home *** CD-ROM | disk | FTP | other *** search
- Path: sparky!uunet!sun-barr!ames!agate!ucbvax!virtualnews.nyu.edu!brnstnd
- From: brnstnd@nyu.edu (Dan Bernstein)
- Newsgroups: sci.crypt
- Subject: Re: New record non-networked factorization of difficult number
- Message-ID: <28087.Jul2514.53.0492@virtualnews.nyu.edu>
- Date: 25 Jul 92 14:53:04 GMT
- References: <BrwFFH.CAt@cs.psu.edu> <1992Jul25.004256.4641@linus.mitre.org> <25367.Jul2504.27.2692@virtualnews.nyu.edu>
- Organization: IR
- Lines: 27
-
- In a reply to Bob Silverman I wrote:
- > (2^488 + 1)/257 had 146 digits, last I counted.
-
- Which proves once and for all that number theorists can't count. :-)
- The factorization is still the record non-networked factorization, of
- course, but (2^488 + 1)/257 has only 145 digits. While I'm at it let me
- correct my assertion that 2^512 + 1 is the record difficult number; the
- record difficult number is (2^512 + 1)/2424833, which has 148 digits.
-
- In case anyone's wondering, I was misled by the UNIX dc program, which
- acts like this:
-
- % dc
- 2 488^ 1+ 257/p
- 310960166879725296199878945847101390767175534833553546544725362515212\
- 1265127859917318281565697486165824574686253229943428928722855633467075\
- 526401
- 10 145^p
- 1000000000000000000000000000000000000000000000000000000000000000000000\
- 0000000000000000000000000000000000000000000000000000000000000000000000\
- 000000
-
- Notice how the first line of output is chopped off a digit early. This
- is what I get for counting digits with the help of a computer instead of
- by hand.
-
- ---Dan
-