2 ms·
The PC + freestanding desk phone was an interesting variant. It went through a few iterations, from sneakernet to dumb terminal to (locally) networked, to full
by dredmorbius 5y ago
The PC + freestanding desk phone was an interesting variant. It went through a few iterations, from sneakernet to dumb terminal to (locally) networked, to fully globally interconnected. I've lived through virtually all of that, and worked through much of it.
The O'Reilly book UNIX Power Tools gives a number of examples of (mostly terminal-based) tools that could be used for office productivity. I went through a number of grep-based (or similar) phonelist iterations myself.
There's a critical difference between the disconnected, and fully integrated system. The ability to look up a number on your desktop (either directly or as a terminal), and manually enter the number on a phone, is useful. It's not fully automated (so no call-centre automation), but it also means that the phone itself cannot exfiltrate data.
Note that phone numbers themselves are typically short and a relatively dense namespace. The US 10-digit number allows a theoretical maximum of 10 billion numbers ... actually somewhat fewer given technical limitations. That's only about 30x the current population of the US, meaning that if each person has an assigned number randomly dialing digits will connect 1 time in 30. That's a tractable search.
If the phone-number length were doubled, then wardialing would connect only about 1 time in 10 billion. Exhaustive search for active numbers would be infeasible.
But we'd all have to dial 20-digit numbers.
Or you could shorten the sequence by switching to lowercase alphabet (14 characters), alphanumeric (12), or mixed-case alphanumeric (11). If you used an 8,000 word dictionary (roughly the length of EFF's Diceware passphrase generator long wordlist), you'd need five words to provide a similar namespace.
That is: there's a trade-off between the conciseness of the namespace and the ability to abuse it for unsolicited calls.
(This is assuming that all phone spammers dial randomly, which of course isn't the case, but it's the start of one approach to addressing the problem.)