##project## MD5 analysis... 3.4028236692093846346337460743177e+38 numbers

[jimson]

i think it is maybe tooks ages for us to have the complete
md5 hash pair data so.. like weed said... is correct about the storage we need to store those pair of 32 chars.

ok.. why dont we just collect the pair result data that start with 0

so (16^32)/16 = 2.1267647932558653966460912964486e+37
so, does it sounds possible ?
or 10000000000000000000000000000000 hex = ? TB?

ok, i think another method
16¹⁶ = 18446744073709551616
18446744073709551616/16 = 1152921504606846976
em, we just having the first 16 char of 32 md5 char.
so we will have
0000000000000000 to 1111111111111111
so we will have pair that generate

aik... ok.. found another solution..
so, our goal is to prove the theory is WRONG...
so, md5('jimsonchang') = 67c356a7e59cda2160d0b86c389264c0

so we just use the computer power to find the 32 chars that can generated "67c356a7e59cda2160d0b86c389264c0"
what ya think?
maybe we will have a list of peeps who join
where their name <32 chars
and database will try to math the list for occurance ?

how u think...?

can anyone tell me, if i run 24 X 7 X 365, how many numbers i can get, not stored in database, just match the list?

sorry weed, about my STYLE... ? <= i got style. ?.. stylish... 🙂 🙁

[eb1024]

edited

[Weedpacket]

Originally posted by jimson

so (16^32)/16 = 2.1267647932558653966460912964486e+37
so, does it sounds possible ?
or 10000000000000000000000000000000 hex = ? TB?

No, (16^32)/16 = 16³¹ is only a factor of 16 difference - in grains of sand measure that's 20 million times the size of the Eath. And that is assuming of course that you can tell in advance which strings hash to MD5 hashes where the first four bits are all 0 and how do you know which ones those are?

Okay, I just generated a thousand different MD5 hashes. Doing it in PHP on my machine it took a bit less than 1/80 of a second, but doing nothing except generating them, and on trivial strings at that. But clever coding and hand-tooled assembler may make things a bit faster. Let's be generous and assume therefore a machine can test 100,000 hashes in a second. Let's say we can get a million machines working on the problem. That's 100 billion hashes tested per second. To test 16³¹ numbers would take only about 450 million times the present age of the Universe. Maybe I should go away and see how Moore's Law would affect this estimate.

ok, i think another method
16¹⁶ = 18446744073709551616
18446744073709551616/16 = 1152921504606846976
em, we just having the first 16 char of 32 md5 char.
so we will have
0000000000000000 to 1111111111111111
so we will have pair that generate

Again, how do you know which ones they are? You're going to crank out MD5 hashes and then try and find matches to each of them in turn? The 450 million above assumes you're just generating strings and then computing the hashes.

aik... ok.. found another solution..
so, our goal is to prove the theory is WRONG...

"The theory" - oh, I love it when people use that phrase; it almost guarantees comedy. Note that the abstract of RFC1321 states that "it is conjectured that it is computationally infeasible...". In other words, there is no theory; just a lot of empirical and theoretical evidence to the effect.

It would be much more interesting to develop a computationally feasible means of reversing MD5 to find which message(s) generate a given hash. That's why I wanted to know if there was anyone with cryptographic experience here.

so we just use the computer power to find the 32 chars that can generated "67c356a7e59cda2160d0b86c389264c0"
what ya think?

How? By developing an algorithm that would work in a feasible amount of time and space for any hash, or by mindlessly chucking strings at this particular hash until one sticks? The first would qualify as cracking MD5, the latter would just be guessing passwords.

maybe we will have a list of peeps who join
where their name <32 chars
and database will try to math the list for occurance ?

Why? You've got their name, just hash it! Who needs to look it up in a big table of hashes?

I think you don't actually understand what a Message Digest is. The next sentence of RFC1321's abstract is: "The MD5 algorithm is intended for digital signature applications, where a large file must be 'compressed' in a secure manner before being encrypted with a private (secret) key under a public-key cryptosystem such as RSA."

sorry weed, about my STYLE... ? <= i got style. ?.. stylish... 🙂 🙁

Hence the quotes around the phrase.

Again, the challenge hash I posted above still stands.

[BuzzLY]

Weed, you are absolutely amazing.

1. You understand what jimson is talking about.

2. You take the time to debate the issue with him, using very sound and convincing arguments.

You strike me as someone that would sit for 4 hours and seriously answer every time a 4 year old says "why?"

So tell me... what do YOU think is the meaning of life? And don't tell me "a1d0c6e83f027327d8461063f4ac58a6," either.

[Weedpacket]

Originally posted by BuzzLY
You strike me as someone that would sit for 4 hours and seriously answer every time a 4 year old says "why?"

Ah, I have a cunning plan to deal with that. They ask me why there are rainbows and I give them a fifteen minute lecture about trigonometry, differential refraction, and atomic energy level transitions.

THAT usually shuts them up 😃.

So tell me... what do YOU think is the meaning of life? And don't tell me "a1d0c6e83f027327d8461063f4ac58a6," either.

Oooh, a bit early to answer that one, don't you think? Probably don't know near enough yet. Come back in five million years' time; maybe we'll have a better idea by then. Myself right now, I reckon it would be a shame if the Universe were to just burn itself out and go cold because no-one was looking after it...

[eb1024]

Solar System will crush with Andromeda in 25 million years according to cientists... :p

[The_Chancer]

I thought it was just used to compare two strings for their "likeness" and the probablilty that the two were the same. Not that it was set in concrete or anything.

Surely for "proper" cryptography you would need an alogorithm as well as a key..... MD5 is OK for hasing a file and letting people have it to compare for example a downloaded file against the original to make sure it has not been tampered with, but to use it as a password "checker" or full encryption would be nonsensical...

(Just my $0.2c)...

[Weedpacket]

I presume by that you mean that the Milky Way galaxy is supposed to collide with the Andromeda galaxy? No, I don't think that the Andromeda galaxy is approaching at nearly 10% of the speed of light. More like one-thousandth of that. So 25 million years? Assuming no proper motion with respect to each other that pushes collision out to something more like 25 bilion years. The solar system will have ceased to be before that.

[eb1024]

If my memory is right Andromeda will colide to our galaxy in 25 million years... discovery channel :p hehe

[jimson]

weed, the method i written before

aik... ok.. found another solution..

is my thought before found out this solution...

currently running on my pc

echo md5('jimson'); // this is the $start.

function md5_loop($start = '892433944a24f19bb32b34516874a6ff') {
	static $prev = '';
	while(md5($prev) != $start) {
		$prev = md5($prev);
	}
	// inform us the $prev values.
}
md5_loop();

i am not guessing the password since i don't provide it with password, i just loop from the 32 chars
since hash of 32 chars will return another 32 chars.
so, if i am right , half of the 16³² is md5 values of another 16^32.

ahh.... why the browser stop after minutes?
any idea to make the script runs forever until it found answer?
my php.ini
resource part
all set to 0 already...

or, i need to start the VB to run this forever... ?
any ideas?

if u know jimson = 892433944a24f19bb32b34516874a6ff
so, there is another one 32 chars that will reffer to 892433944a24f19bb32b34516874a6ff

[jimson]

half of the 16³² is md5 values of another 16^32.

i mean (16^32)/2 md5 values = (16^32)/2 md5 values.
since the pair occupies 2 md5 values.

[jimson]

i don't think there is time for us to run all 16³²
but who know where the "892433944a24f19bb32b34516874a6ff" is located
in the 16^32?
maybe it is in the front part of 16^32.(so we don't spend years to get the result)
ok, this come to another problem,

can we check a md5 chars whether it is in the front, middle or back of series of 16^32?
starting with md5('')?

i know the intended usage of md5 is for message digest
but u see many peeps use it to hash real password and store it in database.
comparing them when user log on.

ok.. just get another thought

if 16³² is the limit, and half of them = md5 of another half of them, then
every pair is basically a knot that tie them together to create a pole of 16³² numbers.
so, it means there is relationship there.(since it is circle like).

buzzLY...

You understand what jimson is talking about.

am i so hard to understand?

anyway, i admit, i preffer if i am 4 years old kid. no bother, no worries...
if everything can start again.

[Weedpacket]

I don't see what you're trying to achieve there. What's so special about the sequence

d41d8cd98f00b204e9800998ecf8427e
74be16979710d4c4e7c6647856088456
acf7ef943fdeb3cbfed8dd0d8f584731
5a8dccb220de5c6775c873ead6ff2e43
76682f743ae018364a082b2e87f2d2f5
0f62265227df1b6d6deec36ab4bc5e76
670df58df5a2ec63b0a33e054418105a
6a0d8f756dfa9c1cd69fed28423ec8ca
628f7643f52b6d6c34accad8a8dc2867
1f16a6c017f12c1539c4947e234de78a
...

That you hope to find

892433944a24f19bb32b34516874a6ff

in it? And what would it prove if you did?

[Weedpacket]

Of course there are periodic sequences - this is inevitable, since every MD5 hash itself has an MD5 hash, and there are only a finite number of MD5 hashes. So? I fail to see (a) where you get this "first half second half" idea from, (b), what's so special about the sequence you're generating that makes it easier to search (and search for what?) than just starting with 00000000000000000000000000000000, and then going on to 00000000000000000000000000000001, 00000000000000000000000000000002, ... (that's easier to search: look there's 892433944a24f19bb32b34516874a6ff! Right after 892433944a24f19bb32b34516874a6fe!)
and (c) what this has to do with finding passwords.

[Weedpacket]

If you were going to look for passwords, a more suitable sequence of generated md5's would be

d41d8cd98f00b204e9800998ecf8427e
93b885adfe0da089cdf634904fd59f71
55a54008ad1ba589aa210d2629c1df41
9e688c58a5487b8eaf69c9e1005ad0bf
8666683506aacd900bbd5a74ac4edf68
ec7f7e7bb43742ce868145f71d37b53c
8bb6c17838643f9691cc6a4de6c51709
06eca1b437c7904cc3ce6546c8110110
89e74e640b8c46257a29de0616794d5d
e2ba905bf306f46faca223d3cb20e2cf
5e732a1878be2342dbfeff5fe3ca5aa3
...

Along with of course strings that hash to them.

[jimson]

i got another thought... thanks God...

okay,
let say
md5('1') = c4ca4238a0b923820dcc509a6f75849b
md5('01') = 96a3be3cf272e017046d1b2674a52bd3
md5('001') = dc5c7986daef50c1e02ab09b442ee34f
...
md5('00000000000000000000000000000001') = e629f196dc612509ea9a61df552423a0

so, just only 1 number can generated 32 md5 numbers.
decrease to 31 when reaching 10

16 ^ 1 = 16
16 ^ 2 = 256
16 ^ 3 = 4096
16 ^ 4 = 65536
16 ^ 5 = 1048576
16 ^ 6 = 16777216
16 ^ 7 = 268435456
16 ^ 8 = 4294967296
16 ^ 9 = 68719476736
16 ^ 10 = 1099511627776
16 ^ 11 = 17592186044416
16 ^ 12 = 2.8147497671066E+014
16 ^ 13 = 4.5035996273705E+015
16 ^ 14 = 7.2057594037928E+016
16 ^ 15 = 1.1529215046068E+018
16 ^ 16 = 1.844674407371E+019
16 ^ 17 = 2.9514790517935E+020
16 ^ 18 = 4.7223664828696E+021
16 ^ 19 = 7.5557863725914E+022
16 ^ 20 = 1.2089258196146E+024
16 ^ 21 = 1.9342813113834E+025
16 ^ 22 = 3.0948500982135E+026
16 ^ 23 = 4.9517601571415E+027
16 ^ 24 = 7.9228162514264E+028
16 ^ 25 = 1.2676506002282E+030
16 ^ 26 = 2.0282409603652E+031
16 ^ 27 = 3.2451855365843E+032
16 ^ 28 = 5.1922968585348E+033
16 ^ 29 = 8.3076749736557E+034
16 ^ 30 = 1.3292279957849E+036
16 ^ 31 = 2.1267647932559E+037
16 ^ 32 = 3.4028236692094E+038

the limit is 16^32,
so, the total from 16¹ to 16³¹ = 3.62967858049E+038
which is obviously larger than 16³² limitation.
and we just use only hex values which is 16 chars only.

okay, right now, i got one question...

WHO HAVE ALREADY FOUND THE FIRST CLASH VALUE.

which means who are the person who have 2 strings or more which can generate the same md5 hash ?

if u got info, please post here.

the first half, second half idea is... i thought
if we just concentrate on the checksum that contained
16 chars first half is = c4ca4238a0b92382
so we need to find what numbers that generated
the 16¹⁶ pairs<=second half. // maybe this thought fail to stand.

[BuzzLY]

jimson, I apologize -- I did not mean to offend you with my post. The way you word things, I find difficult to understand -- perhaps because of your translation from your language to English. I did not mean any disrespect.

And the comment about 4 year olds was not meant to insinuate that YOU were 4 -- only that Weedpacket sometimes seems to have the patience of Job, and would probably spend forever answering a 4-year-old's questions.

Hope that clears things up -- I did not mean any disrespect to you. You seem like a VERY intelligent person -- probably way beyond my intellect! My comments to Weedpacket were intended as a compliment to you.

[goldbug]

Am I the only one that finds it extremely silly to create/run a brute-force style method using PHP (or any other "scripting" language)??

[Mordecai]

Originally posted by jimson
hi, can u please give me a summary what ur crack() function will do...

Umm, look at the code. It tests passwords against another password.

I built an addon (basically just added md5()) to test 1200 words against and MD5'd word.

[jimson]

Buzzly, thank u...
looks like i have to study english... grammar or vocabulary ?
for the VERY intelligent person... well, i dare not take it.
i know only little very little. anyway... goldbug just prove that i am silly... 🙂

Am I the only one that finds it extremely silly to create/run a brute-force style method using PHP (or any other "scripting" language)??

actually, the reason i use php is just to present the concept, if i gonna run for a long-long time, then i will fireup the VB or C, C++ and do some small app to make it runs.
and we are talking about the 16³² numbers... extremely huge
what i know is, we can't run all of them.

since, i believe there is some kinda relationship between each md5 checksum value (increment of 1h, 0x31 0 to 0x31 1), i am thinking a way to predict a range where possible 32 chars will generate the specified md5 checksum.