Thousand of messages are posted in this forum, because someone in the 80's tried
to build up a naming convention for IP adresses in an 8 bit university mostly US
network.
Today we are entering a 64 bits world wide network.The first
question is: was it a good choice?
Second question is: what is the alternative?
The choice was good enough for 20 years, but discussions
show today it is not perfect. Two possibilities: to improve it or to change it. Everyone
in here, even the most active are just talking about patches because they are afraid
of changing things too much.
I believe in the other way: to restart from zero and
see if we can cope with the output. Oddily enough we will do it more easily than
to patch the existing system. The reason why is simple: we are talking about a method
to name human, machine or virtual bodies identified through an ID NR (IP addresses).
Our problem is not a "more TLD" problem, our problem is a "more user problem".
The good naming concept for a medium size data base can be patched into a poor
naming concept for a large data base, or to be replaced by a good naming concept
for a large data base.
The exsiting system corresponds to a standard b-tree, with
a few leaves (TLDs)and was helpfull to university users round the world. Today we
need an object (TLD)oriented modern access system, in tune with the real moving life.
We have such a system which works pretty well for centuries: names.
First name,
last names, nick names, tags on suitcases, cars or towels.
Why not to consider
it?
We will have billion of users with a need for one or several domain names for
their smartphone, smarthome, smartcar, etc.. needing from 10 to hundreds of tags
for their mailboxes, washing machine, home work, virtual caddies, etc. and myriads
of IPv6 IP addresses.
Will a TLD or 100 TLD more realy help?
NO.
What we
have to do is :
- to decide that TLDs are like names, domain names like forenames,
etc...
- to comply with industrial property rules which have decided that TLD
are not part of a trade mark. i.e. TLD cannot be registered.
- to help TLDs developping
as roman "gens", tribues, communities with their own rules. Why the US Gov. would
be indirecty decide the way people call themselves in China or in Moldavia. ARPAnet
was an american network, but the protocol is French, the Web European, and the users
from evey countries.
Is this Utopia? I think limitating the world population to
250 TLD (family names)is Utopia. Let see some of the objections.
- it will
be a revolution! No. To the countrary: the DNS will not change. Only its vision.
New development avenues will be open. I will only know we are not blocked any more,
but if we do not want to change anything, or open a few TLD more: no problem. Insteaf
of a patch to nowhere, we will have a vision and a strategy.
- it will be
a mess. The same way as a telephone directory is a mess with tens of thousands of
different family and business names.
- it will be unmanageable. I do not
see why: the larger the number of TLD, the lesser the size of the files. IPv6 leads
to very large routing data bases anyway.
- you could not move back. Wrong.
Existing TLD will increase values as the top "clan". We can easely collapse new TLDs
in a super TLD if needed: all the new TLDs, like ".abc", will become ".abc.tld" as
there are ".com.uk" or ".gov.fr".
- the start-up will be awfull. I do not
think so : there will be no gold rush if cybersquatters know there will be thousands
of so called gold sushes.
- there will be no financial warranty. No more
than when you buy a house or an icecream. Some TLD will become more valuable because
of these warranties.
- what about cost? You do not pay for your name every
year. Unless ICANN wants to create the world wide tax, the internet name should be
free as belonging to people or companies from birth, creation or ias registration
to death or business end. Many organizations will want your name to be registered,
starting with credit cards and banks. Why to pay for it. Todays sites give you your
domain name for free if they host you.
- what a fight for ontaining TLDs!
I do not believe: ICANN asked who wanted to start operating new TLDs. The response
is very limited. The reason is that you must organize, invest, make known. I suppose
that progressively TLD management softwares and hosting services will develop in
different langagues. You know what: it might be the t-shirt syndrom. Today Kodak
says "no more TLDs where we will have to protect the Kodak name". Times come when
they will set-up and giveaway URLs built with the ".kodak" TLD!
- nobody
will be able to manage! True. Is there a naming authority to give a name to every
human beeing. There is one (CCITT) to give them (indirectly) telephone numbers. IP
adresses have to be regulated, not the way friends call me. Actually, people will
have aliases for the same IP address. Many of them. As they have nicknames.
-
this will never happen! Sorry, it is already here. Just put 4 figures into your Windows
parameter set to get your name service from an altenative source, and new TLDs. I
am sure soon a virus will do the job for you giving you access to ".hack", ".pirate"
etc.. TLDs.
- if this is true, it will be the end of the internet with
10 different managements authorities for the ".biz" TLD. Right. This is why ICANN
should not discuss "ifs" but "hows", not "legalities" but "practicalities" and not
fees, rates, discounts, trade marks but cooperative management with private organizations
from all over the world.
- OK, but why to keep TLDs and not to allow http://kodak?
For a simple reason: this is not the way life is. There are Mr. John Kodak around
and several trademark classes. Keeping at least a dot will please the "dot com company"
and help URL reading.
- your scheme is crazy: and you are the only one to
propose it! There is a phrase which says if you are alone of your opinion you are
most probably right, because the world is not 99.99% wise.
- at least for
it to stay understandable let keep it in English. OK if "Smith" is English in http://john.smith!