[ih] Question on Flow Control

[Jack Haverty]

Whenever there's a question about who invented some protocol or 
algorithm in the computer world, IMHO it's useful to remember that 
humans have been communicating for a very long time -- well before 
computers existed.  My impression has always been that many "inventions" 
in the computer world were simply translations of older "protocols" and 
"algorithms" that humans used before computers were available to do the 
tedious work.

In 1960s high school, I got involved in ham radio, and in particular in 
"message traffic".  We passed messages (think "telegrams") from sender 
to receiver through as many intervening "hops" as needed to get from 
source to destination.  Routes were dynamically determined, depending on 
who showed up on the radio channel at the time.  Radio is a lossy 
environment, with fading, static crashes, and such events corrupting 
messages in transit between two operators.  There was a rather elaborate 
protocol for detecting such errors, retransmitting missing pieces, and 
making sure the message got through intact on each hop through the route 
from sender to recipient.

Ham operators didn't invent such protocols and algorithms.   Morse code 
had existed for a century or more, with professional operators 
performing similar protocols and algorithms.  Before electricity and 
telegraphy, messages were passed using balloons, lights, and signal fires.

Tom Standage's book "The Victorian Internet" tells a fascinating tale of 
how all that pre-computer technology was developed and used.  From 
Amazon:  "The Victorian Internet tells the colorful story of the 
telegraph's creation and remarkable impact, and of the visionaries, 
oddballs, and eccentrics who pioneered it, from the eighteenth-century 
French scientist Jean-Antoine Nollet to Samuel F. B. Morse and Thomas 
Edison. The electric telegraph nullified distance and shrank the world 
quicker and further than ever before or since, and its story mirrors and 
predicts that of the Internet in numerous ways."

Also in high school, I had a teacher who was interested in Greek and 
Roman history.   We learned a bit about communications as practiced 
several thousand years ago.  The Roman Empire covered much of Europe and 
extended into Asia and Africa.   Communications was a big problem, 
especially between the generals in the battlefield and the decision 
makers back in Rome.

They addressed that need with technology.  Couriers carried messages, 
and an extensive network of roads really did lead to Rome.  Today we use 
other terminology, such as "datagrams" and "packets".   We use fiber 
optics in place of paved roadways.

Messages were on scrolls.   I recall an Internet meeting sometime in the 
late 1970s when Vint explained the origin of the term "protocol".  At 
the beginning of a scroll was the "protokolon" (Greek terminology).   It 
contained a short description of what that scroll contained.  Today we'd 
call it a "header".

If a message was very important and critical, the sender would send 
multiple copies, and use alternate routes.  Slaves were cheap and in 
good supply.  One slave courier might be sent overland, and hopefully 
survive encounters with enemies, tribal warlords, and other hazards 
along the way.  Another might be sent by sea, in a trireme (boat with 
human-powered engine) that might survive encounters with pirates.

Latency was pretty high, measured in days, weeks, or even months, so 
acknowledgement and retransmission schemes were impractical.  Better to 
just send lots of datagrams.   Errr, -- I meant couriers.

Security was often a concern.  You didn't want the enemy to know your 
plans and weaknesses.  Scrolls could be ripped into pieces, and each 
piece sent by a separate courier, travelling over different routes.  
Putting the pieces together at the destination would reveal the 
message.   Today, we have "fragmentation" and "reassembly" in the 
Internet.   Our computers are the "scribes".  Our circuits are the 
"couriers".

At the Internet meetings in the late 1970s, I remember discussions where 
we talked about such "prior art", and used the lessons of history to 
make Internet design decisions.  The Internet was designed to serve as a 
communications infrastructure for the US/Nato military, which had the 
same needs as their predecessors in Roman times.   Some of the protocols 
and algorithms used today were invented long ago -- probably well before 
even Roman and Greek times.

/Jack Haverty



-------------- next part --------------
A non-text attachment was scrubbed...
Name: OpenPGP_signature.asc
Type: application/pgp-signature
Size: 665 bytes
Desc: OpenPGP digital signature
URL: <http://elists.isoc.org/pipermail/internet-history/attachments/20251229/44bb66bf/attachment.asc>