 |
|
You are not logged in! F.A.Q Log in Register |
 |
|
 |
 Now online (2)recycle belb ...and 334 guests Last 5 registered Oplandisks nothingstar N_loop yipe foxtrotromeo Browse members... |
 |
Members 8025Messages 2614103 Today 16 Topics 127542 |
|
|||||||||
| |||||||||
|
|
What I think the result must be, is that dices can actually create truly random sequences of 6 mill. throws, even though they're not perfect. And that's why a real dice can be enough to be "practically" random. You do have a point however when you'd repeat the experiment. But only after many repeats the imperfections of my reasonable real dice (eg. no silly things as sticky sides or other "forgeries") would show up. But, nevertheless, the sequences it creates are random. |
||||||||
|
|
|||||||||
| |||||||||
|
|
Sorry for the ramble. I was having "one of those moments". |
||||||||
|
|
|||||||||
| |||||||||
|
|
agreed. the distribution of outcomes might be skewed (i imagine your average dice landing slighty different from a perfect cube because of imperfections and all), but all in all, who gives a shit as the process that creates the event (given that you don't cheat when you throw the dice) is 'real-world', not just playing with the modulus of your computers system clock, and thus unpredictable. butterfly effect, laplace, heisenberg, they're all sitting on my shoulder having a mad party, and i believe the stuff about only quantum events being truly random is bullshit. take random source, normalize, done. COFFEEEEE HRAHRAHRA. |
||||||||
|
|
|||||||||
| |||||||||
|
|
also i named my wireless networks 'scylla' and 'charybdis' HOOT HOOT |
||||||||
|
|
|||||||||
| |||||||||
|
|
w00t + GenuineLOL = RandomFun |
||||||||
|
|
|||||||||
| |||||||||
|
|
Fine, write a paper on the matter. I'm sure if you do it in an academic enough way it'd be published; journals love stuff that disagrees with the accepted wisdom. Your bit, "if you really want to do it the fancy way, feed the rotation and launch velocity from a random source like currents in a stirred water tank" is the weak part of your argument though; what you're talking about is make x more random by combining it in some way with y. The thing is that yes, these compound sources of randomness do make things nearer to true randomness and often iron out some of the glitches in just using one source. The thing is, it's like using a word to describe itself: You can't make something truly random by combining it with something that isn't random. Unless the second source is truly random (in which case, you should just use that), it won't ever make the first truly random! Bit of a mindfuck, I know, but that's why it interests me. Again, this part aids my argument: "currents in a stirred water tank, notoriously unpredictable too." Yes, turbulence is incredibly unpredictable (I remember a renowned scientist giving the answer "why turbulence?" as the question he would ask God is given the opportunity!) but 'unpredictable' is the key word- it's not truly random. goDel: I hear what you're saying - it's possible the die could have come up '6' 6,000,000 times in a row by chance, it's no indication that it will the next time. There are, however, actually several accepted mathematical tests on randomness. They basically work on the principle that over a very large sample set (millions) around 99% should be distributed as expected. Yes, this isn't certain either, but the statistical likelihood of it being false dwindles the larger the sample set. Everyone: Interesting hearing all these different views on the matter, that's why it interests me so much! :) |
||||||||
|
|
|||||||||
| |||||||||
|
|
Whilst I completely agree with this for practical uses (even for encryption at national security levels), the big thing isn't for using it now, it's for 20-50 years time once quantum computing is properly developed. 1000 years ago, a lot of things back then considered "unpredictable" and useful as a source of random numbers (what the weather might be doing in 50 days time in Britain, for example ;-), we can now work these things out in seconds using computers. As computing power increases, so will our ability to model these things incredibly accurately improve. If it gets to the stage where all people have to do is work out what you're doing to generate your random numbers (IE your method rather than numbers themselves), security will be lowered dramatically. dave_g: What you say about working and studying is so true. On my course the part time students who work in the field consistantly piss over the other students from a great height out of a class of 60 odd students, the top 5 were all part timers. There are only 7 part timers in total! As well as being more experienced, we're generally a lot more motivated. " I do think there are too many people at university. The government policy is ultimately at fault here. A degree is not necessarily analogous with sucess." 100% correct. I hate the snobbery directed at the trades in this country. I know lots of people who are leaving IT and training to be electricians/plasterers etc. far better to do that straight away and have yourself fully trained 3 years sooner (not to mention no debt, paid 3 years extra, 3 years extra pention, so nearer retirement, etc.) than wasting 3 years studying something you're not into. DM: You talk about it being "as good as". This is why I like the idea of true randomness, it's almost like computing/maths to the level of a philosophical truth. Once you attain it (for the same price/ease of use), all your "nearly random" ones, no matter how close, are effectively rendered null and void. |
||||||||
|
|
|||||||||
| |||||||||
|
|
I know how you can get something TRULY random: start a thread on xlt and see what happens! Hawhawhaw!! |
||||||||
|
|
|||||||||
| |||||||||
|
|
Not true. You can guarentee with 100% certainty that Monoid will depress you, dog_belch will call you a cunt, and Hanal will show you his arse. |
||||||||
|
|
|||||||||
| |||||||||
|
|
Ceri: For a cheap random number generator, couldn't you simply digitise a white noise source? How about this for a cheap 'computerisable' system. Build a small circuit which amplifies the thermal (white) noise produced over a p-n junction on a device such as a BJT transistor or a zener diode. Connect this to the line-in on a soundcard and sample it. If the instantaneous level of noise is higher than fullscale/2 call it 1, else call it 0. et voila. Construct the whole thing in an earthed shielded box to keep out external interference, and use an external DC power supply and filter it to remove external noise. Place the whole assembly in a temperature controlled "oven" enclosure so that it operates at say 70C so outside temperature changes have no effect. As far as I'm aware all variables in the equation should now be constants, so the thermal noise will be pure white = 100% random. What do you think? |
||||||||
|
|
|||||||||
| |||||||||
|
|
that would be sooo much better if it included a waterfall and you had to go to the waterfall with a microphone to decode the information |
||||||||
|
|
|||||||||
| |||||||||
|
|
That's planned for version 2. Also a range of xpress on covers and +260 extra pixels. |
||||||||
|
|
|||||||||
| |||||||||
|
|
Digitising white noise is a recognised cheap way of doing it. Yours is a good description of it and I particularly like the protection against external intereference. I'll have to read up on it a bit more (still writing this proposal), but it's definately one I'll consider. Ta for that. |
||||||||
|
|
|||||||||
| |||||||||
|
|
dave_g , What kind of work do you do? |
||||||||
|
|
|||||||||
| |||||||||
|
|
I do electronics as a career. The company I work for makes test and measuremnt equipment. Currently I'm working on a series of modular RF/microwave instruments. Interesting stuff! By night I'm an IDM superstar DJ. |
||||||||
|
|
|||||||||
| |||||||||
|
|
your avatar says it all... ;) |
||||||||
|
|
|||||||||
| |||||||||
|
|
PISSED_ HGAHAHAHAAH! |
||||||||
|
|
|||||||||
| |||||||||
|
|
university in action |
||||||||
|
|
|||||||||
|
Messageboard index
|
|||||||||
