Half a croissant, on a plate, with a sign in front of it saying '50c'
h a l f b a k e r y
Bunned. James Bunned.

idea: add, search, annotate, link, view, overview, recent, by name, random

meta: news, help, about, links, report a problem

account: browse anonymously, or get an account and write.



Striped Fish Randomizer

Nature's Barcode
  [vote for,

According to mathematicians getting a truly random number can be tricky. I figured that tricky meant profitable, so I would like to give it a go...

Nature provides unique patterns in the shape of animals with stripes. Such animals can be 'read' using the simplest of equipment, a jeep, laptop, bar code reader, bespoke software and a long stick. The zebras of the Serengeti are plentiful, offering a large enough population to be statistically valid.

Indeed, a non-repeating set of random numbers could be generated quite simply using zebras, hungry lions and confined spaces.

Obviously, a 'home-user' might not possess a long stick, so a more portable device would be required. The solution lies with zebra fish, smaller and more practical then their equine cousins. A home user's device would take the form of a fish tank and a Bingo machine, capable of pushing a fish randomly to the surface.

However, a 'power-user' might need more randomness then nature can supply. Fortunately, some clever scientists have provided the answer,'mutant zebra fish', including one variety called a 'half baked' - see link.

riposte, Mar 22 2001

Clever Scientists http://www.fi.edu/q...journals/part6.html
Mutant Fish [riposte, Mar 22 2001, last modified Oct 21 2004]

Cuddly Zebras... http://www.serengetipark.org/
...Hungry Lions [riposte, Mar 22 2001, last modified Oct 21 2004]

LavaRand http://lavarand.sgi.com/
"... harnessing the power of Lava Lite® lamps to generate truly random numbers since 1996." [egnor, Mar 22 2001, last modified Oct 21 2004]

HotBits http://www.fourmilab.ch/hotbits/
"Genuine random numbers, generated by radioactive decay" [egnor, Mar 22 2001, last modified Oct 21 2004]

Maybe we could use Mr Mandela's former neighbours. http://news.bbc.co....ci/tech/7475654.stm
[AbsintheWithoutLeave, Jun 28 2008]


       remind me again why we need random numbers?????
Susen, Mar 22 2001

       Random numbers allow zebra-fish to maintain anonymity through encrypted sessions on the internet. I have yet to catch one out...
riposte, Mar 22 2001

       Like PeterSealy mentioned, anything that relies on genetics and animal behavior is unlikely to be truly random.   

       Additionally, as far as encryption goes, all that's necessary to crack it is stealing the fishtank or going to the same region and trying all the keys.
bookworm, Mar 22 2001

       Cryptographically secure random numbers are indeed desirable, but there are much, much easier ways to harvest them. A small, cheap electronic component will do the job; absent that, most systems get by just fine collecting entropy from local random data (keypress timings, etc).   

       There are a couple of other silly projects (mostly done quite tongue in cheek) to harvest random data from physical observables. These are not actually useful, mind you, but I've added links.   

       I'm finally just a little surprised that after a lifetime of seeing natural objects (even relatively pedestrian ones like trees, grass, snowflakes, heads of lettuce, ...) you're still able to conclude that the patterns of zebra stripes are somehow uniquely random. Random variability (usually derived from sensitive dependence on local environmental conditions at some point in the past) is the rule, not the exception. When was the last time you saw a perfectly conical tree?
egnor, Mar 22 2001

       I'm for this one. Surely the randomness has more to do with the reading than with the individual zebra. Zebras are not striped evenly all over their bodies so a different number will be produced depending how you swipe them (top left to bottom right v top right to bottom left etc). Even if you swipe them in the same place, rippling muscles of a running zebra will make some stripes thicker than those of one casually grazing.
Gordon Comstock, Mar 22 2001

       So, egnor, if I read your last paragraph correctly, you're saying the inventor of this idea might as well just scan trees, or their neighbor's vegetable garden instead of an exotic fishtank, right?
PotatoStew, Mar 22 2001

       What susen said.
iuvare, Mar 22 2001

       Yes, PotatoStew, that's right.
egnor, Mar 22 2001

       Practice Random Acts Of Scanning And Senseless Acts Of Crypography
thumbwax, Mar 22 2001

       Or, with a sufficiently sensitive instrument, random floating dust particles in the air could be scanned... random numbers could literally be pulled out of thin air.
PotatoStew, Mar 22 2001

       iuvare+susen: Random numbers are used for cryptography (ie allowing you to perform secure transactions over the net). They`re also used, bizarrely enough, for some pretty mundane operations, such as sorting numbers quickly (using the Quicksort algorithm), which you`ll appreciate if you use (directly or indirectly) large databases.
Pallex, Jun 11 2001

       In a world of automation we need to solve the 'long stick' design problem. How about we borrow one of the surveillance satellites from the US government, and tinker with the software so that it can read barcodes. This method allows the satellite to locate a zebra from space, read the barcode and beam the signal down to the place where the random number is needed most.   

       Of course, there is always the possibility that you could select the same zebra. And as we can't randomly choose a zebra then the only solution seems to be that we destroy the zebra after use. However, a counter argument may be that possibly the zebra population would decrease to 0 in a day! should we worry about selecting the same zebra again, because the lines of the zebra are not linear. And what are the chances of scanning the same area of the zebra? fairly small if you take the fact that the earth is rotating. Of course this means that the satellite could NOT be geostationary, cause they circle the earth on a non linear orbit.
u8djf, Sep 19 2001

       Since most zebra (not in zoos) are in Africa, I'm not sure a random orbit for the satellite would be appropriate.
phoenix, Sep 19 2001

       Humans have random number generators built in. 0 9 44 81 36 123 18 62 457 92. Why exactly do we need to extract random numbers from zebras?
Bukkakinator, Jun 26 2008

       //Why exactly do we need to extract random numbers from zebras// Because humans are crap at random number generation. For example, your sequence clusters strongly between 0 and 100 (despite extending as high as 457), and has no consecutive numbers with the same last digit. It also tends to oscillate (large/small/large), avoiding successive close numbers or trends. It also favours alternating odd and even numbers.   

       Human beings are simply incapable of avoiding patterns.
MaxwellBuchanan, Jun 26 2008

       ...or sometimes too good at avoiding patterns. For example no one chooses 1, 2, 3, 4, 5, 6 as their lottery numbers, despite it being just as likely to come up as any other random choice of lottery numbers.

The problem with the zebras idea is separating the randomness from the order. So the number of stripes a zebra has on its back, and the thickness of those stripes are, in fact, not random. They probably form a bell-curve-like distribution within a smallish range. If you know the precise shape of this distribution, you can extract a random number from the location of any specific zebra within this range, but that's a big "if".
hippo, Jun 26 2008


back: main index

business  computer  culture  fashion  food  halfbakery  home  other  product  public  science  sport  vehicle