Kuro5hin.org: technology and culture, from the trenches
create account | help/FAQ | contact | links | search | IRC | site news
[ Everything | Diaries | Technology | Science | Culture | Politics | Media | News | Internet | Op-Ed | Fiction | Meta | MLP ]
We need your support: buy an ad | premium membership

Feasability of alternative namespaces

By dsilverman in Internet
Fri Nov 17, 2000 at 09:23:26 AM EST
Tags: Politics (all tags)

ICANN is a failure. The organization was founded with a bad charter, has been abused by the "temporary" appointed members, and has too much control without enough public influence or involvement. The Corporation has held biased and geographically-oriented elections, and then ignored the results when they found them to be unacceptable. Does anyone else understand why the board has approved new gTLDs without the elected board deciding on them?

My proposal is to seriously consider what it would take to develop a viable and powerful alternative DNS system without fracturing the existing name space, and how to ease a transition to such a system. This proposal is not to abandon the existing system, but an idea for how we could go about creating an alternative system, and what the pitfalls of such a system would be.

A new namespace management system would need several components:
  1. Formation of a non-profit corporation. - "Domain Management Authority" sounds appropriate. I would suggest doing this in partnership with a major institution. MIT comes to mind. Harvard Law School's Berkman Center for Internet and Society is a very open-minded community as well. This would give the movement legitamacy. We should invite prominent members of the field of internet law, such as Lawrence Lessig, to be on our founding board, and to oversee the development of a constitution and bylaws.

  2. Establishment of a working paper - Using a collaborative medium like K5, we should create a set of bylaws and goals. Once again, the Berkman Center has experience in this - they have worked on open and collaborative brief writing through their OpenLaw list. Creation of an OpenDomain list or something similar would perhaps be useful. The community should work to form a concise, simple, and informative document, outlining what is wrong with the current DNS and how we will work to improve it. The constitution should have clear procedures for amendments and elections. Once approved by the community at large and the board, the constitution should be immediately implemented, and elections held.

  3. Transparancy - Immediately from the formation of the organization, all meetings and correspondance should be available, unaltered, on the corporation's web site. All meetings should be conducted either with audio streaming or through online chat, allowing anyone to listen in. The organization, being a true representative of the people, should be completely open and accessible. Anyone should be able to attend physical meetings, of such meetings occur, or participate in online meetings.

  4. Creation of alternative root name servers - Immediately the organization should establish a small network of root name servers, and should mirror the official name files. We should encourage people and ISPs to use our servers as an alternative to the real root servers, which are already bogged down. We should actively solicit grant monies to conduct research into web usability and domain space policy.

  5. Gradual name approval - Over time, the group should establish different gTLD policies. We must make sure not to conflict with ICANN's system. All gTLDs created through the Domain Management Authority ("dTLDs?") should be accessible by anyone on the internet through a DMA subdomain, and to all DMA users through the DMA root servers. All new TLDs should follow strict guidelines. dTLDs for non-profits, for example, should only be given out to real and registered non-profit organizations, as recognized by international law (The UN's "NGO" classification). Various alternatives to .com, etc. should be established, such as .geo and .sex. In this way, different systems for addressing internet space can be used. The goal is, after all, to test various new methods of addressing information, not just to embrace and extend the existing system.

  6. Petition the government - This is important. The DMA should petition the United States government for an official sanction to operate our services. In the spirit of international cooperation and world ownership of the Internet, we should call on the US Department of Commerce to grant DMA the same status as ICANN. We should first request simple recognization, and then solicit funds and access to the real root name servers. This should be handled by real experts in law and technology, not a geek uprising. We should follow legitimate channels, funding prominent internet lawyers through grant monies to lobby and present our case. Eventually, we should aim to usurp ICANN with a better system.
To coin a phrase from "The Death of a Salesman," ICANN is intelligent, but it is not well-liked. They do not keep the people's interests first. ICANN is not personable and respectable. They do not play Washington's game well. This leaves us a great opportunity for an organized political assault on the organization. More details of media campaigns and lobbying should be left for a more intimate discussion, but suffice to say, we should bring them down, Washington style. Only by playing the game of politics can we succeed in reforming the system. By receiving endorsements from major intellectuals and, possibly, members-elect of the ICANN board itself, and by acting quickly, we could, if not take over DNS control, at least bring DNS issues to the forefront and help to change the system.

A complete outline of what would need to occur is left for later. For now I ask simply, what do you think? How feasible is it? And hey, why the hell not?!


Voxel dot net
o Managed Hosting
o VoxCAST Content Delivery
o Raw Infrastructure


What do you think of an alternative DNS?
o It would fracture and seperate the Internet landscape. 18%
o It would alleviate overcrowding and ineffective bureaucracy. 44%
o It is simply not feasible. 13%
o Undecided 23%

Votes: 86
Results | Other Polls

Related Links
o Berkman Center for Internet and Society
o Lawrence Lessig
o embrace and extend
o Also by dsilverman

Display: Sort:
Feasability of alternative namespaces | 40 comments (37 topical, 3 editorial, 0 hidden)
opennic? (3.71 / 7) (#1)
by bgalehouse on Fri Nov 17, 2000 at 12:33:48 AM EST

Sounds like you want to have everbody go get involved with opennic.

Negative (3.00 / 3) (#4)
by dsilverman on Fri Nov 17, 2000 at 12:41:35 AM EST

The OpenNIC story has some good ideas for how to get wide acceptance of an alternative DNS. However, my proposal is to create a new governing body completely through community involvement, and to use the new DNS for experiments in alternative forms of addressing, instead of just throwing out hundreds of TLDs. We should be measured, calm, and intelligent in our decisions. We should reach consensus, and we should keep administration in-house. I don't want a for-profit managing my internet, I want a non-profit with no goals of world domination, which charges only what it needs to keep the system running.

"Speak softly, and carry a big stick."
[ Parent ]
I'm sorry, but... (none / 0) (#23)
by bgalehouse on Fri Nov 17, 2000 at 01:50:21 PM EST

I still don't really see the difference. You go on about a non-profit governing body making'measured, calm and intelligent decisions through consesus. I'm not real well acquainted with OpenNIC, but what I have seen seems to match these goals. Why are you not working with them?

If you differences are major, what are they? If your differences are minor, why aren't you presenting this as a direction that they could go in the future?

[ Parent ]

a better solution (2.42 / 7) (#2)
by enterfornone on Fri Nov 17, 2000 at 12:34:27 AM EST

I think the multiple tld's is defeating the purpose of having a domain name heirachy in the first place. We may as well scrap dns and go with something similar to AOL keywords on a first in basis.

A better way I think would be to have free domain names subject to a strict policy similar to the usenet naming system.

Your proposal BTW, is way too US centric.

efn 26/m/syd
Will sponsor new accounts for porn.
Of course it is (3.50 / 4) (#3)
by dsilverman on Fri Nov 17, 2000 at 12:37:58 AM EST

I purposely made my proposal very US-centric. This is because the United States currently has complete control over the DNS. Thus, we must work WITH Washington in order to achieve our goals, if these are indeed them. I believe that the eventual system should be completely decentralized, however I believe that an appeal to the US Commerce Dept. is the fastest and most effective route.

"Speak softly, and carry a big stick."
[ Parent ]
Re: Of course it is (3.50 / 4) (#12)
by dreamfish on Fri Nov 17, 2000 at 06:07:52 AM EST

The US government, when it passed over control the ICANN, ran away as quickly as possible so it wouldn't need to get involved with all the problems ICANN currently has. Not only that, they stressed that all future development of the Internet needs to be handled on a global basis.

I understand it would be politically expedient to aim this at Washington but you may end up alienating the rest of the world at the same time. A new proposal would be the best way of trying to create a system acceptable to all countries, especially if they were all involved from the start.

[ Parent ]

Then get involved with... (3.85 / 7) (#5)
by h0tr0d on Fri Nov 17, 2000 at 01:09:50 AM EST

OpenNIC. I really think that you need to look into OpenNIC some more. And if you have questions then go to this story at NerdPerfect and ask them because Robin Bandy has graciously agreed to answer posted questions.

-- It appears that my spleeing chucker isn't working again.

my personal thoughts (3.63 / 11) (#8)
by blaine on Fri Nov 17, 2000 at 01:52:35 AM EST

Honestly, I think the main reason we have so many problems with the DNS system is that there are no real enforced rules. Quite frankly, most companies just scoop up COMPANYNAME.* , and see nothing wrong with that.

What I would like to see done is a simple, feasible, and (IMHO) effective solution:

There should be only 5 TLDS:

  • .com : companies only, noone else may purchase them
  • .org : non-profit organizations only, noone else may purchase them
  • .net : ISPs only, noone else may purchase them
  • .xxx : pornography only, noone else may purchase them
  • .alt : individuals only, noone else may purchase them

If the rules were strictly enforced, we would pretty much eliminate the problems we currently have. Joe Schmoe can have nike.alt if he wants it, and Nike can have nike.com , and both are happy. Not only that, Nike can't sue over nike.alt, because just by being a .alt, it is obviously not nike, and therefore they can't claim it would be mistake for Nike.

On top of this, by having .xxx , and enforcing the rules surrounding it, you pretty much guarantee that filtering becomes as simple as possible. One line in your /etc/hosts and you're set. (or, equivalently, block it at the gateway/proxy/etc).

Anyways, I just think that most of the problems come from the fact that since anyone can buy anything, TLDs are really nonexistant. Instead of a hierarchical namespace, we have a flat one, because everybody just purchases the .com, .org, and .net for whatever they register. Not only that, since there isn't a defined TLD for individuals to own, companies get all pissy because somebody took "their" domain name.

Of course, the solution I REALLY want to see is not at all feasible. I would like to see policies set forth that say that noone can own a second level domain. For example, Microsoft should be something along the lines of "microsoft.software.com" . That way, if somebody else wanted "microsoft.homepages.com", that would be fine, as homepages.com is obviously not where you'd find a software company.

However, that will never happen. The dot-com craze that has overcome the world will see to that. It is too bad too, because it would make the net make more sense, as it would basically be categorized without even having to spider or anything.

Anyways, just my opinions.

and I'm a dumbass! (3.00 / 2) (#9)
by blaine on Fri Nov 17, 2000 at 01:58:50 AM EST

I meant 6 domains. We can't forget the lovely .edu , which should always be around (and should be available to institutions around the world, not just in the US!)

[ Parent ]
Couple more thoughts (3.80 / 5) (#14)
by J'raxis on Fri Nov 17, 2000 at 09:22:23 AM EST

I think this is exactly what should be done to the system, but with a couple minor changes.

  • COM/NET/ORG have already become cesspools and should be phased out entirely. Possibly make new TLDs for the new system, such as .INC, .ISP, and .NGO respectively.
  • .ORG is also currently used by opensource software groups. Where would they fit in a new system? Maybe another TLD for these (.SRC?)

(Also, I think you missed .gov)

-- The JRAXIS.ALT Raxis

[ J’raxis·Com | Liberty in your lifetime ]
[ Parent ]

Some others (3.00 / 1) (#17)
by flieghund on Fri Nov 17, 2000 at 10:51:29 AM EST

(Also, I think you missed .gov)

As well as .mil. If you think the US military is going to give up .mil, you haven't been studying world history lately. 8^)

And then there's .int. I used to think nato.int was the only .int out there (may have been true a few years back), but now there are all kinds of "international organizations" that use it. Not sure how to go about getting one... but I would like to know. 8^)

Using a Macintosh is like picking your nose: everyone likes to do it, but no one will admit to it.
[ Parent ]
sorta intentional (3.00 / 1) (#18)
by blaine on Fri Nov 17, 2000 at 11:31:56 AM EST

I left .mil and .gov out mainly because they have nothing to do with the public. They aren't acessible now, and won't be acessible then :P So I figure we can ignore those, and the government can do what they want with them.

[ Parent ]
Some useful .mil and .gov sites are open (none / 0) (#38)
by Ricdude on Sun Nov 19, 2000 at 06:54:09 PM EST

I hit tock.usno.navy.mil once a week, and have had no problem reaching it.
Also, www.usps.gov comes in handy for estimating postage for packages.
The rest of the world might not need them, but their handy for me. =)

[ Parent ]
not what I meant (none / 0) (#40)
by blaine on Sun Nov 19, 2000 at 11:38:38 PM EST

What I meant was that we couldn't purchase them then, and wouldn't be able to if we went with my plan. The government can do whatever they want with them.

Although, I honestly think .gov should be opened to all governments. Just use .us.gov , .uk.gov , etc.

Or .gov.us, .gov.uk . Whatever.

[ Parent ]
Why ditch com/net/org? (none / 0) (#25)
by wesmills on Fri Nov 17, 2000 at 02:40:44 PM EST

As the subject says, why should we ditch the existing com/net/org names? The only way a system like any of us proposes will be accepted is if it has backwards compatability to the existing registrants.

I have had wyvern.org for over 3 years, and many domains exist that predate that length of registration. Why should we lose ours because some others came in later and took advantage of a broken system? Heck, even the individual who registered his ".net" just 5 minutes ago should be able to keep it.

If you're a new entrant, late to the ballgame, and are "stuck" with a .alt for your individual site, I'm sorry. Much like 214 numbers in Dallas were a sign of prestige (before the overlay, but I digress), yes those who have legacy names will look different from the newcomers, but that simply can't be helped.

----- Signature campaign to support K5, become a member!
[ Parent ]

Slight correction (none / 0) (#39)
by J'raxis on Sun Nov 19, 2000 at 09:08:06 PM EST

I didn't mean drop them as in eliminate them, I meant if a new system is accepted, stop allowing registration of Com/Net/Org, make them obsolete, deprecated, whatever.

From another comment:

As well as .mil. If you think the US military is going to give up .mil, you haven't been studying world history lately. 8^)

Since Edu, Gov, and Mil are already regulated as to keep them to their purpose, there would be no need to phase these out anyway. It's just the publicly available Com/Net/Org that've become a mess.

-- The Corrected Raxis

[ J’raxis·Com | Liberty in your lifetime ]
[ Parent ]

Lines not that clear... (2.00 / 2) (#20)
by paulerdos on Fri Nov 17, 2000 at 12:07:48 PM EST

The plan you propose seems ok at first glance, but then you realize there are a lot of gray areas. Many sites are not strictly "business" or "personal": for example, I could be selling software I wrote on my personal homepage that also has pictures of my dog. And we already know the difficulty in defining what "porn" is: how much skin do you need to show before it's considered porn? The list goes on. Not that I have a better suggestion ;) It seems that there are just too many conflicting goals here, and it's simply not possible to satisfy them all. Meaning, I think the system we have now is as good as we can get.

[ Parent ]
not really (3.00 / 1) (#21)
by blaine on Fri Nov 17, 2000 at 12:13:44 PM EST

How is a flat, limited namespace with no enforceable rules "as good as it gets"?

Honestly, right now, the namespace is SEVERELY limited. By splitting it up into 6 domains that have enforceable rules, you open up a lot more possibilities.

On top of that, notice, I didn't state what domains could be used for business. I simply stated that only companies could buy .com, and only individuals could buy .alt . If you aren't a company but happen to be selling something, you could have that on your .alt without a problem.

[ Parent ]
Money will win. (3.16 / 6) (#11)
by Nyarlathotep on Fri Nov 17, 2000 at 02:43:41 AM EST

If you have try to set up a universal DNA system then it will probably end up like like ICANN, WTO, etc. (i.e. a stooge to corprate interests). The only real solution is to set up a system where the "freedom is partially built into the technology."

You could write an alternative DNS system which enforced some sort of voting system, but realistically it would be the systems smallness which protected it from legislation. I suppose shuch a system might eventually win (without being corrupted) by developing cowaperative enforcment strategies (ala Usenet Death Penalty), but I expect that the Usenet Death Penalty would be illegal now if it had allowed some joke site to keep gwbush.com.

Anyway, the real solution is to make it just not matter, i.e. use AlterNIC or OpenNIC will provide a more interesting backwards compatible system while something which dose not make names valuble would become the long range substitute. It could be as simple as having a web browser which automatically did searches based on the site you were viewing, i.e. look at slashdot and see smokedot and the slashdot parodies, look at microsoft and see microshit.com off to the side.

Campus Crusade for Cthulhu -- it found me!
Code As Law (3.00 / 2) (#16)
by dsilverman on Fri Nov 17, 2000 at 10:10:10 AM EST

Building regulation into the code of the technology is a viable idea, one that has been explored in Lawrence Lessig's recent book, Code and Other Laws of Cyberspace. Notice that I did not discuss any specific methods of enforcement, however there are many possibilities, which would need to be agreed upon by the community. The important thing is transparency, non-profit status, and community involvement. Then, reforms can follow.

"Speak softly, and carry a big stick."
[ Parent ]
Don't you really mean 7? (2.60 / 5) (#13)
by slambo on Fri Nov 17, 2000 at 08:29:31 AM EST

What happened to .gov? However, since the TLDs are global in nature, maybe this would be better in .gov.us, .gov.mx, .gov.jp, etc.
Sean Lamb
"A day without laughter is a day wasted." -- Groucho Marx
A simple solution? (3.50 / 4) (#15)
by OmniTurtle on Fri Nov 17, 2000 at 09:56:21 AM EST

Your idea for a non profit organization is a good one, however I would allow any organization to be a top level register provided they paid some nominal fee and provided one root server for each tld they wish to register. The root server must be able to meet certian performace levels (100ms pings from other root servers.. etc). By this method the burden of the non profit organization having to maintain all those root servers would disappear, and the burden would be placed on those who have an incentive for things to run smoothly.

This would also greatly speed the adoption of the new system, as initially it wouldn't take much to sign up with the new organization, plop your nameserver on a T1 and start registering sub domains. I imagine who ever registers .sex will make a killing. :)

Distributed DNS (3.33 / 3) (#19)
by r0cket on Fri Nov 17, 2000 at 12:06:03 PM EST

The registry system really does need to be overhauled, not only in handing out names, but also in the system architecture. The Other Site (tm) has an interview w/ Ian Clarke on the development and applications of Freenet, one of which is an idea for a distributed DNS service. A quote from the article: "One of the initial applications for Freenet that occurred to me was to replace the Domain Name System, because Freenet can attach a piece of information to an identifier. So the identifier might be the name of the computer and the piece of information might be the computer's IP address. So one of the opportunities will be for Freenet to replace [the current, centralized DNS] mechanism." The Freenet technology could give the Domain system back to the users and improve the efficiency of the system. The full interview can be found here on the O'Reilly Network.

Political solutions are not the way (3.00 / 1) (#22)
by aminorex on Fri Nov 17, 2000 at 01:38:26 PM EST

A technological solution is the way to go. Embrace and extend. I propose to replace the current DNS system with a peer system. The notion of TLD root set can be determined by automated peer review, using voting rounds. Let anyone create a TLD, and propagate it virally to every volunteer root server. If there are collisions, execute voting rounds for peer review. Any node suggesting a discarded conflicting map loses reputation, and becomes excluded from the system by consensus. That puts ICANN out of business, and indeed eliminates the need for any central authority.

FreeNet may be the best approach (4.00 / 2) (#24)
by DontTreadOnMe on Fri Nov 17, 2000 at 02:37:03 PM EST

There is some discussion in an interview Ian Clark of FreeNet fame gave, in which he suggests FreeNet may have an alternative solution to DNS altogether.

On the Domain Name System:

"One of the initial applications for Freenet that occurred to me was to replace the Domain Name System, because Freenet can attach a piece of information to an identifier. So the identifier might be the name of the computer and the piece of information might be the computer's IP address. So one of the opportunities will be for Freenet to replace this mechanism."

As a decentralized approach this has the potential to avoid political pitfalls to which any centralized organization, no matter how benevolent, can succumb. Perhaps we should consider how such an approach might be integrated with existing system calls.
http://openflick.org - Fighting Copyright with Free Media

Might work, but ... (none / 0) (#28)
by Simon Kinahan on Fri Nov 17, 2000 at 05:59:07 PM EST

Freenet does not seem like the best technology. It seems more geared up for larger data objects, but I guess thats not necessarily important.

The biggest issue though is that Freenet does not gaurantee that if a data item is present in the network it will be found. Thats OK for person to person communications, but for a piece of vital network infrastructure its not sufficient. Similarly, a Freenet network can contain many matches for a single key, and which one wins out is really a matter of fluke as no algorithm for conflict resolution is defined.


If you disagree, post, don't moderate
[ Parent ]
People who can't get along (3.33 / 3) (#26)
by HypoLuxa on Fri Nov 17, 2000 at 04:03:38 PM EST

You haven't addressed the problem of domain name disputes. This is currently a problem with the current ICANN setup, seeing that you can essentially buy a judgement in your favor with any disputed domain. There are also questions of ICANN policy contradicting local trademark or fair use laws. This is a realy murky area, and it's going to get a lot worse before it gets better.

I think you have brought some good ideas to the table, particularly regarding transparency, bu there are so many other problems that this doesn't even begin to address. Domain name disputes, aleviating the crunch on root name servers, and who can use which TLDs when are just a few of the questions.

I would also be loathe to "solicit grant monies" and "petition the United States government". The problem with the mess we have now is that it's too controlled by the financial interests of the US government and the major corporations who fund it. One is NSI, the other is ICANN. We've been down this road before, and we are now where it leads.

I'm guided by the beauty of our weapons.
- Leonard Cohen

Plenty of possibilities... (3.00 / 1) (#27)
by vastor on Fri Nov 17, 2000 at 05:35:47 PM EST

But I'd think one of the easiest ways of doing it would be to create parallel domains.

So offer .com to general commercial entities but in a .com2 hierarchy to avoid conflict with the existing .com domains.

So amazon.com could register amazon.com2 and everyone is happy. Then also have in the existing setup a 'map' for those not using the alternative DNS mechanism. So if amazon.com didn't exist but amazon.com2 did, then people not in the system could find it by doing a lookup of amazon.com2.altdns.com or something like that (just duplicating the new structures under an existing one - if people advertised their domains as blah.com2.atldns.com under the old system and blah.com2 if you're using an advanced ISP, then that'd create pressure for ISPs to adapt to deal with the new domains so that they weren't left looking outdated as customers started asking if they supported the .com2 domains). [This might actually be said in the gradual name approval section, it's not too clear]

Most countries have suitable definitions of non-profits orgs. Like to get a .org.au domain you have to be a fairly genuine non-profit organisation.

NGO is surely non-government organisation rather than non-profit organisation even if there is a lot of overlap there may well be commercial NGOs out there (otherwise it seems they named it foolishly, the name certainly doesn't sound like it'd be exclusively non-profit orgs).

If say .shop was launched then it could be worth having 2nd tiers below it like pet.shop and computer.shop before people start buying up domains.

As for transparency - why not just let people outright vote on the final proposals. Have a $10 annual membership or something along those lines, it'd raise money to covery any expences and stop ballot stuffing as you'd get if the voting was totally open. Maybe even create a members domain for those in the organisation.

However the options are endless really, people have proposed plenty already in this discussion. If the transition to an alternative is made easy and it doesn't bring any problems with existing domains then I don't see why it couldn't work as long as people using the "old" system are also able to access the "new" alternative one via an easily done remap or something similar.

I don't go with all that "playing the game of politics" stuff - this is something that can be done by purely technological means IMO. I doubt people petitioned the US gov't about usenet2, so there is no reason why a domain2 should have anything to do with it. If it's worthwhile people will transfer to it of their own choice, if it isn't then they won't (and perhaps this would be a good way to shake US gov't shackles of control - demonstrate that its something that can be run without intervention). Make it big enough and ICANN will compromise to work with it or fade from view.

More tlds don't solve the problem (none / 0) (#32)
by ciaran on Sat Nov 18, 2000 at 09:43:15 AM EST

The problem with this is that if the alternative namespace becomes popular, any company with a .com will just buy the corresponding .com2 . Domain squatting would pretty much rule the tld, as opportunists would snap up microsoft.com2, yahoo.com2 etc. . At this point, microsoft would take legal action to get 'their' domain back, any US court would rule in their favour, and the system would end up just the same as the current mess. The obvious solution to this is to take the system out of the jurisdiction of the US courts ... however for this to work they would basically not be able to trade inside the US at all, which would cut them off from their largest market.

So basically the system would either end up exactly the same as the current system, and basically controlled by the whims of the US courts, or removed from a large portion of its market and probably unable to do business. Either way it would not be particularly useful.

What is being proposed here is an entirely new root server system. In theory, this is a good idea, however I don't believe that the US courts would allow such an independent system to exist, as it would violate trademarks and copyrights, which companies could not allow to happen, so the legal pressure would force them to close it down. Locating it in an inaccessible place (Sealand ?) could help this, however blocking at routers along international barriers is technically possible, if extremely impractical. This is exactly the kind of solution favoured by courts in dealing with technical problems - shifting the problem to someone else - in this case forcing the isps to block traffic from these 'illegal' root servers.

So basically, any new dns systems are doomed to end up as regulated as the current one and subject to the rules of the US courts, despite its claim as an international system.

[ Parent ]
TLDs server (4.00 / 1) (#29)
by PacketMaster on Fri Nov 17, 2000 at 09:46:43 PM EST

As much as a limited number of TLDs causes angush amongst the community, they do server a useful function. The root servers, using TLDs, provide excellent versioning and data integrity control. If you start competiting with the TLDs with all these "alternative" TLDs mass confusion will reign.

I this editorial has a good idea. I'd like to see work done on an ICANN competitor. If we can force the registers like NSI to have to open and share their databases, why can't ICANN be forced to compete? Isn't that what the free market is all about? Let another organization be in charge of another set of TLDs. These TLDs could not compete with anything in ICANN. All ICANN essentially is is an organizational control mechanism that basically decides what TLDs there are and handles issues like domain disputes, keeping the registries in control, etc... I'd say get a strong, well-support and LEGALLY UPSTANDING organization going and then petition the Dept. of Commerce to grant status to it. Under the Sherman Anti-Trust act, ICANN can be successfully toppled as monopolistic given the right lawyers. Maybe get some of the organizations that got their TLD proposals reject to band together and get something going. That $50,000 entrance fee took a huge bite out of them and they might like to bite back. Regardless of the branding of the US Government as a corporate-controlled entity, people need to realize and accept the fact that the US Gov't holds all the cards right now regarding Internet administration. They can topple ICANN as fast as you can SAY "topple"

Just some food for thought

needs (none / 0) (#30)
by matman on Sat Nov 18, 2000 at 01:56:50 AM EST

Well, there are already alternative namespaces out there - I dont recall what they're called :) Essentially, I think that what needs to be possible, is to use two different DNS spaces at once - a master and a slave. When you do a query, you usually do a recursive one to a name server, and it goes out and does an interitive one. I suggest that someone writes a name server that lets you use more than 1 root server and specify which have priority over others. When you do a query, it queries the primary, and if there's no such host resolved, checks the next priority tree, and so on. You'd also need to be able to specify what tree to search if you didnt want to let the name server take care of it for you, so you could use special first parts of the domain, like dnstree1.domain.com or dnstree2.domain.com (the host dnstree2.domain.com on dnstree2 would be dnstree2.dnstree2.domain.com).

That would make it so that you could still get at all dns spaces, without confusion, and a transition to a new better managed space could be made with ease.

Good idea, but... (none / 0) (#31)
by GreatUnknown on Sat Nov 18, 2000 at 07:19:00 AM EST

...how are you going to get people to use your alternate DNS? Everyone on the internet uses the ICANN controlled (I think) DNS servers. What's the incentive to change, or to query an alternate DNS? If anything's going to happen, it would have to be more than just the "geeks" that use the servers.

You'd have to have a lot of good content only available on your DNS before anyone actually uses it, and if it's content that will attract a lot of hits, why wouldn't the owner register it on the "normal" DNS.

Great idea, but I doubt if it would work because the general public have no need to change, so ISPs have no incentive to use your DNS as well as the normal ones.....

"Pirate" DNS (4.00 / 1) (#35)
by vryl on Sun Nov 19, 2000 at 02:13:37 PM EST

Well, with the current trademark and 'intellectual property' bullshite on the net, I can see the possibility of a 'pirate' DNS springing up for those 'in the know' to get all the stuff they want/need, or use URLs to publish anti corporatist parodies or whatever.

I think namespace have a functioning alternate DNS and you can configure your nameserver to support it.

Its worth investigating, I think.

[ Parent ]

Thoughts on namespaces (none / 0) (#33)
by cybin on Sat Nov 18, 2000 at 02:37:24 PM EST

I remember a few years ago when AlterNIC started up, pushing alternative namespace... they seem to have been absorbed or have morphed into a regular registrar now by the looks of the site www.alternic.net points to.

everybody doubted back then that it would work too... looks like it really didn't. i'm not familiar with the (if any) alternative DNS systems in place... however, i've always been confused as to why the U.S. gets special treatment... com, net, org... why? we don't even have the largest number of computers on the internet per capita, that would be finland last time i checked...

i wonder how feasable it would be to implement some kind of system that is able to tell what country you are in. like this: if i'm in the U.S. and i want to access WHiD.net (a site of lunatics i'm involved with) i would type http://www.whid.net/, since i'm in the U.S. and the domain is registered in the U.S. if someone in germany wanted to access it, they'd have to type http://www.whid.net.us/. if the site was in germany, i'd have to use www.whid.net.de, etc.

kind of like PBX's, like the one we have at our university. you dial 4 digit extentions on campus, off campus you have to dial the exchange. same thing for area codes... if other switched networks work like this, why can't the internet? :)

i think we also are perhaps too attached to com,net,org,etc. we need more good, ambiguous TLDs like these :) i would disagree that only ISPs can use .net (as blaine mentions above) because, what would a site like K5 do? they aren't really a company, but they're not really nonprofit nor an ISP. is org appropriate for K5? yes. is org appropriate for "the other site" ? i would say no. they should have a dot-com.

just my $0.02. -matt

Die, DNS, Die (none / 0) (#34)
by moshez on Sun Nov 19, 2000 at 09:28:22 AM EST

We should not "replace", or "augment" DNS. DNS should
die, and we should help it die gracefully.

How do we keep name->ip? In a decentralized manner.
Each popular group could have it's own name->ip.
In case of conflict, then the user should be asked.
Here's how it would go:

A site that would want to be *a* DNS root could put up
a webserver.
If I want to query *that* webserver where my resolution
should go (say, www.moshez), the query would go:

GET /moshez/www HTTP/1.0
Host: query.dns

(The HOST header is for seamlessly integrating with other
servers that might be on the host.)

URLs would either give IP, or name and the IP of the server that manages them, like that


If we just integrate this into Mozilla (it's backwards compatible), then IE users who can't click on links will demand that MS embrace and extend it.

[T]he k5 troll HOWTO has been updated ... This update is dedicated to moshez, and other bitter anti-trolls.
We don't need any central root domain servers. (none / 0) (#36)
by Skapare on Sun Nov 19, 2000 at 04:43:59 PM EST

We don't even need any entity trying to tell us what TLDs we may, or may not, be able to look up. In fact we can do anything and everything the way we want to do by just running our own DNS.

To be more practical, anyone who is running a DNS sever, which can be an ISP, a business that's online with their own servers, or anyone who runs one for whatever reason, can simply substitute the "dot" zone with zone data of their own. That would let them control their own view of the TLD space. An ISP could also do this in multiple different ways for their customers to choose from.

And, such control allows you (the DNS admin) to control where the zone information comes from. You can choose to get .com from InterNIC, or you can choose NOT to if that's what you like.

Of course, this does mean you have to set up a "dot" zone with ALL the TLDs you want, which means you do have to know where to get all that information (all the NS and A records). I thought about what might be needed for that a couple years ago and created a web page form that builds a "dot" zone from TLD selections. The concept is called Grass Roots Servers because it would be a "grass roots" effort at running "root servers" everywhere (instead of in some bureacrat's control in a central location).

The Grass Roots Servers site is a couple years old with old data. But it was intended as showing a concept. If people want to do this (want to control their own view of namespace) they they can do this. Not only can they use Grass Roots Servers to build their own "dot" zone, but they can also run their own "dot" zone builder web site, since the software is free (GPL). Download it from there. Or just build your own if you like (the concept is not that hard to do).

Will it fragment the namespace? Of course it will. But the world of business is all about fragmenting everything. It's called competition. He who has what more people prefer ends up providing it to more people. It's that simple. And you can always go back to InterNIC any time you like (if they survive).

It's voluntary, but Bernie is on "their" (none / 0) (#37)
by isdnip on Sun Nov 19, 2000 at 05:29:15 PM EST

DNS is a voluntary system. As others have noted, any server can point at any root. So if you don't want to point at an ICANN root, you can point at your own choice. If everybody chose an alternative, than ICANN's root would be moot. This is, technically, a free-market decision. ICANN has no statutory authority. They simply walked in upon Jon's death and claimed it.

But the market's not quite free. ICANN has the incumbency. So it would take a concerted effort, and a united opposition, to create a useful alternative root. Alas, that probably means that the ISPs would have to go along. And the biggest ISP is Bernie Ebbers' UUNET. And the new GHP of ICANN is Vint Cerf, who works for Bernie. So Bernie thinks he has ICANN under his thumb, and he couldn't give a goat's bzadeh about the public's concerns.

I suggest that an alternative would have to fall back on ICANN's root, interposing itself in case of any conflict. So for instance the alternative root would point to Virtual Works for vw.com, assuming that the rightful pre-WIPO owner of that domain wanted it back. But the alterantive root wouldn't bother to re-register everything else; it would, finding no answer in its own database, fall back on an ICANN-rooted source. Eventually, if the registries found it viable, the alternative might grow up to be a full-fledged competitor.

Feasability of alternative namespaces | 40 comments (37 topical, 3 editorial, 0 hidden)
Display: Sort:


All trademarks and copyrights on this page are owned by their respective companies. The Rest 2000 - Present Kuro5hin.org Inc.
See our legalese page for copyright policies. Please also read our Privacy Policy.
Kuro5hin.org is powered by Free Software, including Apache, Perl, and Linux, The Scoop Engine that runs this site is freely available, under the terms of the GPL.
Need some help? Email help@kuro5hin.org.
My heart's the long stairs.

Powered by Scoop create account | help/FAQ | mission | links | search | IRC | YOU choose the stories!