
The Special Operations Executive (SOE), a British WW2 group infiltrating Reich-dominated Europe, had during the War's early and middle years a continuing problem in certain parts of France. They would train new agents, drop them into French territory, note their contact with a local agent... and they were lost, presumed captured or killed. Two things needed to happen fast: first, a new network had to be built so fresh agents would not be compromised by the older, discovered network. And second, a code generation method must be implemented that did not give a field agent knowledge of how other field agents generated similar messages into encrypted form (knowledge that could be extracted by torture). The answer to the second problem was called a "one time pad", a method still in use today and which had life-saving results almost immediately in the Allied war effort.
-b) this is still relevant today for (especially amateur radio) morse code operators
https://www.amazon.com.au/Between-Silk-Cyanide-Codemakers-19...
He's the guy who came up with the idea of using silk for codes, as silk won't be felt during pat-downs (paper will be felt). There's lots in the book of trying to work with local business who were obviously not willing to part with their valuable silk.
The book goes much more into details of transmitting these codes, the hassle of bruteforcing when somebody misremembered their code or mistranposed (there was an army of women working 24/7 trying out all possible codes). Highly suggested if the OP piqued your interest!
Steve Bellovin wrote a review of "Between Silk and Cyanide" back on the old cryptography@c2.net mailing list that prompted me to buy the book: https://www.mail-archive.com/cryptography@c2.net/msg04572.ht... (The whole thread is good, BTW.)
The seriousness of the topic really left a mark with me (thinking about agents being tortured and killed as a result of poor opsec or cryptographic vulnerabilities). I'm glad that the work I do doesn't have those consequences.
