Talk:Middle-square method

136*136

C++ code
Anyone who knows enough C++ to understand that code probably doesn't need it to be here. In any case, I don't think it is Wikipedia's place to provide code samples -- it is an encyclopedia. It would be better if how it worked was actually explained clearly, rather than via code, in my opinion. --Fastfission 04:40, 5 May 2006 (UTC)

And I think the code provided is incorrect. It returns a uint32, x = (x>>32) | (x<<32); but x is a uint64. Sliding it right 16 would be a start toward getting it right Friendly Person (talk) 22:13, 4 December 2019 (UTC)

Reference
Where does 1946 come from? Some cite J. von Neumann: Various techniques used in connection with random digits. Monte Carlo methods. Nat. Bur. of Stand., Appl. Math. Ser. 12, 1951. Has anyone access to this article? —Preceding unsigned comment added by 139.18.75.122 (talk) 11:38, 1 June 2010 (UTC)

... I found a short hint on "about 1946" in D.E. Knuth's book "The Art of Computer Programming", vol. 2 (3. ed.) in section 3.1, page 3. A primary source would be necessary anyway.


 * I've seen the article; I don't think he mentions 1946 in it, but I don't have it in front of me, just notes. I've added a little bit. I don't know if 1946 was the first time he used it, it would be interesting to know for sure. I can date him describing it in 1949, which is at least pretty concrete. --Fastfission (talk) 23:10, 16 August 2010 (UTC)


 * I ran across the article again. He doesn't mention 1946 in the 1949 article. --Fastfission (talk) 20:36, 23 June 2011 (UTC)

Middle Square Weyl Sequence PRNG
This section reads like it was written by the publisher of the RNG himself. Have been unable to find independent evaluations thereof. --41.169.80.41 (talk) 14:34, 18 March 2021 (UTC)

Use "Tortoise and the Hare" instead of an array for finding loops !
The Example code would use memory for every possible number. this limits the test to small numbers. Use The Tortoise and the Hare algorithm instead. You can test it for integers of any size with that. 91.56.201.135 (talk) 20:20, 30 December 2021 (UTC)

Ekeland, Borges, Brother Edvin
Can anyone provide evidence, independent of Ivar Ekeland's telling of the story in The Broken Dice, supporting the claim that the middle-square method was invented in the 13th century by a Norwegian monk named Brother Edvin? The claim is quite extraordinary. If it's true, it represents a landmark in the history of randomness as well as an important data point in the spread of decimal numeration in medieval Europe. But I have been unable to find any mention of Brother Edvin or his algorithm that does not derive directly from Ekeland's work.

In the absence of corroborating evidence, I can't help wondering if Brother Edvin is an elaborate joke. BetaPiEta (talk) 15:22, 23 June 2022 (UTC)

9999
99999 116.110.0.66 (talk) 15:56, 23 September 2022 (UTC)