Henry W Miller
mill0440 at gold.tc.umn.edu
Fri Jun 13 11:55:05 EDT 1997
On Fri, 13 Jun 1997, Adam B. Roach wrote:
> First, it may be a good idea to modify some of the clients
> so that, when they are unable to contact the key server
> (which happens a lot with my machine, since my dialup provider
> is flaky), instead of going idle, they start testing random
> keys. It's not going to exhaust the keyspace any faster, but
> it does give an outside chance of "tripping" on the answer.
It's been considered and droped more then once. get a buffering clinet
if this is a problem for you.
New clients are in testing, with this feature. IF they work right we
will release them.
> Second, you may want to set up some proxies that communicate
> with HTTP. I know this sounds silly at first, but the company
> I work for (and I know several other companies do the same thing)
> have a non-SOCKS firewall in place, with a web proxy available.
> Since the amount of communication with the server is minimal,
> the slowdown effected by going through a web proxy would be
> insignificant. I could add a ton of resources, including a few
> high-end multiprocessor machines, to the soup if you were to set
> something like that up.
I belive the latest mac client does this. This is a feature of the new
To unsubscribe, send email to majordomo at llamas.net with 'unsubscribe rc5' in the body.
More information about the rc5