home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky sci.crypt:7052 sci.math:18557
- Newsgroups: sci.crypt,sci.math
- Path: sparky!uunet!usc!sdd.hp.com!nigel.msen.com!fmsrl7!lynx.unm.edu!umn.edu!csus.edu!netcom.com!rcain
- From: rcain@netcom.com (Robert Cain)
- Subject: Re: RSA cracked? I don't think so..., HUGE primes?
- Message-ID: <1993Jan21.070521.9787@netcom.com>
- Organization: Netcom - Online Communication Services (408 241-9760 guest)
- References: <1993Jan20.231041.4691@zip.eecs.umich.edu>
- Distribution: na
- Date: Thu, 21 Jan 1993 07:05:21 GMT
- Lines: 52
-
- gilgalad@quip.eecs.umich.edu (Ralph Seguin) writes:
- : A friend of mine posessing a masters in mathematics said that he had heard
- : and saw a flyer in his local math dept that RSA had been cracked. I told
-
- Could this flyer be obtained?
-
- : him that I thought that this was a load of bunk. We got into a heated
- : discussion and he claimed that methods of factoring large composites had
- : been developed that were several orders of magnitude faster than older
- : methods. I told him that this was irrelevant, since you could generated
- : arbitrarily large composites to form your key. Even going 12 orders of
- : magnitude faster (1 trillion times faster), it could still take easily
- : several billion years to factor some of the smaller sized keys. The
- : question that I have is: am I completely wrong here (ie, have I stuck
- : my foot and a good portion of my leg in my mouth ;) ? Personally, I feel
-
- If in fact there was a fundamental breakthrough in factoring it could make
- the size of the composite irrelevant but I am sure we would know about
- that.
-
- : that the NSA posted that flyer to get people to refrain from using something
- : they are unable to crack ;>
-
- Not even the NSA is that stupid (I hope :-)
-
- :
- : My second question is: what is the best means available for generating
- : truly huge primes (200+ digits)? Fermat and Mercene (sp?) numbers
- : are NOT secure.
-
- Generating big random numbers and then testing them probabilistically for
- primality. Sounds crude but it works. The density of primes up in the
- range desired is higher than intuition would lead one to believe. I think
- that 100 attempts is larger than what is encountered on the average (and
- I'll be corrected if not :-) The probibalistic test can be iterated
- for an arbitrarilarily high theoretical probability that one is prime.
- Well, I know that one is prime, you understand what I mean. :-)
-
- I am less sure of this but I think that if the probabilistic test for
- primality actually yields a composite then the product of this with a
- real prime is about as hard to factor as a two prime composite. The
- probability of getting two such composites is exceedingly small.
-
- Bob
- --
- Bob Cain rcain@netcom.com 408-358-2007
-
- 'The meek shall inherit the earth--the rest of us will move on..'
- Sameer Parekh
-
-
- PGP 1.0 or 2.0 public key available on request.
-