tldns.com
Jim Kingdon
The reason I ask is that they seem to have gotten a small amount of
attention on dmoz.org (see
http://www.azport.com/odp/newsletter/7.shtml - there is a URL there
for a forum, but that is only for dmoz editors, I just browsed it and
it was basically people who were clueless about the difference between
alternate DNS and legacy DNS - there was a number about how tldns.com
apparently registered something like 50,000 domains in their first
day).
I guess my first reaction is that it seems less promising than OpenRSC
or TINC, because they (apparently) want to own the root as a
for-their-own-profit thing (some of us think that is OK for a top
level domain, but I don't think I've heard anyone say it is OK for the
root). Perhaps someone else has a comment?
Jim Kingdon
> I've heard anyone say it is OK for the root
Er, I think I worded that imprecisely.
s/OK/something that someone would want to point their nameserver at/
But the other interesting thing I noticed since my previous post is
that they are trying to get people to put this into resolv.conf rather
than into a named's root.cache. This is doomed (in the sense that
they'll need to lay in some serious bandwidth/CPU if they hope to keep
up - and I don't see where the $$$ would come from) but also cool (in
the sense that it lets individuals play, rather than just telling
individuals "bug your ISP" like some previous alternate DNS efforts
have done). http://www.tldns.com/download/faq.htm
I don't know, I'm starting to get intrigued by tldns.com. You see,
some people think of the net as something very fragile which must be
protected against people who are trying to fragment the DNS. I see
the net as something very resilient which is is willing to let people
have a shot at new protocols, organizations, and the like - with a
shrug and a "well, we'll see if they can make it work".
Russ Allbery
> I don't know, I'm starting to get intrigued by tldns.com. You see, some
> people think of the net as something very fragile which must be
> protected against people who are trying to fragment the DNS. I see the
> net as something very resilient which is is willing to let people have a
> shot at new protocols, organizations, and the like - with a shrug and a
> "well, we'll see if they can make it work".
Microsoft has a huge investment in people being able to put
www.microsoft.com in their browser and always get back the same site. I
personally could care less if they get Microsoft's site and see no
particular reason why they should have to. I think this is why I'm not
nearly has hung up about DNS staying consistent.
Inconsistent DNS is a self-correcting annoyance as far as I'm concerned.
But I don't have billions of dollars at stake in trying to shove a
particular URL in everyone's face.
--
Russ Allbery (r...@stanford.edu) <URL:http://www.eyrie.org/~eagle/>
Richard J. Sexton
Sigh. It's name.space without the attitude. Kind of
a flat namespace. Robin dropped ito orsc for
a while and I think we got the idea of a heirarchy across
to him. They're from the porn universe.
--
Richard Sexton | ric...@tangled.web | http://dns.vrx.net/tech/rootzone
http://killifish.vrx.net http://www.mbz.org http://lists.aquaria.net
Bannockburn, Ontario, Canada, 70 & 72 280SE, 83 300SD +1 (613) 473-1719
Jim Kingdon
> But I don't have billions of dollars at stake in trying to shove a
> particular URL in everyone's face.
Well, Red Hat does have quite a bit at stake (although I don't have as
much at stake personally as those who are called out by name on the
Red Hat Wealth Monitor :-)).
I guess the difference in mindset is between people who are control
freaks (who always live in fear that their leading position will be
taken away and try to beat back anyone who might be able to do it) and
people who like to think that their leading position has deeper roots
than that, and that the key to maintaining it is to, well, lead.
Jim Kingdon
> a flat namespace. Robin dropped ito orsc for
> a while and I think we got the idea of a heirarchy across
> to him.
Hmm, interesting.
Namespaces want to be flat, because that is the way the human brain
works. The following might not be the best example, but I was just
looking for "Red Hat Wealth Monitor" in a bunch of search engines
(without success, somewhat to my surprise). Did I remember anything
about where it was hosted or anything like that? No, just the name
(and even that I might have slightly garbled). This is what drives
the everything-in-.COM phenomenon - even a relatively good web
directory like google or dmoz doesn't take away the desire for
memorable names.
Namespaces want to be hierarchical because flat namespaces run out of
names. So do you auction off the good names? Or first come first
serve (which can end up being somewhat the same thing)?
> They're from the porn universe.
A group which is much underrepresented in internet governance,
compared with their percentage of the net
traffic/sites/names/whatever ;-).
Greg Skinner
Russ Allbery <r...@stanford.edu> wrote:
>Microsoft has a huge investment in people being able to put
>www.microsoft.com in their browser and always get back the same site. I
>personally could care less if they get Microsoft's site and see no
>particular reason why they should have to. I think this is why I'm not
>nearly has hung up about DNS staying consistent.
Hmmm ... what is your feeling on whether or not people are able to get
consistent results to www.stanford.edu?
(Not that tldns.com is likely to interfere with Stanford, but you get
the point.)
--gregbo
gds at best.com
Richard J. Sexton
They figured out they have to refer com/net/org and we were able
to convince them to point to a few other tlds, too.
The rest is flat though.
Russ Allbery
>> Microsoft has a huge investment in people being able to put
>> www.microsoft.com in their browser and always get back the same site.
>> I personally could care less if they get Microsoft's site and see no
>> particular reason why they should have to. I think this is why I'm not
>> nearly has hung up about DNS staying consistent.
> Hmmm ... what is your feeling on whether or not people are able to get
> consistent results to www.stanford.edu?
The intended target audience of www.stanford.edu is mostly a mystery to
me, and if people were looking for particular research I'd think they'd go
via a search engine rather than poking around from our home page. I think
most of the people who read the news reports and whatnot that are posted
there are actually already at Stanford.
So I think my feelings are about the same.
If people's search engine results point them at completely different sites
than the search engine sees, they'll get annoyed, and they'll find ways to
fix it. Self-correcting.
Brian Palmer
> Greg Skinner <g...@nospam.best.com> writes:
> The intended target audience of www.stanford.edu is mostly a mystery to
> me,
Prospective students/faculty?
> and if people were looking for particular research I'd think they'd go
> via a search engine rather than poking around from our home page. I think
> most of the people who read the news reports and whatnot that are posted
> there are actually already at Stanford.
>
> So I think my feelings are about the same.
>
> If people's search engine results point them at completely different sites
> than the search engine sees, they'll get annoyed, and they'll find ways to
> fix it. Self-correcting.
But what if they don't know there's a problem? Say, for instance,
there's a DNS server which points to my machine as amazon.com. I act
as a proxy between the user and amazon.com --- but I also happen to
log all credit card numbers used on my machine.
Hmm... sure, cryptography could come into play there, but I could do
similarly confuse issues things by lifting content and just replacing
ads.
--
Brian Palmer
"Every nonzero finite dimensional inner product space has an orthonormal
basis. It makes sense, when you don't think about it."
Russ Allbery
> Russ Allbery <r...@stanford.edu> writes:
>> The intended target audience of www.stanford.edu is mostly a mystery to
>> me,
> Prospective students/faculty?
Maybe, to some degree. But people are still going to apply to Stanford
without us trying to push our name in their face. I think this goes to
Jim's point about market share based on marketing vs. market share based
on doing something well.
> But what if they don't know there's a problem? Say, for instance,
> there's a DNS server which points to my machine as amazon.com. I act as
> a proxy between the user and amazon.com --- but I also happen to log all
> credit card numbers used on my machine.
Don't point your root DNS at servers you don't trust. *shrug* I don't
see a security issue in allowing choices; obviously, you don't use a root
server that will let people do things like that. If you do, you're only
hurting yourself and the people you provide DNS resolution services for.
Greg Skinner
Russ Allbery <r...@stanford.edu> wrote:
>Brian Palmer <bpa...@leland.stanford.edu> writes:
>> Russ Allbery <r...@stanford.edu> writes:
>>> The intended target audience of www.stanford.edu is mostly a mystery to
>>> me,
>> Prospective students/faculty?
Not only those. The current student/faculty/staff body, tourists,
publishers of college guides, and many other groups, I would imagine,
have an interest in www.stanford.edu taking them to Stanford
University's home page (particularly because it has done so for quite
some time). I imagine most search engines have indexed Stanford
University's home page.
>Maybe, to some degree. But people are still going to apply to Stanford
>without us trying to push our name in their face. I think this goes to
>Jim's point about market share based on marketing vs. market share based
>on doing something well.
People are still going to buy Microsoft products despite Microsoft
pushing their name in everyone's face. Whether or not Stanford or
Microsoft "do something well" (whatever that means) is beside the point.
If people *don't* get information reliably from alternative providers,
they will just turn to those sources that do provide it reliably. The
existence of a market for these services guarantees that it will get
provided reliably.
>Don't point your root DNS at servers you don't trust. *shrug* I don't
>see a security issue in allowing choices; obviously, you don't use a root
>server that will let people do things like that. If you do, you're only
>hurting yourself and the people you provide DNS resolution services for.
This is the rub. The alternative DNS providers need to gain the trust
of the Internet community. For what it's worth, I don't think that
alternative providers are going to lie (that would not be in their
best interest). However, there is likely to be concern that
alternative providers are more likely to have their security
compromised. Denial of service attacks are also a potential problem.
--gregbo
gds at best.com
Kai Henningsen
> Namespaces want to be flat, because that is the way the human brain
> works.
Say what?
>The following might not be the best example, but I was just
> looking for "Red Hat Wealth Monitor" in a bunch of search engines
> (without success, somewhat to my surprise). Did I remember anything
> about where it was hosted or anything like that? No, just the name
> (and even that I might have slightly garbled). This is what drives
> the everything-in-.COM phenomenon - even a relatively good web
> directory like google or dmoz doesn't take away the desire for
> memorable names.
That's what's good for a _search engine_. That's not how you _remember_
stuff.
> Namespaces want to be hierarchical because
... human brains are wired to understand hierarchy. Witness human
societies all over history.
Kai
--
http://www.westfalen.de/private/khms/
"... by God I *KNOW* what this network is for, and you can't have it."
- Russ Allbery (r...@stanford.edu)