home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!att!dptg!ulysses!ulysses.att.com!smb
- From: smb@ulysses.att.com (Steven Bellovin)
- Newsgroups: alt.hackers
- Subject: Re: Prime Number Generator
- Message-ID: <17166@ulysses.att.com>
- Date: 31 Aug 92 22:33:27 GMT
- References: <grtyj5-@rpi.edu> <1992Aug31.160225.26724@gateway.novell.com>
- Sender: netnews@ulysses.att.com
- Lines: 22
- Approved: xyzzy@plugh.cave.org
-
- In article <1992Aug31.160225.26724@gateway.novell.com>, alex@otis (The Console
- DJ) writes:
- > Someone correct me if I am wrong but the whole idea behind the encryption is
- > the large prime numbers are difficult to calulate??? Actually there should
- > be no known algorithm (except brute force try every number thingy).
-
- OK, you're wrong... Generating large prime numbers is quite easy -- one
- generates random odd numbers, does a few trial divisions, then applies
- as many iterations as desired of probabilistic prime number generators.
- The Sieve of Erastosthenes or exhaustive division trials would take far
- too long for numbers of cryptographic interest, which are several hundred
- bits long, at the very least.
-
- The hard problems are things like factoring large numbers, or calculating
- discrete logarithms in finite fields. If you're interested in the subject,
- read sci.crypt for a while. Or read some books or papers on the subject.
- The book I recommend is Davies and Prices's ``Security for Computer
- Networks''.
-
-
- --Steve Bellovin
-
-