Re: dns discovery for agenda? [Re: chairs and charter]

[Johan Ihren <johani@autonomica.se>]

Pekka Savola <pekkas@netcore.fi> writes:

> On 15 Mar 2003, Johan Ihren wrote:
> > I'm all in favour of talking about DNS discovery, that's a worthwhile
> > subject. However, to my mind the real issue is discovery mechanisms in
> > general, not only DNS. As long as we stay micro focused on "dns
> > discovery", "foo discovery" and "bar discovery" I fear that we'll keep
> > on finding individual mechanisms rather than a generic method.
> > 
> > Unfortunately the more generic discussion doesn't seem well suited for
> > DNSOP.
> 
> Indeed, this is a worry.
> 
> However, DNS could be argued to be sufficiently special.
> 
> For instance, I have run DHCPv4 here in the IETF network basically just to
> find the DNS servers.  If I wanted to run in IPv6-only operation, I'd have
> to configure the server manually, run DHCPv6 or whatever.  I'm not at all
> interested in configuring search paths, NTP, Foo, or Bar in "visiting"  
> networks.

That you're not interested in configuring ntp, foo and bar is fine.
But that is in no way an argument for a multitude of discovery
mechanisms. I still think that *one* generic discovery mechanism is
very much better than several.

Much better that you just configure your client not to import the
stuff you don't care about than inventing more mechanisms to support
every form of granularity that people can invent.

Let's keep it simple. 

Johan
#----------------------------------------------------------------------
# To unsubscribe, send a message to <dnsop-request@cafax.se>.