Here are some. What are some other?
* timing of radioactive decays
* recording background noise into .wav files
* live video files
* session logs from chat lines
* keystroke latencies for typed text
Asked
Active
Viewed 262 times
1
124697
- 22,097
- 68
- 188
- 315
-
Your question might be better (and on-topic) if it asked how to create a truly random (and not pseudo-random) numeric sequence inside a computer. Actually, that question has been asked several times already; see here: http://www.google.com/search?hl=en&source=hp&biw=1920&bih=1113&q=create+truly+random+numbers+site:stackoverflow.com&aq=f&aqi=&aql=&oq=&gs_rfai= – Robert Harvey Dec 31 '10 at 16:32
-
[dice rolls](http://xkcd.com/221/ "dice rolls") – David Dec 31 '10 at 16:33
-
@user521180: it depends on your definition of randomness ;) I mean, at one point it becomes some kind of a metaphysical question! – SyntaxT3rr0r Dec 31 '10 at 16:36
-
From the "Related" sidebar: [Random numbers from physical sources on the Internet](http://stackoverflow.com/q/782210/2509) for sources of such numbers ready-made; [Generating totally random numbers without random function?](http://stackoverflow.com/q/3834102/2509) a duplicate of [True random number generator](http://stackoverflow.com/q/37702/2509). Not a duplicate as such, but most of the answers have been discussed in these and other questions already. – dmckee --- ex-moderator kitten Dec 31 '10 at 16:57
-
@SpoonBender It really doesn't. It's simply a matter of clearly defining what you want. – Nick Johnson Dec 31 '10 at 23:00
-
The trickier part is measuring entropy. – President James K. Polk Dec 31 '10 at 23:15
-
@Nick Johnson: it really does... – SyntaxT3rr0r Jan 12 '11 at 21:05
-
There's nothing "metaphysical" about asking about sources of random data. It's well defined in information theory - see shannon information, for example. – Nick Johnson Jan 12 '11 at 22:40
4 Answers
3
- Mouse movements
- Soil moisture
- Number of ants passing through a point in an antfarm.
- Checksum/hash of data from the frontpage of a very dynamic website (like twitter).
Danny Staple
- 7,101
- 4
- 43
- 56
2
The nice random.org site by Mads Haahr has been providing hardware RNG draws via a web interface for many years.
My random package for R uses this:
R> library(random)
R> rn <- randomNumbers(100, 1, 100, 5)
R> head(rn)
V1 V2 V3 V4 V5
[1,] 96 99 68 95 90
[2,] 84 11 62 20 40
[3,] 83 79 38 32 90
[4,] 28 18 32 58 97
[5,] 15 31 42 68 98
[6,] 65 93 6 15 55
R> rs <- randomSequence(1, 20)
R> head(rs)
V1
[1,] 14
[2,] 4
[3,] 6
[4,] 11
[5,] 13
[6,] 20
R>
There is ample documentation in an introduction by Mads which I include with the package
Dirk Eddelbuettel
- 360,940
- 56
- 644
- 725
0
JavaScript creates its "random" numbers from an internal clock in milliseconds, which is quite random. Other sources may include atmospheric pressure and public keys.
Kevin Ji
- 10,479
- 4
- 40
- 63
-
2JS random numbers (and most the 'random' call in pretty much any language) are pseudo-random number generators (http://en.wikipedia.org/wiki/Pseudorandom_number_generator), not truly random. And how are 'public keys' a source of random data? – Jonathan Rupp Dec 31 '10 at 16:44
-
Clocks are random now? I'll have to remember that next time I'm late. – Nick Johnson Dec 31 '10 at 23:00
0
http://en.wikipedia.org/wiki/Hardware_random_number_generator
In short - best random sources are such those roots are based on laws of quantum mechanics or chaos theory.
Agnius Vasiliauskas
- 10,935
- 5
- 50
- 70