> :-)
  > 
  > Sorry if my note seemed condescending or something... I
  > didn't mean it that way.

=> oh no I didn't interpret it that way, I just wanted to
make sure that we talk about the same thing.

  > >The point was: using tunnelling to evade scope boundaries.
  > >This can be done in a zillion ways for different addresses,
  > >if we're not careful how the tunnel is setup or if a malicious
  > >node is inside the site and can fool the firewall (if one exists).
  > 
  > I completely agree...
  > 
  > The current discussion is (I think) attempting to compare the
  > security implications of two ways of addressing a private
  > network:
  > 
  >          - Using site-local addresses that are filtered at
  >                  an SBR due to site configuration.
  >          - Using global addresses that are filtered at the
  >                  border of the private address space using
  >                  routing filters.
  > 
  > In neither case would the addresses being used on the private
  > network appear in global routing tables.
  > 
  > I still don't really understand why these cases are substantially
  > different in terms of security... 

=> I think that if you put Tony Hain's email together 
with Rich Draves' last email, you would get a good
idea about why it's better and what the "default"
should be. I think they both make sense. 

For a host, I think that the default should be 
1 site per interface. 

Not that I'm claiming to be
  > a security expert or anything, but I would like to understand
  > what the difference are, if any.

=> Sure, I think most of the mails on this thread are
clarifying this. 

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