Here's Paul Baran's RAND-published September 1962 justification for distributed communications networks summed up in a single chart. It's also the same paper where the famous "centralized vs decentralized vs distributed" triptych of graphs comes from. You still see this exact diagram, uncited, in modern presentations on the decentralized/distributed web.
Paul Baran is not nearly as famous as Bob Kahn or Vint Cert or the like, but the guy literally invented packet switching in ~1964 as an outgrowth of this paper and similar work.
Here's a link to that paper, btw https://www.rand.org/pubs/research_memoranda/RM3763.html
Okay it never really occurred to me that in the 1960s a crytopgraphic key could be literally just a single punch card that you held onto. It presumably was under lock and key itself!
(Image from a related paper by Baran, "Security, Secrecy, and Tamper-Free Considerations". https://www.rand.org/pubs/research_memoranda/RM3765.html )
I also just learned that store-and-forward packet switching was also referred to as "hot potato routing", which, why is it not still called that??
Paul Baran: also apparently doing the xkcd thing 50 years before xkcd
Oh wow. So that really famous centralized vs decentralized vs distributed diagram that I quote above? Well, in one of the companion papers, Baran provides this iteration on it, which in my opinion is far superior and I'm going to start using in my presentations.
@darius thanks for sharing all this lovely archival material! You're consistently blowing my mind with your toots as of late.
That's a great base for illustrating other system characteristics like how authority, costs, and complexity are concentrated or distributed
Also hinted at... Networks aren't typically homogenous, by layer or node. You won't get consensus between an urban neighborhood and a rural village on network topology because their needs differ. You could, however, get consensus on hybrid approaches that allow different local topologies
Subtooting ActivityPub v SSB
I'm more for
2a) decentralized with federation
2b) decentralized and distributed
@darius my family always alerts me of things somewhere between B and C
@darius Because they're different. There are lots of ways to do packet switching, hot potato routing is just one of them.
@darius yubikey should make a wide and holed version of their dongles.
@darius that RAND logo is *everything*.
It's your friendly, deco military-industrial-complex-consulting-firm!!!
@slightlyoff their letterhead is equally exceptional, if I come across it in the next couple hours I'll send it
@slightlyoff weirdly I couldn't find the actual letterhead but these are neat
@darius aaah, thanks! i've wanted to look it up so many times and didn't do it yet so I'm very glad you are sharing this :)
*shows second picture to anyone*
"which of these is the better net"
"number #3 obviously the other two... aren't nets"
"so why does the interNET work like the first one HUH?"
@emsenn I actually prefer 2 to 3 because it maps more cleanly onto human social structures that make sense to me. 1 is totalitarian control, 3 is peak libertarian every person fends for themselves. (I understand the utility of talking about it like physical nets to some people.)
You can always tell whether someone is pitching decentralized or distributed topographies by how well-connected the nodes of the decentralized system are in their graphs 😂😂😂
uncited in this talk... but even centralized networks can easily acieve redundancy, today, and obviously that paper was before following folks. :D the way information is spread from node to node in distributed networks has other huge advantages, though.
[Katharina Nocun, 32c3]