>
> => Well IMHO, the default router is not a server. 
> For a node to communicate with other nodes outside
> the link, it requires a router. Hence, the dependency
> is already there (on the default router). There is 
> no other node involved. We can discuss the terminology,
> but the point is, there is no reliance on a third
> node involved. 

so as long as the piece of code runs on a (default) router, the
requirement for server-lessness is fulfilled? e.g this could be a DHCP
server?


=> OK so there are lots of word games going on :)
I wonder how consistent this scrutiny of proposals
is. Anyhow, it is good to eb precise so let's 
continue. 
The answer to your question is in the question
itself. You said:

' requirement for server-lessness is fulfilled? e.g this could be a DHCP
server?' 

If it's a server, then how could you call the solution
server-less? 
Also the IETF does not mandate architectures, we can't
force people to install DHCP servers on each router. 
Even if we can, what are the consequences for management
cost of running multiple servers? Is it worth it?

Hesham



--------------------------------------------------------------------
IETF IPng Working Group Mailing List
IPng Home Page:                      http://playground.sun.com/ipng
FTP archive:                      ftp://playground.sun.com/pub/ipng
Direct all administrative requests to [EMAIL PROTECTED]
--------------------------------------------------------------------

Reply via email to