> > ...  in other words, we need to abandon this idea
> > that a host can have an arbitrary number of addresses with
> > varying scopes and lifetimes and connectivity,
> > and it's up to hosts/apps to sort out which one is best to
> > use. ...
> 
> We have heard you say this many times over, but you are refusing to hear
> that other people have figured out how to use them for their specific
> app or topology. 

Just because some apps happen to work with SLs or are easily modified
to work with SLs does not mean that SLs are either harmless or justified.

> Since you have identified the case that a multi-party
> app can't do the right thing when the name/address resolution is only
> capable of a single scope, you should write a BCP to that effect.

I have no interest in describing how to impair apps to deal with an 
impaired network.  I find it far saner, and hopefully more productive,
to keep the network from being impaired in the first place.

> Otherwise your argument ammounts to a demand that we not allow doing
> more than what we always have done, simply because it is easier.

No, I'm arguing for a more flexible Internet, one that can support
a wider range of applications with less complexity.  more warts does 
not equate to more flexibility,  and SL is a good example of a wart 
that decreases flexibility.

Keith
--------------------------------------------------------------------
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