ICANN opens up Pandora's Box of new TLDs

Phil Regnauld regnauld at catpipe.net
Tue Jul 1 15:43:24 UTC 2008

Phil Regnauld (regnauld) writes:
> John Levine (johnl) writes:
> > d) 280
> # dig @f.root-servers.net axfr . | egrep 'IN[[:space:]]NS' | awk '{ print $1 }' | sort -u |wc -l
>      281

Interesting extract from a transcript of tICANN board meeting in Paris.
It doesn't say much about what was originally envisioned,
but sheds light on the considerations that were made.


[It's in the first statement from Dave Wodelet:]

I just think it's important for the public record to make some
comments about adding new gTLDs to the root. While conceptually I
agree and see the benefit to the community with adding more TLDs
to the root, there are still some concerns about how scalable in
the long term this will be.

How many can we truly support? Well, from the best guess we have,
and I do stress the word "guess," somewhere around 5,000 or so TLDs
seem to be realistic.

But how high can we actually go? We really don't know.  There are
both technical and administrative issue limits to the scaling. And
it looks like the administrative issues may be more limiting than
the technical ones.

Certainly, what we do now administratively will certainly need to
change to support even the 5,000 or so that I mentioned earlier.
So how many will we have to support? Well, if we just look at the
number of place names, there seems to be somewhere between 5 and 6
million place names in the world. And if every one of these wanted
a TLD, that might not be possible.

And the 5 to 6 million place names doesn't include the number of
commercial TLDs businesses may want, and this 5 to 6 million doesn't
include the vanity names people may want as well, nor does this 5
to 6 million include what we may need in the future for names of
planets, planetary colonies, which may, indeed, happen within the
life of our Internet.

So I am a bit concerned about spending our TLD name inheritance for
future generations of Internet users. As we know, everything has
limits, like IPv4. We all know that has a limit, and that's why
we're looking at IPv6.

More information about the NANOG mailing list