I read today that the DNS bosses want to start selling top-level-domains, which
is funny because I had a similar idea and was thinking about it yesterday!
I don't want to sell them though, but make them availble for free, with
conflicts resolved by the community (wiki/forum style), and decisions being
based on merit, not money.
I started this project ages ago, but didn't put much effort in to it.
Now might be a good time to continue it, before the DNS bosses do introduce this
new money-making scheme.
For example, my nickname online is usually 'sswam' or 'ssw4m'. This is fairly
unusual, so with freens, I could lay claim to 'sswam' as a top-level domain,
or 'ssw4m' if there is contention and I lost out!
A domain like 'fred' would probably go to the wikipedia page about that common
name, not to any particular person called 'fred' who happened to get in first.
Something like 'mcdonalds' would go to that famous franchise,
because that's the common binding for that name.
Technically, it's simple to set up alternative DNS root servers, which defer to
the offical root servers in case a domain is not found. freens would start off
like an overlay over the normal DNS namespace. We wouldn't touch .net .com
.org, etc; they can remain 'paid for' addresses. Everything else is open for
free use as of yet. It would only work for people who 'opt in' by using our
nameservers, that's okay.
For the purposes of email addresses, every top level domain e.g. ssw4m can also
be found at ssw4m.free within the system, or under some other domain outside
the system (a sub-domain under some regular domain we have).
Does anyone think this is interesting / worthwhile? or not?
Currently I use bind9 for my nameservers, but I would also like to try djbdns,
I suppose it is technically better, and it's simpler to configure.
I use simple text files to configure dns at the moment, via a script,
and I would continue that. Here's an example of one for the 'ai.ki' domain,
I trimmed some lines:
The price hasn't been fixed, but will be upwards of $100,000. Still
> You received this message because you are subscribed to the Google
> Groups "mxug" group.
> To post to this group, send an email to mx...@googlegroups.com.
> To unsubscribe from this group, send email to mxug+uns...@googlegroups.com
> For more options, visit this group at http://groups.google.com/group/mxug?hl=en-GB
I think it's interesting, but also an old idea, and I don't think it's
particularly worthwhile. I think if we were going to get around DNS,
we could do something more extravagent. I mean, DNS is really just a
very direct search for the name. So, it could technically be replaced
with a more featured search engine. But it probably needs more thought
than that. I guess I find the limited-ness of domains a little bit
arbritary and perhaps unneccessary. A non-unique system might be
better (but perhaps not practical).
I do think it would be "fun" to have a independent dns system though,
where a group of friends have their own mappings set up. But of
course, this is exactly what you do when you run an internal network.
And I vaguely recall services available years ago that did similar
things, creating domains like '.business' and so on (not .biz). Free
to join/etc (obviously just configure your lookup to there).
I also think the security of DNS is interesting, through DNSSEC.
So in review, I think it's "interesting" but to be worthwhile needs
"more thought". Just my opinion.
Noon Silk | http://www.mirios.com.au:8081 >
Fancy a quantum lunch?
"Every morning when I wake up, I experience an exquisite joy — the joy
of being this signature."
I think you misunderstand my intention.
I don't intend to buy TLDs, I intend to TAKE OVER the DNS system,
making such domains available to people for free
- based on merit and consensus, not cash.
It wouldn't be a hack, rather a choice; people can choose to point their
resolvers and nameservers at my (our) DNS servers rather than the regular
Our DNS servers can then provide additional names over the top of the regular
We could, but I think this makes it much more difficult to get anyone to use
it. At the moment with freens, all they would have to do is change a few
settings in their resolver or nameserver config (the upstream nameservers).
Perhaps we could look at improving or replacing the protocol after we've
already taken over the world! Or, it could be done I guess using a 'local
nameserver program' which they could download. It would speak DNS and also our
new protocol whatever that might be.
I think simplicity is a good thing; for me, DNS is not simple enough!
> A non-unique system might be better (but perhaps not practical).
Yes it might be interesting, if everyone's view of the world through DNS were
somewhat different, however it might be weird or dangerous too! There would
have to be a particular 'variable' scope I think. random extra TLDs might be a
good choice for such a scope, I don't know.
> I do think it would be "fun" to have a independent dns system though,
> where a group of friends have their own mappings set up. But of
> course, this is exactly what you do when you run an internal network.
yes it'd be fun. In a LAN, you don't normally do things like making 'mxug'
point to our website (we'd need more than DNS, also to go via an http redirect
as things stand!)
> And I vaguely recall services available years ago that did similar
> things, creating domains like '.business' and so on (not .biz). Free
> to join/etc (obviously just configure your lookup to there).
ok wow, I didn't know there was such a precident - nothing new under the sun!
I know I've done it before, anyway.
To make it work, and popular, it would have to be very easy to use and well
done. I suggest we could ship it along with a popular product suite - perhaps
something where people can set up their own website / a facebook-like system.
I have a domain waiting for that: fr1end.net (with a 1 for #1 - chosen in
advance for when it is!) Distrubuted / without ads.
> I also think the security of DNS is interesting, through DNSSEC.
Yes, I don't know much about that as of yet. It does seem important.
> So in review, I think it's "interesting" but to be worthwhile needs
> "more thought".
true - but I did give it some more thought already ;)
Extra helpful thought like yours is much appreciated, that's why I posted
of course, to get some feedback.
And by "more thought" i was kind of meaning "needs a working group of
people thinking for a few months about all the things it would need to
Your proposal is just custom-managed domain servers. Sure, it's
something. But my point was I think DNS is considered "broken" so,
instead of just re-implementing the central nameservers, try and "fix
it all". But of course, there are lots of proposals like that.
This wikipedia page:
<http://en.wikipedia.org/wiki/Alternative_DNS_root#Open_RSC> seems to
have some more prominant ones (the one I have in mind but can't
remember isn't on there, I don't think, so clearly this is just a
small subset of what has been tried).
Nevertheless. It would make for a fun discussion topic one night, though ...
I wouldn't attempt to do that right away, just stick with existing DNS for a
start else will have 0 users for sure! Then if ever get some users, can look
at making improvements.
I'll put you down for 'noonsilk' then.
End of days are upon us:
Noon Silk | http://dnoondt.wordpress.com/ >
Fancy a quantum lunch? http://groups.google.com/group/quantum-lunch?hl=en