Wikipedia:Reference desk/Archives/Computing/2019 January 16

= January 16 =

What is an "electronic roulette wheel"?
When I were a lad, I read about A Million Random Digits with 100,000 Normal Deviates in some youth-oriented book, maybe the "Mathematics" issue of the Time–Life something something, and they said that the digits had been generated by "an electronic roulette wheel" or some such. Our article says it was "an electronic simulation of a roulette wheel attached to a computer".

I don't think I've ever seen any details on how this "electronic roulette wheel" worked. Did it have a moving wheel, or was there maybe some sawtooth signal that was somehow sampled at random times? --Trovatore (talk) 02:18, 16 January 2019 (UTC)
 * see https://www.rand.org/pubs/monograph_reports/MR1418/index2.html section "Production of the Random Digits". Some random timed pulses, about 100,000 per second were counted over a 1 second period. More on the topic here: https://www.rand.org/content/dam/rand/pubs/papers/2008/P113.pdf Graeme Bartlett (talk) 02:55, 16 January 2019 (UTC)
 * Thanks, Graeme. I did find that, after I asked the question.  It doesn't completely clear things up, though.  To quote,
 * A random frequency pulse source was gated by a constant frequency pulse, about once a second, providing on the average about 100,000 pulses in one second.
 * The text goes on to explain (essentially) that the count was taken mod 32, which makes it similar to a "roulette wheel" with 32 positions for the ball to fall into.
 * But what exactly is a "random frequency pulse source"? Maybe a radioactive source and a Geiger counter?  I haven't been able to find that out. --Trovatore (talk) 05:28, 16 January 2019 (UTC)


 * ERNIE was a major example of such, used to select the winners for the British Premium Bond lottery, a huge list of numbers needing a very rapid, yet demonstrably fair, generator. This used random signal noise in gas-filled neon tubes.
 * During the 1970s, much smaller roulette wheels and other random generators were popular projects in the hobbyist electronics magazines. The random number generators were often related to white noise generators, as used for electronic music, and used particularly noisy semiconductor junctions, usually in Zener diodes. Roulette wheels were less common and typically showed a rotating lamp or LED for the obvious roulette or 'wheel of fortune' effect. These used a sequential counter (the 4017 was a perennial favourite), fed by a pulse generator, often running at a frequency which fell slowly. Most importantly, the clock signal was gated by a circuit which stopped the pulses after a while to stop the count. This could be based on a relaxation oscillator, usually with a unijunction transistor (originally the neon tube again, when voltages had been higher), as it's easy to adapt that circuit to become unpredicatable. Andy Dingley (talk) 03:00, 16 January 2019 (UTC)
 * Thanks Andy, good info. However I'm asking specifically about the machine used to create those million digits. --Trovatore (talk) 05:28, 16 January 2019 (UTC)


 * I seem to recall someone at SRI telling me about a long history of adversarial research into the RAND random digits book. As I remember, there was a lot of study into just how good the randomness actually was in this specific table, and others like it.  (After all, this table was for many decades the standard source of randomness, and folks at SRI and elsewhere needed that kind of randomness to build secure cryptography algorithms, scientifically-valid computer physics simulations, and so on: they needed really real real randomness, so they tore that table  to shreds with every mathematical tool known to computer-scientists).
 * It was an endeavor probably best summarized by von Neumann, not to be confused with the other Neumann at SRI who specialized in computational and numerical analysis of random digits.
 * Regarding the machine used by RAND: as described in the book's introduction, they used a pulse counter (like the one in the April 1970 issue of HP's Measure Magazine - the HP5326A. That's a standard piece of lab kit from a very different era in digital electronics.  The RAND machine was another couple of decades older - buf its digital counter probably looked and worked very similarly.  A digital counter circuit is a pretty textbook design.  They probably fed it with a thermal noise source, or a photomultiplier, which is what they're calling the "pulse generator;" in both cases, it would probably be a very high gain vacuum-tube amplifier.  The true source of entropy would be either optical photon incidence, or thermal randomness; and on a very deep level, these are really the same thing occurring at different temperatures (frequencies).  With suitable signal-conditioners, that physical source of entropy was getting sampled by the digital system for conversion into random decimal digits.
 * RAND cites Juncosa (1953), a paper that is not easy to find on the internet; but there are better methods than search-engines! From DTIC, here is Gorn and Juncosa 1953, On the computational procedures..., straight from Aberdeen and quite literally citing the authors' original classified work conducted on no other computer than ENIAC.
 * Nimur (talk) 06:41, 16 January 2019 (UTC)


 * From the excellent discussion at [ http://bitbabbler.org/index.html ]:


 * "It turns out that it is actually quite challenging to build a genuinely good entropy source, and that unless you are extremely diligent, and skeptical, and fully committed to the task of proving that the RNG you have just made is bad, it's very easy to be deceived by the initial results of some simple statistical tests." and "To make it work we just needed to turn to the dark side and find a good way to eliminate as much of the signal in a circuit as possible while cultivating every available source of noise into a fully grown, unpredictable, string of zeros and ones … Quick, easy, and seductive, what could possibly go wrong? Of course it turns out that getting rid of any trace of a signal is at least as hard as getting rid of the noise in a conventional design." (Full disclosure: the bit babbler is the HRNG I use and recommend. I have no other connection to the product or the company)


 * I also highly recommend reading [ http://www.ciphergoth.org/crypto/unbiasing/ ]. --Guy Macon (talk) 08:06, 16 January 2019 (UTC)

TLS Certificate Chaining
I'm trying to install TLS certificates on a Linux server running RHEL 6.10. I have a root certificate, two intermediate certificates, and a host certificate. The instructions I'm following assume that I only have one intermediate certificate, not two. I know that I need to import the root cert to the Java keystore; the instructions then say to append the intermediate cert to the root cert and import that into the keystore as well. Since I have two intermediate certs, do I append both of them to the root cert? Or do I append intermediate 1 to the root and intermediate 2 to intermediate 1? Basically, I'm not sure how many certs I can chain together. OldTimeNESter (talk) 14:00, 16 January 2019 (UTC)

Wlan over Powerline in a small building
Is a wlan powerline feasible in a small building where all apartments have their own surge protectors? Would this (or something else) mess with the signal? --Doroletho (talk) 19:16, 16 January 2019 (UTC)
 * Probably it will not work. See this. Ruslik_ Zero 20:58, 16 January 2019 (UTC)

Do you have the option of removing an SP outlet & replacing with a standard outlet? (being sure to label it as such) — Preceding unsigned comment added by 50.39.223.26 (talk) 19:07, 18 January 2019 (UTC)


 * This is not just a SP outlet. Each apartment has its own home surge protector. Removing or circumventing this would imply rewiring the power lines. It would be easier to lay a wired network around the building.--Doroletho (talk) 00:12, 19 January 2019 (UTC)


 * Also the signal will be spread out and become weaker when it has to travel through more places, so it is unlikely to work. I also like to discourage this sort of use as it increases electromagnetic interference. Graeme Bartlett (talk) 03:38, 20 January 2019 (UTC)