>> Kind of sounds like the international postal system.� Or shipping
>> packages internationally.�
>
> About 20 years ago, in Internet for Dummies, I analogized the
> operation of the Internet to paper mail. The pre-TCP protocols were
> sort of like registered mail, where each package has great value* and
> is carefully logged in and out every time it is sorted or transported
> to be sure it doesn't get lost.
>
> TCP is more like certified mail, where the package itself is of no
> value, only its message ("usually a letter from your insurance company
> saying your policy has been cancelled.") Certified mail is only
> logged when mailed and delivered. If it isn't delivered after a
> while, you just send another copy.
>
> I further tortured the analogy by saying you were mailing a copy of
> the ten-pound manuscript of your novel, but the regulations limit each
> package to one pound so you divide it into pieces and mark each one
> PART 1, PART 2, and so forth. The packages arrive in whatever order
> the post office delivers them, and the recipient puts them back in
> order.
>
> R's,
> John
>
> * - in the 1800s the government shipped gold bars by registered mail
>
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>

[Richard Bennett]

Presumably paper replaced clay tablets because it was lighter and therefore faster to carry. Among other things.

> On Jul 6, 2018, at 12:00 PM, Jack Haverty <***@3kitty.org> wrote:
>
> While we were debating and evolving TCP2 into TCP4, circa 1979-80 or so,
> I remember thinking about my four years of Latin classes in high school,
> where I coincidentally learned quite a lot of Roman history and what
> life was like in the Roman Empire.
>
> It turned out that the Roman empire had a problem of scale. As it got
> larger, it became increasingly difficult for outlying parts of the
> Empire to reliably and securely communicate with Rome. Caesar wanted
> quick, reliable, and secure communications with his Armies and
> Governors. And he couldn't find a suitable ISP...
>
> The commanders in the field used a variety of techniques to communicate.
> Messages were written (Packetized?) on some convenient media by
> Scribes. They were handed to Couriers. To improve reliability, several
> Couriers would be sent with the same message, by diverse means and
> routes (one on foot, one on a ship, etc.). Local knowledge (soldier in
> charge of the local garrison) could tell the courier where best to head
> next (Routing), based on current conditions. If no reply was received
> in a reasonable time, more Couriers could be sent. More resources could
> be allocated, i.e., more Couriers and Scribes (Multi-Channel?) for
> really important messages. Slaves were cheaper than T1 circuits. For
> really secret messages, the text could be split apart (fragmented?) and
> pieces put into separate messages, so that the message could be
> understood only when the pieces were re-united (reassembled?) at the
> destination. Caesar had reliable, secure, and state-of-the-art fast
> communications.
>
> Several thousand years later, the Postal System adopted similar
> techniques, adapted to the technology of the day. Circa 1980, so did
> The Internet.
>
> Caesar probably wasn't the first... Maybe the Assyrians?
>
> /Jack
>
>
> On 07/06/2018 09:49 AM, John Levine wrote:
>> In article <11fd50bf-d495-c4a5-ce77-***@meetinghouse.net>,
>> Miles Fidelman <***@meetinghouse.net> wrote:
>>> Kind of sounds like the international postal system.ￜ Or shipping
>>> packages internationally.ￜ
>>
>> About 20 years ago, in Internet for Dummies, I analogized the
>> operation of the Internet to paper mail. The pre-TCP protocols were
>> sort of like registered mail, where each package has great value* and
>> is carefully logged in and out every time it is sorted or transported
>> to be sure it doesn't get lost.
>>
>> TCP is more like certified mail, where the package itself is of no
>> value, only its message ("usually a letter from your insurance company
>> saying your policy has been cancelled.") Certified mail is only
>> logged when mailed and delivered. If it isn't delivered after a
>> while, you just send another copy.
>>
>> I further tortured the analogy by saying you were mailing a copy of
>> the ten-pound manuscript of your novel, but the regulations limit each
>> package to one pound so you divide it into pieces and mark each one
>> PART 1, PART 2, and so forth. The packages arrive in whatever order
>> the post office delivers them, and the recipient puts them back in
>> order.
>>
>> R's,
>> John
>>
>> * - in the 1800s the government shipped gold bars by registered mail
>>
>>
>>
>> _______
>> internet-history mailing list
>> internet-***@postel.org
>> http://mailman.postel.org/mailman/listinfo/internet-history
>> Contact list-***@postel.org for assistance.
>>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

—
Richard Bennett
High Tech Forum <http://hightechforum.org/> Founder
Ethernet & Wi-Fi standards co-creator

Internet Policy Consultant

[Leo Vegoda]

On 6 July 2018 at 11:25, Richard Bennett <***@bennett.com> wrote:
> Presumably paper replaced clay tablets because it was lighter and therefore
> faster to carry. Among other things.

Cheap paper was an outcome of the development of steam powered
machinery. Until the 19th century, most paper was made from recycled
textile fibres and until weaving was industrialized, textiles were
expensive. Weavers in Britain would often pin pound notes to their hat
to demonstrate their wealth!

Papyrus is relatively light and was used from about 5,000(ish) years
ago but clay tablets continued to be used into the Iron Age. Such an
overlap suggests that there were considerations beyond the weight of
the messages to be sent.

[Brian E Carpenter]

On 07/07/2018 06:00, Jack Haverty wrote:

> Caesar had reliable, secure, and state-of-the-art fast
> communications.

But did he have avian carriers?

Brian

[John Day]

Indeed this is a good analogy of how the Internet worked. However, this approach of a connectionless datagram network layer with a reliable end-to-end transport originates in 1972 with the CYCLADES network. CYCLADES was the first to have the classic 5 layer model. The link layer was HDLC-like because of the error-prone lines of the time, over which ran a connectionless network layer, called Cigale. The primary purpose of the end-to-end Transport Protocol, called TS, was end-to-end reliability with flow control and recovery from loss due to congestion.

> On Jul 6, 2018, at 12:49, John Levine <***@iecc.com> wrote:
>
> In article <11fd50bf-d495-c4a5-ce77-***@meetinghouse.net>,
> Miles Fidelman <***@meetinghouse.net> wrote:
>> Kind of sounds like the international postal system.� Or shipping
>> packages internationally.�
>
> About 20 years ago, in Internet for Dummies, I analogized the
> operation of the Internet to paper mail. The pre-TCP protocols were
> sort of like registered mail, where each package has great value* and
> is carefully logged in and out every time it is sorted or transported
> to be sure it doesn't get lost.
>
> TCP is more like certified mail, where the package itself is of no
> value, only its message ("usually a letter from your insurance company
> saying your policy has been cancelled.") Certified mail is only
> logged when mailed and delivered. If it isn't delivered after a
> while, you just send another copy.
>
> I further tortured the analogy by saying you were mailing a copy of
> the ten-pound manuscript of your novel, but the regulations limit each
> package to one pound so you divide it into pieces and mark each one
> PART 1, PART 2, and so forth. The packages arrive in whatever order
> the post office delivers them, and the recipient puts them back in
> order.
>
> R's,
> John
>
> * - in the 1800s the government shipped gold bars by registered mail
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

[John R. Levine]

> Indeed this is a good analogy of how the Internet worked. However, this approach of a connectionless datagram network layer with a reliable end-to-end transport originates in 1972 with the CYCLADES network. CYCLADES was the first to have the classic 5 layer model. The link layer was HDLC-like because of the error-prone lines of the time, over which ran a connectionless network layer, called Cigale. The primary purpose of the end-to-end Transport Protocol, called TS, was end-to-end reliability with flow control and recovery from loss due to congestion.

I know about CYCLADES but this was already pushing how much nerdage I
could put in a For Dummies book.

R's,
John

>> On Jul 6, 2018, at 12:49, John Levine <***@iecc.com> wrote:
>>
>> In article <11fd50bf-d495-c4a5-ce77-***@meetinghouse.net>,
>> Miles Fidelman <***@meetinghouse.net> wrote:
>>> Kind of sounds like the international postal system.??? Or shipping
>>> packages internationally.???
>>
>> About 20 years ago, in Internet for Dummies, I analogized the
>> operation of the Internet to paper mail. The pre-TCP protocols were
>> sort of like registered mail, where each package has great value* and
>> is carefully logged in and out every time it is sorted or transported
>> to be sure it doesn't get lost.
>>
>> TCP is more like certified mail, where the package itself is of no
>> value, only its message ("usually a letter from your insurance company
>> saying your policy has been cancelled.") Certified mail is only
>> logged when mailed and delivered. If it isn't delivered after a
>> while, you just send another copy.
>>
>> I further tortured the analogy by saying you were mailing a copy of
>> the ten-pound manuscript of your novel, but the regulations limit each
>> package to one pound so you divide it into pieces and mark each one
>> PART 1, PART 2, and so forth. The packages arrive in whatever order
>> the post office delivers them, and the recipient puts them back in
>> order.
>>
>> R's,
>> John
>>
>> * - in the 1800s the government shipped gold bars by registered mail

[John Day]

;-) Understood. You were explaining the Internet, not the history.

Although their process to arrive at that solution is interesting.

> On Jul 6, 2018, at 15:47, John R. Levine <***@iecc.com> wrote:
>
>> Indeed this is a good analogy of how the Internet worked. However, this approach of a connectionless datagram network layer with a reliable end-to-end transport originates in 1972 with the CYCLADES network. CYCLADES was the first to have the classic 5 layer model. The link layer was HDLC-like because of the error-prone lines of the time, over which ran a connectionless network layer, called Cigale. The primary purpose of the end-to-end Transport Protocol, called TS, was end-to-end reliability with flow control and recovery from loss due to congestion.
>
> I know about CYCLADES but this was already pushing how much nerdage I could put in a For Dummies book.
>
> R's,
> John
>
>>> On Jul 6, 2018, at 12:49, John Levine <***@iecc.com> wrote:
>>>
>>> In article <11fd50bf-d495-c4a5-ce77-***@meetinghouse.net>,
>>> Miles Fidelman <***@meetinghouse.net> wrote:
>>>> Kind of sounds like the international postal system.??? Or shipping
>>>> packages internationally.???
>>>
>>> About 20 years ago, in Internet for Dummies, I analogized the
>>> operation of the Internet to paper mail. The pre-TCP protocols were
>>> sort of like registered mail, where each package has great value* and
>>> is carefully logged in and out every time it is sorted or transported
>>> to be sure it doesn't get lost.
>>>
>>> TCP is more like certified mail, where the package itself is of no
>>> value, only its message ("usually a letter from your insurance company
>>> saying your policy has been cancelled.") Certified mail is only
>>> logged when mailed and delivered. If it isn't delivered after a
>>> while, you just send another copy.
>>>
>>> I further tortured the analogy by saying you were mailing a copy of
>>> the ten-pound manuscript of your novel, but the regulations limit each
>>> package to one pound so you divide it into pieces and mark each one
>>> PART 1, PART 2, and so forth. The packages arrive in whatever order
>>> the post office delivers them, and the recipient puts them back in
>>> order.
>>>
>>> R's,
>>> John
>>>
>>> * - in the 1800s the government shipped gold bars by registered mail

[Wayne Hathaway]

A naive question: I don't recall anybody mentioning "The Victorian Internet" by Tom Standage.

I enjoyed it quite a bit many years ago, and it addresses many of the topics raised in this thread.

Thoughts?&nbsp; Opinions?


Wayne Hathaway
***@playaholic.com
&nbsp;

[Scott Bradner]

yup - a good (and informative) read

Scott

> On Jul 7, 2018, at 8:58 AM, Wayne Hathaway <***@playaholic.com> wrote:
>
> A naive question: I don't recall anybody mentioning "The Victorian Internet" by Tom Standage.
>
> I enjoyed it quite a bit many years ago, and it addresses many of the topics raised in this thread.
>
> Thoughts? Opinions?
>
>
> Wayne Hathaway
> ***@playaholic.com
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

[Vint Cerf]

i enjoyed this book, too - Tom is a good friend now. If you replace
"telegraph" with "internet" you see the same hype we had about Internet in
the 1995-2000 period showing up during the telegraph evolution - especially
when it went trans-Atlantic! and then wireless in 1901!

v


On Sat, Jul 7, 2018 at 8:58 AM, Wayne Hathaway <***@playaholic.com> wrote:

> A naive question: I don't recall anybody mentioning "The Victorian
> Internet" by Tom Standage.
>
> I enjoyed it quite a bit many years ago, and it addresses many of the
> topics raised in this thread.
>
> Thoughts? Opinions?
>
>
> Wayne Hathaway
> ***@playaholic.com
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>


--
New postal address:
Google
1875 Explorer Street, 10th Floor
Reston, VA 20190

[Miles Fidelman]

On 7/7/18 5:58 AM, Wayne Hathaway wrote:

> A naive question: I don't recall anybody mentioning "The Victorian
> Internet" by Tom Standage.
>
> I enjoyed it quite a bit many years ago, and it addresses many of the
> topics raised in this thread.
>
> Thoughts?  Opinions?
>
Yes.  A great book.  About telegraphy, all the way back to the days of
towers, mirrors, and flags, if I recall correctly.

Miles

--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra

[Joly MacFie]

*https://books.google.com/books?id=uFT4AgAAQBAJ
<https://books.google.com/books?id=uFT4AgAAQBAJ>*

> No eBook available. :(

But there is audio.

https://play.google.com/store/audiobooks/details?id=AQAAAABHUHVQIM



On Sat, Jul 7, 2018 at 1:14 PM, Miles Fidelman <***@meetinghouse.net>
wrote:

> On 7/7/18 5:58 AM, Wayne Hathaway wrote:
>
> > A naive question: I don't recall anybody mentioning "The Victorian
> > Internet" by Tom Standage.
> >
> > I enjoyed it quite a bit many years ago, and it addresses many of the
> > topics raised in this thread.
> >
> > Thoughts? Opinions?
> >
> Yes. A great book. About telegraphy, all the way back to the days of
> towers, mirrors, and flags, if I recall correctly.
>
> Miles
>
> --
> In theory, there is no difference between theory and practice.
> In practice, there is. .... Yogi Berra
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>



--
---------------------------------------------------------------
Joly MacFie 218 565 9365 Skype:punkcast
--------------------------------------------------------------
-

[John Levine]

In article <33D65BF2-E292-4526-9F62-***@serissa.com> you write:
>Now this does not appear to be packet switching, but rather message switching, although perhaps the telegrams
>were limited in length.
>
>I don’t know how systems like TELEX worked, for telegram routing. The Wikipedia article suggests it was
>circuit switched, over POTS, but it is not much of an article. It does claim TELEX as the forerunner of email

Telegrams and Telex were not the same thing. Telegrams were indeed
message switched, originally written by hand, later with torn paper
tape. There were young women on roller skates carrying messages from
one station to another at Western Union at 60 Hudson St.

Telex was circuit switched teletype. Telex machines had phone
numbers, so you dialed a call and then either typed your message or
more usually sent a prepared paper tape since you paid by the minute.
The connection was two way so for some kinds of transactions it was
common to negotiate by typing back and forth.

If you are interested in this stuff, the reference is "The Story of
Telecommunications", written by George Oslin in 1992. Oslin spent
much of his career at Western Union and was 93 when his book was
published, so he was largely writing about stuff he knew directly.
It's long out of print but it's easy to find used copies.

R's,
John

[John Levine]

In article <CAC20D2Mb0GL0mB9_H_-***@mail.gmail.com> you write:
>The UUCPnet grew incredibly fast because it was easy and reasonably cheap
>to attach ... but quickly the 'routing problem' emerged. In traditional
>UNIX style, UUCP had been design without worrying about some problems --
>UUCP was thinking small scale so how mail (packets or whatever got there)
>was not an issue. You did your own routing.

Right. A lot of this involved hiding the phone bills in company overhead budgets,
or carefully figuring out what was a local call. I have a beach house on Long Beach
Island in New Jersey, and careful reading of the first pages of the phone book revealed
that even though the Atlantic City airport was a long way away, quirks of geography
made it a local call, and the uucp node at the FAA tech center at the airport kindly
gave me a uucp feed for the summer.


> But with many, thousands of nodes, this was a huge problem.

>Netnews gets layered on top of UUCP and because of the growth, and the wild
>nature, this lack of support for routinr quickly becomes an issue.

Actually, for netnews it didn't matter since you flooded your news to
all your neighbors and it didn't matter what was multiple hops away.
The mapping project was for e-mail.

Everyone could use the route info to source route their uucp mail to
anyone else on the map. Considering what a kludge it was, it worked pretty well,


R's,
John

[Dr Eberhard W Lisse]

After Namibia’s independence calls to South Africa remained VERY cheap for a while and only gradually increased. So I dialled Grahamstown with UUPC at first and later managed to have the local University absorb calls to Pretoria from their Olivetti (!) box.

Then Linux appeared and Taylor/UUCP with long and sliding packets. We then found an early version of smail making gzipping and batching possible which together made this very efficient.

Once I figured out M4 I could set up sendmail up to do this in 9 lines :-)-O and when we had enough “paying” users to cross the leased lime threshold it took them a week or so to figure out that mails took minutes instead of hours :-)-O


I still like (Taylor) UUCP and it might actually be more efficient over satellite phones than doing this over TCP/IP, especially if one could hack up something dropping all attachments and rejecting messages over a certain size to block circumvention :-)-O

This might actually be helpful in disaster recovery :-)-O

el

Sent from Dr Lisse’s iPad mini 4
On 6 Jul 2018, 22:41 +0100, John Levine , wrote:
> In article <CAC20D2Mb0GL0mB9_H_-***@mail.gmail.com> you write:
> > The UUCPnet grew incredibly fast because it was easy and reasonably cheap
> > to attach ... but quickly the 'routing problem' emerged. In traditional
> > UNIX style, UUCP had been design without worrying about some problems --
> > UUCP was thinking small scale so how mail (packets or whatever got there)
> > was not an issue. You did your own routing.
>
> Right. A lot of this involved hiding the phone bills in company overhead budgets,
> or carefully figuring out what was a local call. I have a beach house on Long Beach
> Island in New Jersey, and careful reading of the first pages of the phone book revealed
> that even though the Atlantic City airport was a long way away, quirks of geography
> made it a local call, and the uucp node at the FAA tech center at the airport kindly
> gave me a uucp feed for the summer.
>
>
> > But with many, thousands of nodes, this was a huge problem.
>
> > Netnews gets layered on top of UUCP and because of the growth, and the wild
> > nature, this lack of support for routinr quickly becomes an issue.
>
> Actually, for netnews it didn't matter since you flooded your news to
> all your neighbors and it didn't matter what was multiple hops away.
> The mapping project was for e-mail.
>
> Everyone could use the route info to source route their uucp mail to
> anyone else on the map. Considering what a kludge it was, it worked pretty well,
>
>
> R's,
> John
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

[John Levine]

In article <CAM9VJk35M776iuUO3gjfQp_zYDf3U8XfBsyxoo86tExC-***@mail.gmail.com> you write:
>*https://books.google.com/books?id=uFT4AgAAQBAJ

>> No eBook available. :(

Lots of used copies available for about $10.

https://www.abebooks.com/servlet/SearchResults?sts=t&cm_sp=SearchF-_-home-_-Results&an=&tn=&kn=&isbn=9781620405925

Or you can take it out of the library. If you live in NYC, the library has
six copies.

If you want really early networking there's "The Early History of Data
Networks" by Holzmann and Perhson, IEEE CS press, which is mostly
about the era of optical telegraphs from the 1700s until the 1840s.
It's now ridiculously expensive, about $95 (my copy sure didn't cost
that much) but the library has it, too.

R's,
John

[Bill Ricker]

And if one wants to re-live / historically interpret the Victorian
Internet, there is an on-line community enabling analog current loop
telegraphy with tunneling over IP.
(IDK if it's UDP or TCP; i kinda hope it's multicast UDP at the
bottom. I'm not looking, don't need another rabbit hole!)

You can do good old American Morse wireline telegraphy over the internet today.
http://kob.sdf.org/morsekob/interface.htm … &
http://morsetelegraphclub.org/InternetTelegraphy.html …
(This original form of the code used the spacing of clicks subtly
differently than the International Morse dots and dashes most of us
learned as (pre)teens.)

(and of course some Hams are keeping International Morse on wireless
-- the Edwardian Internet ?? -- alive on the shortwaves. HI HI 73 OM )

Every decade or so, G.Marconi's daughter the Princess Elletra visits
Marconi National Historic Site on Cape Cod and/or related sites in
Canada and UK etc; she came again this last May:
http://www.arrl.org/news/radio-contact-planned-between-cape-cod-and-newfoundland-during-marconi-daughter-s-visit

[Lyndon Nerenberg]

> (and of course some Hams are keeping International Morse on wireless
> -- the Edwardian Internet ?? -- alive on the shortwaves. HI HI 73 OM )

... who are not nearly as rabid about this as the ex-coastal radio
operators! :-)

http://radiomarine.org/

--lyndon (VE0WX)

[John Levine]

In article <***@ary.qy> you write:
>Or you can take it out of the library. If you live in NYC, the library has
>six copies.

If you don't mind taking out a pirate scan, the Internet Archive has a scanned copy, too.

R's,
John

[Richard Bennett]

Please don’t promote piracy.

> On Jul 7, 2018, at 6:36 PM, John Levine <***@iecc.com> wrote:
>
> In article <***@ary.qy> you write:
>> Or you can take it out of the library. If you live in NYC, the library has
>> six copies.
>
> If you don't mind taking out a pirate scan, the Internet Archive has a scanned copy, too.
>
> R's,
> John
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

—
Richard Bennett
High Tech Forum <http://hightechforum.org/> Founder
Ethernet & Wi-Fi standards co-creator

Internet Policy Consultant

[Miles Fidelman]

But Richard,  "Information wants to be free!" :-)


On 7/7/18 6:06 PM, Richard Bennett wrote:
> Please don’t promote piracy.
>
>> On Jul 7, 2018, at 6:36 PM, John Levine <***@iecc.com
>> <mailto:***@iecc.com>> wrote:
>>
>> In article <***@ary.qy
>> <mailto:***@ary.qy>> you write:
>>> Or you can take it out of the library.  If you live in NYC, the
>>> library has
>>> six copies.
>>
>> If you don't mind taking out a pirate scan, the Internet Archive has
>> a scanned copy, too.
>>
>> R's,
>> John
>>
>> _______
>> internet-history mailing list
>> internet-***@postel.org <mailto:internet-***@postel.org>
>> http://mailman.postel.org/mailman/listinfo/internet-history
>> Contact list-***@postel.org for assistance.
>
> —
> Richard Bennett
> High Tech Forum <http://hightechforum.org> Founder
> Ethernet & Wi-Fi standards co-creator
>
> Internet Policy Consultant
>
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra

[Joly MacFie]

If you actually follow the link, this is a technologicly-enabled lending
library. One just "borrows", one doesn't pirate, which is something that,
as I understand it, involves plunder on the high seas.

joly



On Sat, Jul 7, 2018 at 11:31 PM, Miles Fidelman <***@meetinghouse.net>
wrote:

> But Richard, "Information wants to be free!" :-)
>
> On 7/7/18 6:06 PM, Richard Bennett wrote:
>
> Please don’t promote piracy.
>
> On Jul 7, 2018, at 6:36 PM, John Levine <***@iecc.com> wrote:
>
> In article <***@ary.qy> you write:
>
> Or you can take it out of the library. If you live in NYC, the library has
> six copies.
>
>
> If you don't mind taking out a pirate scan, the Internet Archive has a
> scanned copy, too.
>
> R's,
> John
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>
> —
> Richard Bennett
> High Tech Forum <http://hightechforum.org> Founder
> Ethernet & Wi-Fi standards co-creator
>
> Internet Policy Consultant
>
>
>
> _______
> internet-history mailing listinternet-***@postel.orghttp://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>
> --
> In theory, there is no difference between theory and practice.
> In practice, there is. .... Yogi Berra
>
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>


--
---------------------------------------------------------------
Joly MacFie 218 565 9365 Skype:punkcast
--------------------------------------------------------------
-

[Joly MacFie]

I see it. https://archive.org/details/victorianinterne00toms

Does not seem to be trivial getting it on my android phone to read on the
subway, but I'll try.

On Sat, Jul 7, 2018 at 8:36 PM, John Levine <***@iecc.com> wrote:

> In article <***@ary.qy> you write:
> >Or you can take it out of the library. If you live in NYC, the library
> has
> >six copies.
>
> If you don't mind taking out a pirate scan, the Internet Archive has a
> scanned copy, too.
>
> R's,
> John
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>



--
---------------------------------------------------------------
Joly MacFie 218 565 9365 Skype:punkcast
--------------------------------------------------------------
-

[John Levine]

In article <CAHxHggf7L0qH5o-B6-***@mail.gmail.com> you write:
>The basic methods of queueing theory apply to both message and packet
>switching.

Seems to me the key insight is that you can number the packets and the
recipient can reorder them so the network doesn't have to worry about
keeping them in order.

R's,
John

[Brian E Carpenter]

On 08/07/2018 12:39, John Levine wrote:
> In article <CAHxHggf7L0qH5o-B6-***@mail.gmail.com> you write:
>> The basic methods of queueing theory apply to both message and packet
>> switching.
>
> Seems to me the key insight is that you can number the packets and the
> recipient can reorder them so the network doesn't have to worry about
> keeping them in order.

Maybe, but ATM went for small fixed size packets to get to what they thought
was the sweet spot in queueing theory**, but they had to stay strictly in order.
And anybody who's written reassembly code knows what a bad thing out-of-order
variable-length packets can be.

** and the queueing theory is definitely simpler with fixed size packets,
because one of the distributions becomes a constant. In my misspent youth,
I slogged through both volumes of Kleinrock's book. I don't remember much,
but I do remember that M/D/1 is a lot simpler than M/M/1.

Brian

[Jorge Amodio]

Believe or not there are some instances of earlier use of the word
"Internet."

In my native country (Argentina) in the '50s it was a brand of panties ;-)
http://blog.internet-argentina.net/2010/05/se-me-cae-la-bombacha-de-la-emocion.html

That site is a crude personal attempt to put together some of the AR net
history, the Bombacha Internet picture is legit.

Cheers
Jorge

[Vint Cerf]

Jorge, that's hilarious!


reminds me of my surprise to find there was a popular
brand of bread, BIMBO, in Spain...

v


On Sat, Jul 7, 2018 at 9:41 PM, Jorge Amodio <***@gmail.com> wrote:

>
> Believe or not there are some instances of earlier use of the word
> "Internet."
>
> In my native country (Argentina) in the '50s it was a brand of panties ;-)
> http://blog.internet-argentina.net/2010/05/se-me-
> cae-la-bombacha-de-la-emocion.html
>
> That site is a crude personal attempt to put together some of the AR net
> history, the Bombacha Internet picture is legit.
>
> Cheers
> Jorge
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>


--
New postal address:
Google
1875 Explorer Street, 10th Floor
Reston, VA 20190

[Jorge Amodio]

Hi Vint,

yes it is, I was surprised many years ago when somebody pointed me to the
story.

Funny fact is that the person who created the brand in 1953, Ernesto
Rigoni, obviously didn't have anything to do with networking and didn't
speak a word in English.
This is a video that shows the ad in a popular magazine of 1956
https://youtu.be/wUirFOSxmWU

BTW, the BIMBO brand still exists, the company was founded and still have
headquarters in Mexico. Our kids love their toasted bread slices.
https://www.grupobimbo.com/en/ourhistory

Cheers
Jorge


On Sun, Jul 8, 2018 at 4:09 AM, Vint Cerf <***@google.com> wrote:

> Jorge, that's hilarious!
>
>
> reminds me of my surprise to find there was a popular
> brand of bread, BIMBO, in Spain...
>
> v
>
>
> On Sat, Jul 7, 2018 at 9:41 PM, Jorge Amodio <***@gmail.com> wrote:
>
>>
>> Believe or not there are some instances of earlier use of the word
>> "Internet."
>>
>> In my native country (Argentina) in the '50s it was a brand of panties ;-)
>> http://blog.internet-argentina.net/2010/05/se-me-cae-la-
>> bombacha-de-la-emocion.html
>>
>> That site is a crude personal attempt to put together some of the AR net
>> history, the Bombacha Internet picture is legit.
>>
>> Cheers
>> Jorge
>>
>> _______
>> internet-history mailing list
>> internet-***@postel.org
>> http://mailman.postel.org/mailman/listinfo/internet-history
>> Contact list-***@postel.org for assistance.
>>
>>
>
>
> --
> New postal address:
> Google
> 1875 Explorer Street, 10th Floor
> Reston, VA 20190
>
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.
>
>

[John Levine]

In article <CAHxHggc0BB6N28HvmF1UHkSfB9r4_+***@mail.gmail.com> you write:
>Jorge, that's hilarious!
>
>reminds me of my surprise to find there was a popular
>brand of bread, BIMBO, in Spain...

Bimbo is the largest commercial baker in the US, but outside of Latino
markets they use other brand names like Thomas, Arnold, Entenmanns,
and Sara Lee.

Speaking of Spain, UNIX makes fire extinguishers:

http://www.cs.grinnell.edu/drupal6/sites/default/files/museum/89_0.jpg

R's,
John

[Miles Fidelman]

And there's the classic story of the ad campaign for Drek soap - because
there wasn't a single Jew on Madison Ave. to tell folks that it meant
"Sh*t" in Yiddish.

Miles


On 7/8/18 10:33 AM, John Levine wrote:
> In article <CAHxHggc0BB6N28HvmF1UHkSfB9r4_+***@mail.gmail.com> you write:
>> Jorge, that's hilarious!
>>
>> reminds me of my surprise to find there was a popular
>> brand of bread, BIMBO, in Spain...
> Bimbo is the largest commercial baker in the US, but outside of Latino
> markets they use other brand names like Thomas, Arnold, Entenmanns,
> and Sara Lee.
>
> Speaking of Spain, UNIX makes fire extinguishers:
>
> http://www.cs.grinnell.edu/drupal6/sites/default/files/museum/89_0.jpg
>
> R's,
> John
> _______
> internet-history mailing list
> internet-***@postel.org
> http://mailman.postel.org/mailman/listinfo/internet-history
> Contact list-***@postel.org for assistance.

--
In theory, there is no difference between theory and practice.
In practice, there is. .... Yogi Berra

[Bill Ricker]

>>reminds me of my surprise to find there was a popular
>>brand of bread, BIMBO, in Spain...
>
> Bimbo is the largest commercial baker in the US, but outside of Latino
> markets they use other brand names like Thomas, Arnold, Entenmanns,
> and Sara Lee.

They're slowly raising their head above the parapet.
Bimbo USA HQ is outside Philadelphia, so they've bought shirt
sponsorship for the MLS Philadelphia Union.
https://en.wikipedia.org/wiki/Philadelphia_Union#Sponsorship

So we have an American pro-athlete team proudly wearing "Bimbo" at every game.

(And for what it's worth, "bimbo" is masculine in the original
Italian, and so originally was a synonym for gigolo.
There was an attempt to retrocoin "bimbette" in the 1980s but it
hasn't taken off.
See Google Ngram.
)

[John Levine]

In article <CAM9VJk34+EvmCjttCsirfEjoCq_QK+L=-sNjND=LG8+***@mail.gmail.com> you write:

>If you actually follow the link, this is a technologicly-enabled lending
>library. One just "borrows", one doesn't pirate, which is something that,
>as I understand it, involves plunder on the high seas.

The Archive ships containerloads of books to China to be scanned, then
lends out the scans. They assert that's equivalent to lending out the
physical books. They keep the books, which are shipped back to the US
in containers stacked up somewhere in San Francisco.

That is a highly dubious assertion. US copyright law in 17 USC 108(c)
allows libraries to lend copies if the original "is damaged,
deteriorating, lost, or stolen, or if the existing format in which the
work is stored has become obsolete" and they can't find a replacement.
The law defines obsolete to mean that one can't reasonably get the
device needed to read it. Eight-track tapes and microfiche, sure,
paper books, no.

R's,
John

PS: My union is looking for a test case.

>On Sat, Jul 7, 2018 at 11:31 PM, Miles Fidelman <***@meetinghouse.net>
>wrote:
>
>> But Richard, "Information wants to be free!" :-)
>>
>> On 7/7/18 6:06 PM, Richard Bennett wrote:
>>
>> Please don’t promote piracy.
>>
>> On Jul 7, 2018, at 6:36 PM, John Levine <***@iecc.com> wrote:
>>
>> In article <***@ary.qy> you write:
>>
>> Or you can take it out of the library. If you live in NYC, the library has
>> six copies.
>>
>> If you don't mind taking out a pirate scan, the Internet Archive has a
>> scanned copy, too.