[RC5] C DES core

Mary Conner trif at serv.net
Thu Jan 22 13:24:07 EST 1998


Given some of the reactions that I have seen to the released x86 clients,
releasing a functional but slow C DES core may have backfired.  I have
seen people saying that they're going to put their cyrixes back on rc5
because the cyrix no longer outperforms the pentium, and this would hurt
their stats.  Even where we had fully functional clients available,
switchover to des was poor (even a week later), and I cannot tell you how
aggravating it was to have people asking how to set up their new rc5des
client to run rc5 instead of des, because, "doing DES will take away from
from my rc5 stats, I don't want to lose ground by doing DES."  Yet,
they'll take the rc5des client because it has a slightly faster rc5 core. 
We have deschall clients for many machines, and I really appreciate the
people who are using them, and I know that many people can't, but I'm
hearing many people who certainly could, but they won't, because the
deschall clients won't do anything for their stats.

Given this, where should our effort go?  In producing a client that would
run at 1/20th the speed people were expecting (doing this wouldn't be for
free, someone would have had to spend time on this instead of something
else), or in what will be the only thing that will get most people to
switch, and that is a client that will report to stats and will be as
fast as they are expecting?  I estimate that a generic DES core would
have got us 5% people committed to doing DES with a client that would
produce crappy stats.  The vast majority of people participating are not
like the people on this mailing list, they run for stats and stats only
and could care less for the overall goal.

> Let's not forget that as well as a crypto problem (crack as many DES
> blocks as possible per day) d.net also has a social engineering problem:
> lots of clients that were ready and willing to crack DES have been left
> out in the cold for a long time. This doesn't build enthusiasm and
> commitment to the cause!

The cold hard truth is that most of the people out there were recruited
with and care about one thing---STATS, and they have zero commitment
to the d.net cause.  That is the largest social engineering problem,
it's not going to go away, and if we want rapid switchover on time
critical projects like this in the future, this factor has to be
addressed.

--
To unsubcribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest



More information about the rc5 mailing list