In article <43o44t$hof@calum.csclub.uwaterloo.ca>, iagoldbe@calum.csclub.uwaterloo.ca (Ian Goldberg) writes: [ summary of Ian's day deleted ] Now imagine what my last 48 hours have been like. :-)
Holger.Reif@PrakInf.TU-Ilmenau.DE (Holger Reif ) was kind enough to verify that the SunOS 4.1.3 version of Netscape generates its keys in _exactly_ the same way as Solaris and HP-UX; he says he'll test other architectures tomorrow. I suspect any big-endian machine with the lrand48() function (which is used in key generation on Solaris/HP-UX; it's disguised in unssl.c as the macro mklcpr()) will be the same. Other Unix flavours should require only minor changes.
Most of the unix machines do the same thing. On SGI machines that have the hardware cycle counter, its value is used in place of the srand48(usec), lrand48() sequence. BSDI the code used srandom and random.
I'm still interested in what Windoze clients do (other than lose).
On windows and mac the first 32bit seed is seconds since 1970, and the second 32bit seed is the "tick count", which I'm told is the number of milliseconds since windows started. --Jeff -- Jeff Weinstein - Electronic Munitions Specialist Netscape Communication Corporation jsw@netscape.com - http://home.netscape.com/people/jsw Any opinions expressed above are mine.