Early proto internet?

I was searching for earlier internet threads and found this comment:

packet switching
We could have built packet switching telegram networks in 1850, using relay logic computers we also could have built. They would have been slow, but much faster than human telegraph operators. A radiotelegraph system could have been built around 1905 with the invention of the triode. Coherers and crystals are too slow for a sensible radio telegraph network before then.

Can anybody explain if these things proposed are possible and what it would mean in layman`s terms? Any help is appreciated.

The orignal link:
https://www.alternatehistory.com/discussion/showthread.php?t=1103
 
I can't comment on the technical feasibility, but it's a very intriguing idea.

What would be the initial use for the telegraph system? Its initial customers or investors would determine a lot about how it ends up being used and viewed.

I can imagine that finance and banking would find a great number of uses though, and very quickly. Checking systems or consumer credit would be much more feasible if that information was readily available. A more efficient financial sector would have its own secondary effects as well, of course.
 
There wasn't any need for packetizing data. It was already translated into Baudot or Morse code.

Human operators hung on until the very end, but there was no technical reason why mechanical senders and transcribers couldn't have replaced them and speeded up throughput quite a bit.

Since telegraph was a point-to-point system it didn't need the packetization overhead that makes TCP/IP much less efficient by contrast.

Both Morse and modern TCP/IP packets are digital data; you could write Morse code on a piece of paper and have a pigeon carry it if you wanted. With TCP/IP, you could write several messages and send them with different pigeons, and the sender could tell if a pigeon never arrived and ask you to resend that particular message.

In fact, there's an RFC for transmision of IP datagrams over avian carriers:

http://www.faqs.org/rfcs/rfc1149.html

Transmission of IP packets has also been demonstrated using bongo drums. But there's still packet overhead compared to Morse or Baudot code...
 
Top