[RC5] Time out and DES-II-2 challenge

Joe Zbiciak j-zbiciak1 at ti.com
Sat Jun 27 01:37:59 EDT 1998

'ALBERTY Pascal' said previously:

| Imagine now that I do the same with DES-II-2 challenge ?
| I got several blocks on my very very old and slow computer and
| (my god !) I got THE key, but my computer take more than
| 10 days to find it.
| Is there a solution for slow computer keeping keys to much time ?

You could do one of the following:

-- Buffer the smallest size block, and always take your "buff-in.des"
   file with you to work and flush it every day.  (For 10 days, that
   shouldn't be a huge headache).  If you pick a 2^28 block-size,
   instead of the default 2^30th, you will receive blocks to work on in
   more manageable doses.  Also, you'll want to buffer no more than you
   can process in about a day.

-- Alternately, you could just leave the 486 out of the DES contest,
   and let it keep working on RC5.  

The DES-II contest really needs short block latency in order to be
effective, since we will probably start reissuing keyspace within a few
(5-6 days) of starting it, if all goes well.  At that point, a block
latency of as small as one day could kill us by causing oodles of
duplicated work.  

Ironically, according to the simulations I've run, the LIFO buffer
arrangement we're using actually helps block latency, and reduce
duplicate work during the "keyspace thrash" period.  The reason appears
to be this:  The blocks at the bottom of the stack get put there almost
immediately, when the client does its first fetch.   Subsequently, if
the client is configured/managed in such a way that it rarely empties
its buff-in completely (thus causing the blocks at the bottom to become
"stale"), then the likelihood of these blocks ever being processed is
fairly low.  What happens, then, is the active portion of buff-in tends
to always have "fresh" blocks, and the "stale" blocks at the bottom
simply never (or extremely rarely) get touched.  Then the key-server
reissues these blocks later, and someone else processes them.
Essentially, the LIFO arrangement causes all buff-in activity to be
concentrated on the most recently grabbed (eg. freshest) blocks, which
is really what's important in a time-limited contest.



  +------- Joseph Zbiciak ------+
  |- - - j-zbiciak1 at ti.com - - -|  "The ability to quote is a serviceable
  | -Texas Instruments, Dallas- |   substitute for wit."
  |- - #include <disclaim.h> - -|                  -- W. Somerset Maugham
To unsubscribe, send 'unsubscribe rc5' to majordomo at lists.distributed.net
rc5-digest subscribers replace rc5 with rc5-digest

More information about the rc5 mailing list