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

John Campbell jcampbel at lynn.ci-n.com
Mon Jun 29 19:07:39 EDT 1998


On Mon, 29 Jun 1998, ALBERTY Pascal wrote:

> Of course I'll do it! But what about people who'll forget to make these
> changes. Or people who'll keep intentionnaly blocks to slow down the project
> ?
> 
	People who intentionally keep blocks will affect d.net's overall
performance by no more than they would by not participating at all. (Well,
plus the negligible amount of time it takes for the master to hand the block
to someone else, assuming that the contest hasn't been won by the time it
gets around to handing that block out again.) 

> For this kind of challenge (with a short time completion estimated), is a
> 'time
> out' for block a good solution (time out handle by key servers)?
> 
	There's already a time-out built into the system. When the key
servers run out of blocks, they start over handing out the ones that have
been handed out but not returned yet, so the time-out period is not
something pre-determined arbitrarily, but the precise amount of time until
the server is required to hand the block out again. This can cause some
duplication of work from clients with pathologically long fetch/flush
cycles, but I can't think of any real way to avoid that, except to shorten
the fetch/flush cycle.

---
John Campbell
jcampbel at lynn.ci-n.com

QotD:  Just go with the flow control, roll with the crunches, and, when you
get a prompt, type like hell.
--
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