MD5 idea/theory
this might sound stupid, but i'll say it. lets SUPPOSE , ok ? dont comme telling me there's more, again i say SUPPOSE the number of all possible MD5 hashes is 100 trillion possibilities (its not infinite). so let's say we get a 64 chars long string, with the same charset of the ** always 32 chars long** MD5 hash.. the possibilities will then be, considering my first supposition, 200 trillion.. so there ARE two strings that have the same MD5 hash right ? and if so, MD5 was made to create "fingerprints" for documents, and documents definately are over 32 chars long, even though the possibility is very low (if my idea is correct), it exists.. right ?
the main idea is that there are some strings that have the same MD5 hash..
It is. MD5 Hash collisions have been discovered quite some time ago.
Here`s a paper on it: http://eprint.iacr.org/2004/199.pdf
Yup, it's possible and has been done. It doesn't necessarily mean you have a lot of extra time on your hands, but imagine that you could make it look like bill gates' will gave everything he owned to you :)
A better type of fingerprint should probably be alphanumeric instead of just hexidecimal, and probably more around 65 chars long… That would be nearly impossible to find a collision for with today's technology
Hum for the IP it's true because in a few there is going to be so much computer and server that they are going to miss adresse. I think the new format should be something like 0123:4567:890A:BCDE it's like 16 char in hexa. For the IP it isn't just a rumor some serious magazine about computer & stuff like that have begin to talk about it. For MD6 … I've never heard about it.