Bay 12 Games Forum

Please login or register.

Login with username, password and session length
Advanced search  

Show Posts

This section allows you to view all posts made by this member. Note that you can only see posts made in areas you currently have access to.

Messages - BinaryBeast1010011010

Pages: 1 ... 9 10 [11] 12 13 ... 19
151
DF General Discussion / Re: Mersenne twister in df
« on: May 30, 2012, 04:22:16 am »
So when you reach the last two values, you use the 2nd, xor it with the first and that's all?
If you want to start again do you use the first value as last and xor it with the result of the first iteration?

After how many iterations like these do patterns appear?
On the usability of rng in games like dwarf fortress, do thousands of rolls are generated then stored or is only the seed stored?
The second would seem more sensible but it cant explain why two same seeded world gen end up being different. The world would change every time you load a game would nt it?

152
DF General Discussion / Re: Mersenne twister in df
« on: May 29, 2012, 02:05:11 pm »
an example :
1100 1101 0111 0100 0011 0101 1111 0110 first 32 bit
0011 1010 0110 1100 1111 0010 0001 1001 second 32 bit

temp would be 1111 and 0011 1010 0110 1100 0010 0001 1001
shift right : 0111 1001 1101 0011 0110 0001 000 1100
not odd, dont xor with 9908B0DF

then you xor that with 1100 1101 0111 0100 0011 0101 1111 0110 first 32 bit?

and next iteration you do it using 0011 1010 0110 1100 1111 0010 0001 1001 with the next 32 bit fragment?

thinking it over again, the greatest bits...
temp would look like :0 1011 1010 0110 1100 1111 0010 0001 100
dont xor with 9908b0df
xor with 1100 1101 0111 0100 0011 0101 1111 0110

start again with the second and third fragment

wich one is right? the second I would guess

153
DF General Discussion / Re: Mersenne twister in df
« on: May 29, 2012, 11:24:51 am »
first :is the seed randomness important in any way? I'd say no but wouldnt be the result even more random?

second : so, a 19937 bit long array wich is the seed, divided into 624 bit long segment, the first 32 bit segment is shifted to make temp, then you xor temp and the first segment (nonshifted), if temp is odd, you xore again with 9908b0df.
then you go on with the next 32 bit segment. is that it?

third question : why 9908b2df (having a 32 bit long value is one thing, but would have anything else worked?)
fourth : would the middle square method be reliable to generate the seed?

154
DF General Discussion / Re: Mersenne twister in df
« on: May 29, 2012, 09:22:42 am »
haaan, I understand now. but when does the seed get used?

155
DF General Discussion / Re: Mersenne twister in df
« on: May 29, 2012, 05:46:05 am »
From what I gather, the twister uses a long mersene prime, divide it in segments then takes the first segement, shift it 11 bit then use the temporary first segment created xoring it with the non shifted first segment. Then do the same op with all the remaining segment of the string.
Then if after being xored it is odd it is xored again with 90BDF (or something like that I dont recall)
Question being : here what is the mersenne prime? Is it the seed? What use is the value used in case the fragment is odd?

156
DF General Discussion / Re: Mersenne twister in df
« on: May 29, 2012, 03:17:57 am »
ptw, this is amazing :o
I am going to print this thread as a pdf I think ^^

157
DF General Discussion / Re: Mersenne twister in df
« on: May 26, 2012, 06:43:19 pm »
I find these places more intimidating in fact. I tried to post on the backtrack forum but I could not create any new thread, post on old ones or even try and pm the moderators. This is quite intimidating.

158
DF General Discussion / Re: Mersenne twister in df
« on: May 26, 2012, 06:25:24 pm »
I used the aircrack suite to try and get in, it is pretty easy once you get used to it. For a first try I must say that I really like linux. I tried myself at the "metasploit unleashed" tutorial but could not find the so called "dradis framework". I muste be doing something wrong since I use backtrack5 r2 kde 32 and it should be there...
About metasploit there are a lot of things I dont understand and the tutorial's writer seems to assume them as widely known so he does not explain them...

159
DF General Discussion / Re: Mersenne twister in df
« on: May 26, 2012, 03:52:47 pm »
I thought the seed was changed every time you gen a world, maybe using the mouse cursor movement or something else.
 if you gen two worlds with the same seed (if it is possible) will they be the same?

160
DF General Discussion / Re: Mersenne twister in df
« on: May 26, 2012, 02:12:07 pm »
Macchanger is the first example that came to my mind (since I have not been playing darwinbots for a while, the rand function came up after), I am new to linux, two weeks ago my girlfriend find out the hard way that someone has been downloading copyrighted material with her internet connection, I decided to learn more about it and after a week of constantly studying and trying to understand I learned that even a quite long key was ridiculous if you use wep. I'll say that I knew nothing of what was the later and only had an hint about the former before that week (duh? wep? isnt this some internet thingie?)...
I dont want to be taken seriously since I am a musician and have no skills in programming, but I can get quite obsessive when I dont understand something and I love to learn.
To get back to the crypto track, how the use of a truely random seed state could result in a predictable string of characters?

(I strayed from df, the combat use of random numbers was quite obvious -facedesk- but for terrain generation... I saw lovely fractal landscapes on the web, with moutains, rivers but I did not know that there was any element of randomness in fractals, quite the opposite I thought)

161
DF General Discussion / Re: Mersenne twister in df
« on: May 26, 2012, 06:41:43 am »
As I understand it it does not start from a seed state so it isnt usable as a key generator for two communicating computers.
Is it the one used in backtrack with the "macchanger -r" function?

162
DF General Discussion / Re: Mersenne twister in df
« on: May 26, 2012, 05:15:21 am »
You did help me indeed thank you. I do apologise if my first sentence seemed harsh.
The cryptographic use random numbers in one time pads always fascinated me, the problem is how to obtain them. How old ladies thinking bingo numbers was not random enough for cryptographic use is one example. Blum blum shub uses big primes but I think we can predict that in some years (if I am optimistic, by the end of this century or the next) we well be able to factorize their product so it will be useless. The future of cryptography (it's only my opinion) lies in one time pads (the quantic computer does, in theory at least,  have a mean to securely transmit the keys). As one time pad is a perfectly secure encryption (provided the key is truely unpredictable) I hoped there also was a way to generate keys for them without using expensive radioactive thingies (numerized lava lamps are cool but if they could be replaced by an human written program wich only needs a truely random seed to create an infinite amount of random numbers it would be even better).

so once the first bit of a(1) followed by the last 31 bit has been shifted right and xored with 9908B0D, (if it is odd it is xored again with the remaining bits of the 623 bit long string)
we keep the result as first part of the random number string. if temp is even do the other bits just stay the way they are?

163
DF General Discussion / Re: Mersenne twister in df
« on: May 25, 2012, 06:06:12 pm »
I know what is binary thank you, the use of xor for non binary numberssurprised me. I do not NEED to know, I'm just interested in the subject. I dont want to know how toady gen all the things (since it is under the hood and such things are not to be known).

I'm interested in mersenne twister because it seems more usable than blum blum shub. Until a while I thought the only way to generate truely random strings of numbers was using radioactive material, I heard about the middle numbers algorithm but it does not seem "random enough" for things like what df does.

about bitwise command I learned them playing around with darwinbots (I know thats no programing school but that's the best I have)
 it's "temp" that I dont understand.

if it is secret knowledge though I wont bother you anymore ;)

nice calling me a troll when I only look for teachings from more competent people.

164
DF General Discussion / Re: Mersenne twister in df
« on: May 25, 2012, 05:20:11 pm »
I understand that it twist them around but :
first, to xor dont you need the bits to be binary?
a here being the whole number and a(1) the segment used during the iteration?
and what does
a(i) = temp shifted right one bit xor
         X'9908B0DF' if temp is odd xor
         a(i+397)
mean?

165
DF General Discussion / Re: Mersenne twister in df
« on: May 25, 2012, 04:26:03 pm »
ok I get it, now, does anyone know how the mersenne twister algorithm works?

Pages: 1 ... 9 10 [11] 12 13 ... 19