On Sun, Jan 06, 2002 at 05:42:03PM +0100, Maarten Geurts wrote:
> a reply to :From: Goran Thyni <[EMAIL PROTECTED]>On Sat, Jan 05, 2002 at
> 02:38:17PM +0100, Maarten Geurts wrote:
> 
>  > -Does any node expect to be connectable to any node, or just a n number
>  > of nodes it found during the discovery of other nodes.
> 
> - Any node if possible.
> 
> please explain "if possible".

if reachable (within timeout)


> Ok, I want to solve this  with an other (an desgin of) transport type.
> "proxy/firewall.". best thing would be if i only have to change The
> connecting node and not do anything of tunneling. However in the protocol
> there has something to be changed that only the firewalled node can initiate
> the connection.
> 
> Timing out should be prevented since we know from the start it is firewalled
> and it makes no sence to try a conventional connect to this node if it has a
> non routable adress (10.* /192.168.*)

It did some investigations around this topic,
and describe what I thought was needed for an implementation.
see:
http://hawk.freenetproject.org/pipermail/devl/2001-December/009726.html

> I think the tunneling function should be a part of the final protocol.
Me too.
I just started too mess around with gcj compilation instead because I think the
resource demands a JVMs is a bigger problem.

happy hacking,
-- 
G�ran Thyni



_______________________________________________
freenet-tech mailing list
[EMAIL PROTECTED]
http://lists.freenetproject.org/mailman/listinfo/tech

Reply via email to