home *** CD-ROM | disk | FTP | other *** search
- Newsgroups: sci.crypt
- From: Owen.Lewis@purplet.demon.co.uk (Owen Lewis)
- Path: sparky!uunet!pipex!demon!purplet!Owen.Lewis
- Subject: Re: Zimmermann's response
- Organization: FidoNet node 2:252/305 - The Purple Tentacle, Reading
- Date: Tue, 12 Jan 1993 18:27:00 +0000
- Message-ID: <42.2B534925@purplet.demon.co.uk>
- Sender: usenet@demon.co.uk
- Lines: 25
-
- -=> Quoting Rob Stampfli to All <=-
-
- RS> It is not my intent to add fuel to any fire,
-
- Likewise.
-
- RS> Pgp seeds certain random number vectors by asking the
- RS> user to type some characters at the terminal and then measuring the
- RS> times between keystrokes.....it probably works well in a DOS environment.
-
- As some may have noticed, I don't type particularly well or fast (I thought),
- using two fingers and a thumb. However, several times, when experimenting
- with PGP2.1 key construction, I have received screen messages requiring me
- to add to the character stream as some earlier characters had been typed in
- too fast. I understand this to mean that the time interval from one
- character to the next was too small for the key generation program (under DOS)
- to use. Though noting the occurences, until Rob's post, I did not give them
- much further thought.
-
- Does anyone know what the limits on the time interval measurement are and
- what effect this will have on the random number vectors?
-
- Owen
- ___ Blue Wave/QWK v2.11
-
-