2024-03-09 00:05:49 is it 64 bit? 2024-03-09 00:08:05 Yeah it's whatever platform you've got 2024-03-09 00:08:17 I think it supports both 32-bit and 64-bit 2024-03-09 00:09:27 Theoretically it supports anything GCC supports but it's probably only tested on x86, ARM, maybe others? 2024-03-09 00:24:36 sadly too involved to change from using brew, to port and making it work in the install script. 2024-03-09 00:24:43 will live with 0.7.3 :) 2024-03-09 00:30:52 Yeah it's a state 2024-03-09 01:23:35 could someone make a container? 2024-03-09 01:24:12 Yes although I'd guess it would be quite big 2024-03-09 01:48:13 Ton of books on vacuum tube circuits here: 2024-03-09 01:48:15 http://www.tubebooks.org/technical_books_online.htm 2024-03-09 03:43:13 KipIngram, thanks. The site has been 'scraped' with a pdf download script, nice old stuff. 2024-03-09 19:48:07 I love our tendency here to "curate" this sort of thing. I think having them around is important. 2024-03-09 19:48:24 "Curators of simplicity." :-) 2024-03-09 19:56:30 simplicity, or robustness... https://www.cs.unm.edu/~ackley/be-201301131528.pdf 2024-03-09 20:58:17 Cool - I just got an Amazon delivery; little 10-channel 12-bit analog-to-digital converter. Not high performance; it was only $15 or so. It delivers all ten voltages about every half-second, just by plugging it in and "cat /dev/ttyUSB0" 2024-03-09 20:58:46 I've been tinkering with battery chemistries and want to be able to log charge and discharge profiles. 2024-03-09 21:00:13 I've got some opamps in the garage I can use to make a current-to-voltage converter. 2024-03-09 21:00:52 And I can pipe this into a Python program to make myself something with a more sensible output format. 2024-03-09 21:03:37 That would tie up my computer for the duration of experiments, though. What I'd really like to do is make a more proper battery test rig controlled with one of those MAX32655 boards. Program it to run a cell through whatever process I want, all unattended. 2024-03-09 21:04:07 https://www.amazon.com/dp/B07V2NNN4P?psc=1&ref=ppx_yo2ov_dt_b_product_details 2024-03-09 21:05:24 That guy has its own A/D, though. 2024-03-09 22:46:58 This is interesting: 2024-03-09 22:47:01 https://www.cryptomuseum.com/crypto/enigma/wiring.htm 2024-03-09 22:47:14 Actually historical wiring information on the various Enigma machines. 2024-03-09 23:41:11 I'm thinking of writing an emulator for it - I'm interested in seeing how straightforward it would be to break it using the tools we have these days. 2024-03-09 23:42:11 enigma is cracked pretty quickly by modern computers, iirc 2024-03-09 23:42:45 I think I'd brute force the three wheels. There are about a million settings for that, so you'd get a million candidate texts, but they'd all just be simple substitution ciphers away from full clear text (the plug board). So things like frequency of double letters and so on would be intact - you could probably rank them on how likely there were to be the right one. 2024-03-09 23:42:52 or even older computers; the Brits kept on selling enigmas to various other nations after the war so they could break those comms 2024-03-09 23:42:53 Yeah, I think so too. 2024-03-09 23:43:09 Doing something a million times is no problem for modern computers. 2024-03-09 23:43:25 But the plug board introduces another factor of like 15 quntillion or something. 2024-03-09 23:43:36 But it doesn't change letter to letter. 2024-03-09 23:44:07 I think they relied on operators forgetting to change the plug board, or routine weather reports with the same text in them, etc 2024-03-09 23:44:48 Yeah. The big weakness, though, was that Enigma never left a letter unchanged. No letter was ever "itself," and they were able to use that. 2024-03-09 23:45:39 They'd guess at some plaintext and then slide it along under the ciphertext until they found one that was "possible." That would then give them a bunch of candidate mappings, and they'd use the kown ring structures to try to assess the settinsg. 2024-03-09 23:46:48 I figured it out earlier today - if you pick a 26-character substitution cipher "randomly," there's only about a 15% chance of getting one with zero self-mappings. 2024-03-09 23:47:05 And when you do get a self-mapping, you have to get another one - they come in pairs. 2024-03-09 23:49:31 like sith lords! 2024-03-09 23:50:05 :-) 2024-03-09 23:50:37 When I was a kid my local newspaper used to publish "Cryptoquotes." famous quotations subjected to a substitution cipher. My grandmom and I used to do them together. 2024-03-09 23:51:04 But these Enigma patterns are simpler, because if say 'r' goes to 'k,' then 'k' has to go to 'r' too, so you can decode with it. 2024-03-09 23:51:55 The rotor wheels aren't wired that way - they're wired with straignt random substitutions. But every character goes through them twice - right to left, then back left to right. 2024-03-09 23:52:20 There's a thing called a "reflector" over on the left that turns the signal arund and sends it back through. 2024-03-09 23:53:03 I figure if I'm going to do this I should make it accurate - that website I linked has everything needed. 2024-03-09 23:54:00 Also, the models tended to build on one another. Like a later one had four wheels selected from eight, instead of three selected from five. but the low numbered wheels were compatible across models. So if I do it right I'll be able to select model.