On Thu, 15 Jan 2004, Nick Rout wrote: > seems to me you only need to scan, as aopposed to rigorously compare, > md5sums. change one bit in the file and the md5sum will change beyond > recognition, ie even the first couple of chracters won't be the same. > > IMHO It would take a _very_ clever scam to produce an md5sum that looked > similar enough to fool a casual glance.
It is not so much YHO, but more the mathematical properties of this kind of hash sums :-) BTW, it is not proven that md5sum's for two different files are always different - on the contrary: it can be shown that there exist pairs of different files with identical md5sum's. But practically you cannot profit from it, because you can not compute the files backwards from the md5sum result. One neat and very compact application of md5sum is the fdupes program, which calculates the md5sum's for all files in given directories, and thus identifies duplicate files, even if they have different names. Very useful for tidying up the hard drive... Kind regards, Helmut. +----------------+ | Helmut Walle | | [EMAIL PROTECTED] | | 03 - 388 39 54 | +----------------+
