<html><head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
</head><body style="background-color: rgb(255, 255, 255); color: rgb(51,
51, 51); background-color: rgb(255, 255, 255); color: rgb(51,
51, 51); font-family: Helvetica Neue; font-size: 14pt;"
bgcolor="#FFFFFF" text="#333333"><div style="font-size:
14pt;font-family: Helvetica Neue;"><span style="font-family: Helvetica
Neue;"></span><br>
<br>
<span>Karl Auerbach wrote:</span><br>
<blockquote cite="mid:56744DEB.3000001@cavebear.com" type="cite"><meta
content="text/html; charset=UTF-8" http-equiv="Content-Type">
<br>
<div class="moz-cite-prefix">On 12/18/15 2:33 AM, Christian de
Larrinaga wrote:<br>
</div>
<blockquote cite="mid:5673E0EB.3030100@firsthand.net" type="cite"><meta
content="text/html; charset=UTF-8" http-equiv="Content-Type">
<div style="font-size: 14pt;font-family: Helvetica Neue;"><span
style="font-family: Helvetica Neue;">Actually DNS is not
working for most of the Internet either, witness we don't have
names resolving to the billions and approaching trillions of
devices and applications services at the edge of data
networks. <br>
</span></div></blockquote>
I've never heard that claim before. I've run experiments with DNS
and found surprisingly few limits on how far it can expand. (For
example, in one experiment [more than a decade ago] we ran Bind with
tens of millions of top level domains and then ran query traffic [in
which we mixed a fair amount of absent names to make it more
real-life.])<br>
<br></blockquote>
<br>
I'm intrigued. Was this done to
establish evidence that a flattening of the hierarchy would not be a
technical problem? Took about thirty years for that shift in
architecture of DNS to come out of the cold. <br>
<br>
I am really referring to the scaling of DNS beyond server side hosts
which are now largely in located in data centres to satisfy the
need for persistent identifiers for all our devices and services. That
has not
happened using DNS registry business model as it has developed and
managed at ICANN. I dare say it could have happened technically. But the
business model doesn't work out to charge $10 or more per an for a
device orientated name service. The DNS has been taken over by those
using it as a pseudo business registration service. A role that the DNS
is bound to fail in satisfying. <br><br>
Incidentally I am not knocking the work that Jon Postel and Paul
Mockapetris started back in 82 ish and many others have done some
amazing work on DNS which we all depend on today. But it seems to have
gone as far as it can. <br>
<br>
<br>
<blockquote cite="mid:56744DEB.3000001@cavebear.com" type="cite"><blockquote
cite="mid:5673E0EB.3030100@firsthand.net" type="cite"><div
style="font-size: 14pt;font-family: Helvetica Neue;"><span
style="font-family: Helvetica Neue;">Sad fact is DNS designed
in an era of big iron...</span></div></blockquote>
DNS was designed in the mid 1980's, and the biggest of computers we
had back then are overmatched even by rather small devices of
today. The laptop I'm using to type this makes the Crays I used
(for magnetic confinement fusion simulations) seem rather weak.<br>
<br>
However, there is an intriguing side vector, which is that DNS is
fading as a user-visible technology.<br>
<br>
This does not mean that DNS is going to disappear, rather that it is
being submerged to become an internal internet name/address
technology. IP and MAC addresses used to be far more visible to
users. They became submerged under DNS names. DNS is now following
that path and being submerged under URI based names and
application-local names (such as Facebook names, hashtags, Twitter
handles, etc.) Even URI names that contain long DNS names and index
data are being submerged under shortened names. I anticipate that
attribute-based naming systems will come to dominate in certain
areas (I am sure, however, that if one were to look inside such
systems that DNS names will be there serving as internal machinery.)<br>
<br></blockquote>
<span style="font-family: Helvetica Neue;">I like your use of the word
"submergence" of DNS. It is a great way to put it. </span><br>
<br>
<blockquote cite="mid:56744DEB.3000001@cavebear.com" type="cite">
There is at least one of the new top level domain offerings that is
based on the idea that this kind of DNS submergence is happening.
It's (partial) focus is on DNS names used to located technical
resources; the human semantics of the names is not particularly
important because it isn't humans who are uttering those DNS names.
On the other hand, because a flexible human has been supplanted by
embedded firmware, the value of long term persistence of a DNS name
is more important than cute words that such a name might contain.<br></blockquote>
<br>
Persistence and global reach of identifiers are critical qualities for
many data applications. <br>
<br>
DNS is continuing to serve
as a naming service in the sense of being submerged
within a
grander URI schema such as with Handles or other registries. But there
is the likelihood of a different identifier model entirely appearing. <br>
<br>
There has been a lot of interesting work that might lead to persistent
identifier routing for data objects or graphs of semantic links to give
two examples. Both would be a move away from the "everything is a file"
Unix metaphor to address content which can lie within and across many
devices, even network boundaries. <br>
<br>
How far the DNS as it is currently structured can usefully serve in such
an environment I don't know but it is likely to become increasingly
"submerged" as you describe and I suspect increasingly routed around. <br>
<br>
<blockquote cite="mid:56744DEB.3000001@cavebear.com" type="cite">
--karl--<br>
<br></blockquote><pre wrap=""></pre>
<div class="moz-signature"><span style="font-family: monospace;">Christian</span><br>
</div>
</div></body></html>