[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
RE: [wg-c] new TLDs
From a psychological perspective, that won't fly. There is just too much
value-add to the mnemonic/semantics angle. Applications developers aren't
going to give that up and end-users would shun the service if they did,
particularly if no one else is dropping that service usage.
This is a lot like the way BroadVision works. This is universally reviled.
> -----Original Message-----
> From: Karl Auerbach [mailto:karl@CaveBear.com]
> Sent: Thursday, December 23, 1999 10:48 AM
> To: Roeland M.J. Meyer
> Cc: firstname.lastname@example.org
> Subject: RE: [wg-c] new TLDs
> > > I'd suggest: a1001, a1002, a2003, ... z9999
> > >
> > > There's 25,000 of those. If we need more, I'd lengthen the
> > > numeric part.
> > Karl,
> > Then what would be the point? Why not simply use the IP
> address and drop the
> > DNS altogether?
> DNS names are stable, they don't change when IP addresses
> change due to
> network reconfigurations, dynamic address assignments, NAT, etc.
> Also, DNS names are beginning to be used as names of services
> -- with the
> actual traffic being directed to any of a number of servers, possibly
> servers that are located in very different locations and with very
> different IP addresses.
> It's the mnemonic part of DNS names that got us in all this
> hot water in
> the first instance. DNS names are simply not expressive
> enough to handle
> all the human semantics we want to paint onto them. That's
> why we need to
> step back and use search engines to dig through lists of (attribute,
> DNS-name) pairs and return lists of potential targets.
> Systems like e-mail already strongly use such search engines
> - we tend to
> call 'em personal mailing lists and they are often built into our own
> personal mail composition tools - which merely indicates that "search
> engine" doesn't mean something on the scale of an altavista
> or yahoo for