Wow, utopianet sounds great. How do you handle fraudulent abuses? Let's say I make it look like someone on Inet is spamming. Bam, they're banned for 5 years before they can even get evidence together to prove their innocence.

And wouldn't it be nice if we could trust a single organization to make all these crucial decisions. Maybe we should trust Google to manage things, I hear they don't want to be evil.

And where are you going to find the manpower to manage all the registrations, support, disputes. I imagine there's a registration fee involved. Mo' money, mo' problems.

And what if variants of my Trademark are a legitimate separate trademark? How do you handle variants that overlap current domain names?

DNS is broken no doubt, but as nice as this sounds, this doesn't seem plausible to me.

I like this idea, but I'd like to tweak it a bit. Here's the issue: forcing pornography into a .XXX TLD isn't enough - it's too hamfisted to get all of the stuff a parent or school might want to block (what about Fredericks of Hollywood and Victoria's Secret?) and it doesn't offer any protection for non-pr0n that people may want to block like gambling or shopping or file-sharing.




I would suggest that I would suggest that IDNS define a set of site content categories (use the categories from WebSense, or St. Bernard, or some other URL companies as a starting point). Then require registrants to select up to 3 categories that best describe their site. Part of the domain registration fees should cover having humans (or smart tools) periodically examine their registered site for accuracy.




By having a standardized list of categories, and a trivial DNS-like way to look up a site's category, browsers could offer high performance, high quality filtering and content access control.




My suggestion is colored by having worked for a couple years in a URL-filtering company (note I said workED - I'm no longer there).

This doesn't make a lot of sense to me. Great, he wants to take over the cultural imperialists at ICANN.

But your article (and, as far as I can tell, what he has come up with so far) only gives ways to make it work (sort of; after everyone rewrites every single internet tool out there, NOT just the web browser)... But no reasons why it is better.

There is lots of promise.. "we'll be utopian, we'll resolve disputes fast, we'll ensure everyone has a valid mailing address" but no word on how any of this is feasable.

Child porn and spam are already illegal in many places, and violate the TOS pretty much everywhere. The problem is, you can't just kick an entire domain (huge corporation, anyone?) off the 'net forever because of one violation by a customer.

And spam is usually sent by botnets, not directly emailed by some guy in his basement sitting on a cable internet connection laughing all the way to the bank because the provider sanctions his behavior... Why? Because it's ALREADY A VIOLATION OF THE TOS.

Under this plan, every single ISP would be banned from the Inet in 24 hours. Genius.

Your friends idea has merit, to be sure. I like the idea of controlling the Internet and elimination users that just make the experience harder for the rest. However, i have to wonder what would prevent the Inet administrators from becoming too powerful. People in such a positition to control such a large industry as ecommerce had better not be corrupt. Really, its like the history lessons you learned in school: too much power on one place can be extremely beneficial, but most of the time the power gets abused. The idea of an organization having such power is frightening when you consider what abuse of that would mean.

that sounds a bit over dramatic, but the point is, theres a lot of rough details to work out and polishing to go before that would hold up in todays world. Something to consider is that the Inet (kindof ired of the 'i' infront of acronyms and product names....ipod...iphone...ect.) the Inet is going to need to be a global organization. This would need to operate independent of the united states laws, or any other countrys laws.

Take, for example, the insanity that was the struggle to allow for free exchange of encryption, in the late 90's. In a global marketplace, companys based in the united states would be restricted y a very different set of laws than a company that ran out of japan or Germany, despite the fact that all people all over the globe, so long as they have Internet, have equal access to that company as much as any other. Physical location causing some websites to operate or behave differently than others is a major part of the issue with spam and the like: websites doing illegal things, such as selling pirated software and movies, or child pornography, have only to buy their own island nation to circumvent international laws and treaties. You wont find nearly as many shady websites in the united states as you will in countries abraud that have little or no legislation concerning such things.

Is this the first step towards one united governemnt for all of humanity?

What's in it for the end user? I don't register domains, I don't see the need to download a patch and fiddle around with names (.xxx or .com .whateveritmaybe) that I don't care about. Anything that evolves does need to be redone at some point - but in a very non-disruptive manner. This is definitely disruptive and I don't see anybody really gaining out of this...

This a harebrained scheme that has no chance of achieving its stated goals.

Strip away the technical details, and the plan is to simply ban/eliminate/make inaccessible all uses of the Internet that don't comply with whatever arbitrary rules are set up by Inet. You think we have disputes, profiteering, and monopolism NOW, wait 'til something like this is put into place.

Bob,

Instead of revamping the Internet, how about getting folks to resolve some huge issues.

1. Email vendors, ISP's, large corporations have still not resolved the spam issue. Here we are years later and still nothing has been done. We can crack the human genome (DNA code) and yet we can't figure out or agree on how to stop spam.

2. IP spoofing has always been a problem. The recent root server DNS attacks used IP spoofing. Why can't we get ISP's to simply check outbound packets. If they don't have an address from their net, then the packets get nulled.

I think things like this have been done already. There have been browser plug-ins or something that would allow for single-word names to get translated to domain names. There was behind it a company of course that hoped to make money off the scheme eventually. Don't know if they are still around.

Beyond that, it sounds like a lot of "easier said than done" to me.

For one thing, I have an anonymous domain. Not because I'm a spamer or porn person, but because I don't want to get a lot of junk mail and junk phone calls to go along with all the junk e-mail I get. My junk mail volume went WAY up with I registered my first domain and I could tell by the way some of the mail was addressed that they got it right off my "whois" info. It will of course take a long time for that effect to taper off now that the information is anonymized, I wish I'd had the knowledge to do it that way in the first place.

I, for one, welcome our new INsEcT overlords!

Can we put Kofi's son in charge?

Well, try it, by all means. But isn't this what CompuServe and AOL and MSN and Apple eWorld were all offering back in the days when they were competing with the Internet? Say, whatever happened to those guys, anyway?

I must be missing something here. It would be nice if things worked the way this scheme seems to suggest - spammers are easy to catch, and once banned, stop spamming - but I just don't get how it's to be achieved. Unless you need a special chip in your arm to access the Inet, I don't see how it's feasible to accomplish any of this.

And is the internet's main problem really that DNS is inefficiently administered?

Oh the irony….Check out the Google ads served for this article --->

Other than ‘identify and persecute’(which is what we have now) how would this system reduce spam?

One of the worst ideas I have ever heard. Rule with an iron fist huh? No thanks, us Americans don't like that kind of philosophy..

I like this idea! Only thing I don't like about it is the "any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated" ...I mean, what if somebody's on vacation for a month and doesn't want to be plugged in? They get back and find their domain's been down for 3 weeks. I don't like that idea.

So where do I download the browser patch? ;)

Wait. This is a joke, right? This is pretty much a laundry list of all the descriptions of the perfect Internet — with no explanation as to how any but the most trivial problems will be solved. "The Inet's e-mail service would incorporate centralized anti-phishing and anti-spam techniques, and would block known spambots." Right, and my new computer chip will do voice recognition and solve NP-complete problems in polynomial time. Is this some kind of joke about David Harrison that I'm not getting? Or maybe there are more people like David Harrison that I don't know about? I don't see the point of the article otherwise.

It had to happen eventually. After years of enjoyable reading of Cringely, he's finally jumped the shark. Oh the humanity.

What good is a DNS replacement that only works with browsers? That's just silly. The network, no matter what the "I" stands for, needs to support a host of protocols beyond HTTP to be useful. SSH and NTP may only be used by/known to the nerds, but SMTP and POP3 are pretty widely used.

The foolishness about needing to reply within 24 hours to any notice from Inet would, by the way, render the network strictly a place for larger players who can afford to provide coverage around the clock. Maybe not 24/7, but certainly 12/7 to be sure of learning of a problem, fixing it, and responding within the 24-hour limit.

Van

That filthy Bush administration! First they imprison the Dixie Chicks, who will not see the light of day until they are WELL into their sixties, and now they're using the internet to take over other countries and make them part of the U.S.! And they're PROUD of this! Careful, Bob. The way the Bushies are knocking down doors and rounding up innocent people these days, this kinda talk might put you in some hot water. Dangerous times, folks. Dangerous times...

Let's take this good idea and make it even better: not one alternative DNS system, but two. The second would be everything Inet is, but with zero advertising. Advertising in any form would not be tolerated. People have the right to browse content without having to be imposed upon by advertising. A no-advertising clause could be included for the TOS for the Inet, subject to the same 5 year punishment for other offenses.

I suspect that the promise of an outright advertising ban would increase the allure of the Inet. For one, advertising offers me nothing. I never click the ads, let alone buy any of the advertised products. All advertising does is clog up websites and use more bandwidth unnecessarily.

Another interesting thought - TCP/IP suffers from an inability to prioritize traffic. And video traffic is exploding. Now imagine a replacement for TCP/IP that does prioritize traffic - how would that affect spam bandwidth? If video is clogging the Internet, that means that it is clogging up the bandwidth available for spam. More video, less spam.

Another possibility - email prioritizing. Mail to self-acknowledged friends gets priority. That way if someone suddenly adds 100 friends to an email account in a fashion that screams "scripted", it gives email providers one more variable to check to identify spam accounts.

Inet sounds like a good idea. It would be sweet to combine it with Google's fiber.

I'm hungry...

Inet, I like it. Where do I sign up?

There must be about a gazillion domain names already registered. As they are ported over, who will check for legitimacy and compliance? Will there be a "report this website" button? Will a wiki-style committee vote on things? 24 hour resolution sounds good, but it takes serious manpower to make it happen.

"People have the right to browse content without having to be imposed upon by advertising."

BS. Seriously. Servers don't pay for themselves. Nor is bandwidth free. Administrators, editors, authors, all like to pay the rent and put food on the table.

Unless you want to dig out your credit card and PAY for every page viewed, you have no "right" whatsoever to demand that people "give" you their content for free. If it has value, then it has value.

Before putting the hands on DNS, the real backbone of all the Internet, I'd prefer to discuss whether the current (application) protocol design (and implementation) fits the needs.
Email (ESMTP+POP+IMAP) badly needs a re-design in order to get a more adequate behaviour.
Authorisation, authentication, signatures, delivery status notifications and so on should all be mandatory parts of the protocol, not application options. I think about something much like SFTP would fit at least part of the solution. And maybe also some technology similar to P2P.
Only later I would start putting the hands on the DNS.

This is the first ever Cringely article that seems like it wasn't proof read:


"Inet DNS calls to servers would be flagged by a bit in the call courtesy of the browser patch. This could be read by website servers using server-side code, and consequently a call via the Inet could result in a different response to a browser call than if it came via a straight check on ICANN's DNS. This means a site can be visible under, or generated for an Inet call, but invisible or not generated for a straight ICANN call, or vice versa."

..and another thing:

Porn sites are already labelled. No need for a .xxx domain.
End of story.

"The Inet's e-mail service would incorporate centralized anti-phishing and anti-spam techniques, and would block known spambots"

How?

So, replace the 'imperialist' ICANN with a system run by one guy and his mum from his bedroom?

What possible incentive would people have to move to an internet based on this world view?

Disclosure: I do internet marketing, sometimes on behalf of quite large companies. I do both (white hat) SEO strategy and paid search work - advertising.

Some users specifically seek out advertising. I know of some searches where the users are looking to buy stuff, and the organic results reflect a preoccupation with research type information. Current search engines are lousy at inferring user intent.

Like it or not, if Inet is to work, there needs to be a legitimate mechanism for advertisers to appear, or they will find ways to appear that are not controlled. If advertising is banned or too tightly regulated on Inet, then all that will happen is that sites will end up with covert cash feeds in return for product placement copy, or some equivalent. That's not an improvement. What will increase the interest in advertising circles is if this seen as a higher value network - fewer advertisers, more real people. The forces that will destroy it are self-generated.

Technologically: probably feasible - I could see this take off in OSS circles. The problem for it is financing.

Why is financing important? See DMOZ. Inet appears to need a large number of reviewers, and without cash, those are volunteers. Ever seen the allegations that (some) DMOZ category editors are industry insiders? I have, and I've spoken to editors who were industry insiders and used "unfair techniques" to promote their favored companies. A chunk of cash insulates the reviewers from that. Like it or not, there are advantages to being a "professional" and one of them is that you can afford to maintain ethical standards.

Until this system has a way to generate enough revenue to make the control mechanism economically sufficient to avoid taint from bribery and corruption, it is of technical interest only and will be subverted as rapidly, or even more rapidly, than the current Internet, through corrupt reviewers. Plato says it all in The Republic: who watches the watchers?

Understand the control system and you know how to control the system.

Bubble. Pin. Pop. Sorry.

As much as I would like to see the Net cleaned out, this idea is a dream (or should we say a nightmare) rather than a feasible system. If we want ambitious goals that could achieve something real, how about starting out with ways to:
- replace SMTP
- improve PC security and ISP ethics
- improve international co-operation in finding, catching and trying the Bad Guys

Well, I'm not saying it's a bad idea... but you really make this sound like it's a fairly trivial change and something that would be easy to police.

Hey, Have I been taken in by an early April Fools? :-)

On the technical side, changing the DNS is not that hard as you say. It's the political / religios / sociaty / people side that gives difficulty.

These utopian ideas make David sound like Annekin Skywalker saying that he would like to govern the universe, and make everything better.

"any communication from the Inet DNS system to the named registrar must be answered within 24 hours"... what happens if I've gone snowboarding, and they send me a query?

There have been court cases which have gone on for weeks and months about domain name disputes. How come these utopians can make all these decisions in one week? Sometimes the questions about who is right/wrong can be much harder to decide.

Trademarks can be valid in one country, but someone else can own them in a different country. Automatically giving them worldwide would not always work.

As I said, I'm not saying it's a bad idea... just much much harder to impliment than you suggest. The really bad guys (spammers, phishers, ...) are easy to pick off and prevent. But the people on the borders between what's illegal and what's just immoral or inappropriate are much harder to police.

For instance, what about someone selling Nazi memorabilia? As far as I'm aware, although this is distasteful to most (including myself) and really sick if you think about what it represents; it's not actually illegal. Should this be banner in Inet?

Anyway, I agree that something like this needs to happen. My point is that it'll be much harder (and more expensive) to put in place.

Steve

It's all politics.
I trust the American constitution and spirit with all its shortcomings over anything else proposed by David.
Who else will accept a spit in the eye and still stand to their right to keep an Internet domain?
Anyone can create his own DNS system ( actually VPN is a good example ). Why haven't anyone done so by now?
PS I'm not an USA citizen

Interesting but naive.

"When you register a trademark as a domain on Inet, you automatically get all of the global alternatives in one go."

Based on what countries trademark laws? What happens when two separate organizations own the same trademark as was the case with World Wildlife Fund and World Wrestling Federation? How does this new system improve the law?

"Domain dispute resolution would be rapid: one week for evidence presentation, 24 hours to decide, and 24 hours for appeals."

Did we just enter a global banana republic? Expedited does not equate to fairness or justice.

"Pornography sites could only register using the .xxx top-level ..."

And peace and harmony would come to all mankind.

Are their technical solutions to improve the Internet? Sure. Will technology be the answer to a variety of legal issues? No. If this idea focused on real solutions to technical issues it would have more merit. Unfortunately, it comes off as a utopian fantasy.

Not good. Replacing one heirarchical, centralized service with another is a mistake. The problem with DNS is that it's a centralized solution mapped onto a distributed network. Get David to think about how a decentralized naming system would work.

Email is hopelessly broken. You explain nothing wrt how your "magic" email system is going to work. ALL email should be encrypted and based upon an opt-in policy. You don't send me things until I let you. This would shut down spam once and for all.

O.k. let's look at this:

1) Domain squatting would not be permitted, either.

So, how do you tell if Mr McDonald is squatting or not?

2) Inet DNS registrants would have a real name, address, and contact details (not a PO Box)

Like the fake addresses that a lot of corporations have? I see an increase in companies offering PO Box service without you having to say "PO Box", instead calling them "Apartments"

3) any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated

So, you want to personally own a domain? No vacation for you! System crash? Hard drive failure? Oops, sorry we're going to revoke your registration! Thank you, come again!

4) Inet’s Terms and Conditions would prohibit child pornography, phishing, fraudulent commercial services, spam,

So, who's held responsible when someone sends a spam through a gmail account? Gmail? How do you define something as spam? Once enough people decide it is? People already have this type of problem with AOL and others who create global blacklists based on user spam reports.

I know I would absolutely love to have my domain yanked because the wiki I ran was vandalized, or someone misconstrued an email as spam.

Interesting point of view, but it would not normally reach critical mass of users, and not because of users (who may or may not subscribe), but because of lack of content. Most content owners will not subscribe it or will not even hear about it. Less content, less users. Less users, less content-owners to subscribe. And setting this up DOES require funds one would probabily not get.

A kind-of similar scenario but maybe with more chances of success would be a similar (yet free,at start) service of rating websites offered by google. You'd make a search, you'd find something useful, you rate it. If it's porn or spam, or with popups, you give it a bad rating.
Take this into google toolbar and add a feature "show ratings for websites that I browse". This will, in time, allow filtering of sites by content based on users input.
One could also help mitigating the spam problem like this. If you get a spam advertising some website selling whatever, you get there and rate it with "spam advertising", generating less purchases (I would personally try not to buy from someone I think is a spammer). Add warnings and filters on this database and we're able to get all the benefits with a very low cost and an early adoption.

"any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated."

Oh, that's just lovely. I have to check my email every single day? There goes my vacation...

Doesn't have to be answered *well*. It just has to be answered. IE: "Hi, you've reached Phil Smith. I'm on vacation from Feb 20 - 25th, but will try and respond no later than the 27th on my return."

I use email responders when I'll be away on vacation. Terms satisfied.

It needs to be asked, but what does Bob think about Cisco and Apple settling for nothing? Does this make it more likely that Apple will open their product line? Or is this just a red herring on their part, because they're popular enough to get away with almost anything?

Deadline pressure must be getting to Cringely. This article is complete rubbish. Bob, I don't trust David Harrison any more than Mr. Hilter and his Bocialist Party.

You talk a lot about how disputes will be settled quickly, but not about WHO will be settling them.

It sounds like the plan sums up to trading a dictatorship by ICANN for a dictatorship by someone else. Meet the new boss, same as the old boss.

"When you register a trademark as a domain on Inet, you automatically get all of the global alternatives in one go. So when Wal-Mart registers walmart.com, they'd get all the similar domains automatically"

"Pornography sites could only register using the .xxx top-level domain scuttled not long ago by ICANN."

These two statements are contradictory. A porn site would only get the .xxx registration, but what is preventing someone else registering the name for the other domains, sowing confusion?

Bob,

You've really gone off the deep end here. Up until now, your occasional political barbs were tolerable because in the meat of your writing, you actually made a good point or prediction. You put it out there for all to see and I respected that.

This week's column however is, not putting too fine a point on it, a piece of garbage. I don't know David Harrison, but he sure sounds like a "wouldn't it be nice if we could all just get along", utopian useful idiot.

ICANN has a million different ways in which it has serious problems. Overall though, it's been slugging through a job where every single decision , no matter which way it came down, was going to make most people unhappy.

And just because the Bush Admin decided NOT to let the UN (or some other "open" body) have control over it, does not make that an affirmative move of using it as "an imperialist tool". We get that you don't like Bush and his policies, but where is the intellectual honesty? This kind of tripe has now turned you into nothing better than the million other Bush-Hating Bloggers.

To paraphrase Winston Churchill: Keeping ICANN "in-house" was the worst of all the possible decisions, except all the others. The US is certainly not perfect, but what you Bush-Haters fail to realize is the same concept: at least we're not like all the rest.

As for the technical merits of the article, they've been pretty well squashed by the previous comments. Unworkable, unreasonable (Phase 3: Profit!), and unusable.

I will however add this non-technical comment:
One oft-overlooked aspect of Freedom means tolerating the speech and legal activities of others that you don't agree with, meaning, you WILL be offended by what others do, but tough.

Stick to Google and Apple and Microsoft and Yahoo, etc, etc... You will have plenty of material to think/write about until the cows come home.

Funny how you produced long, not really interesting article just to draw our attention to the MS case.

I'm glad you haven't been channelling David Harrison in the past, because this is by far the least convincing column you've ever written. Perhaps you can do a follow-up and explain how this is not really an early April Fools joke.

1. "When you register a trademark as a domain on Inet, you automatically get all of the global alternatives in one go." - which alternatives? One-letter wrong? Misspellings? Homophones? Where does it stop? Who decides?


2. "Domain dispute resolution would be rapid: one week for evidence presentation, 24 hours to decide, and 24 hours for appeals." I can see the appeal in this, but who pays for the staff required? After all, the renewal fee is "nominal".


3. "any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated." I think this would cause an undue burden on individuals - what if someone disputes your domain on the weekend or on vacation? You come back to find someone squatting on your domain because the registrar entered a default judgement for non-response.


4. INet's T & C prohibits a long list of things that are bad.

a) Why would this new registrar be any better at preventing zombie networks than ICANN? The point of a zombie network is that the owner is secret after all.

b) Child pornography? Doesn't that mean that someone from INet has to visit every page on the site to check?

c) Fraudulent commercial services? Again, it sounds like someone has to check this. If you depend on complaints, then this doesn't sound any better than the current system - except that the site gets banned before the court case gets under way and the site owner is presumed guilty rather than innocent (the only way I can see to make that one week timeline for dispute resolution).


5. "All known spammers or phishers would, where identified, be banned from the system for five years or life". Do we actually know who the spammers are? As far as I can see, known spammers generally end up in jail.


6. "Anyone operating on behalf of a known spammer or phisher would receive the same punishment". Does this include people whose computers have been rooted and are part of a zombie network that sends out spam?




All of this sounds an awful lot like an idea I heard about a while ago. We should change the IP packet structure to include an "evil" bit field. All evildoers that transmit on the Internet have to flip the evil bit. Being safe on the Internet is then a simple matter of blocking all packets that have the evil bit flipped. I leave it as an exercise for whoever wants to come up with all the specific reasone why this doesn't work, but they all boil down to "evildoers don't follow the rules."

Respond within 24 hours or lose your domain registration? I have my own domain for my home. Just me, no corporation, no employees. Does this mean no vacation for me? What if I am in the hospital for a day or two? What if I just want to take the weekend off and visit friends? What if I work a long day at the office and don't check my email and phone messages until the next morning?

And as for the policies against spam, phishing, etc, nice idea, even utopian as someone else said, but it assumes there's a solid brick wall between freedom of speech and criminal activities. There isn't, and this won't work.

I agree with the other poster, this is your shoddiest column yet. David Harrison sounds like a nutter. One idea a week for the last decade and this is the best he's got?

While I totally agree with Mr Harrison that ICANN is definitely a tool for US imperialism, let me be the first to point out that schemes such as "Inet" have been proposed several times before. Just google for "Alternate DNS Systems" for instance, and I am sure you will find plenty of alternate DNS organizations that quite willing and able to register your domain site.

The fact is, the DNS system works not because it's US-independant (it is very US-centric, but because cheap domain name registrars exist, and are doing a good business of registering names for all comers.

One reason, for instance, that many French sites are actually registered as US ".com" or ".org" is that France's AFNIC (www.afnic.fr) had decided to apply totally unreasonable rules when it came to giving out domain names. It turned out that many French companies, tired of dealing with the bureaucracies, turned to the american registrars instead. Try www.liberation.com for an example of this.

Great. So now you channel David Harrison who obviously is channeling the ghosts of Stalin and Mao??

What's up for next week? A follow up channeling Hu Jintao (President of China) telling us his glorious visions for the internet?

wow..this is boring... roughly the same equiv as alternate TLDs proffered for DECADES online but have faded away due to browser manufacturers caving.. Must be march break coming up, and this was an autopost? :)

This sounds like a dream for a utopian internet rather than a feasible plan. To say that Inet would eliminate cybersquatting or online porn is wishful thinking. Even with human eyes viewing the sites, what's to say that content won't change once someone from Inet has viewed it?

I would agree with what others have noted...this is not one of Cringley's best columns. But hey, everyone's entitled to have an off day....

two things:

1. Rather than have a single organization responsible for filtering out spammy sites, I would prefer collaborative filtering -- if a site gets enough complaints, or if an email is marked "spam" enough times, then it gets taken down. Of course this idea needs some work, but I don't think we can rely on one organization to do the footwork and do better than ICANN

2. I'm waiting for a world where we don't need to know any domain names. It is a bad user experience to have to remember exactly the address of a site. As I see it, intelligent searching is the new navigation paradigm on the web. Rarely do I type in URLs anymore. Of course URLs are important because they're unique, but as far as I'm concerned the visibility of URLs is going to become less and less as people rely more and more on intelligent search, built right into the browser. And this is a good thing.

Lead by example.

Start it, then tell people what you have done.

If you tell people what you are going to do, they will ether tell you its a great idea, or a dumb idea, and nothing will get done.

Ignore the details for a moment, and focus on Mr. Harrison's idea. The rules he mentions are an example of what could be, not what would be, so they really aren't that important. Further, his idea isn't really about replacing the whole internet, just one part, and not by throwing anything away but by setting up something that runs in parallel.

In a nutshell, he's saying the internet needs a car pool lane.

Like the car pool lane, his idea has more stringent requirements than the other "lanes" do, but if you meet those requirements, you'll be able to travel faster than you otherwise would -- particularly if the "normal" lanes are congested. If there's someplace you want to go that isn't reachable by car pool lanes, you can still get there by hoping onto the "old" lanes, it's just that the ride may not be as smooth. Eventually, if enough folks decide to "car pool", you may not need anything except car pool lanes, but since this is the internet we're really talking about, the old lanes won't necessarily disappear.

For example, imagine if every university in the world could wave a magic wand and switch over to this new system tomorrow. There have been complaints for a while from scientists and researches that there's little bandwidth left over for serious work; setting up an "Independent Network" that only includes universities, and only those universities that agreed to certain rules, would make a great deal of sense. All of these universities would still have their public face on the "old" Internet, so they'd remain reachable by folks who aren't on their Independent Network, those folks just just wouldn't get all the benefits that being on the network provides.

Now, it could end up that more "everyday" companies decide they want to opt into this "Independent Network" too. It would probably start with other research institutions who do a lot of collaboration with these universities, but if it grows far enough beyond that, "everyone who's anyone" could eventually be on this "new" Inet, and the old Internet -- while it won't disappear -- would be less relevant.

All that doesn't sound so far fetched to me.

The 24-hour response requirement is extreme. What if you're on vacation or a storm knocked out your connection, and you return to find your domain is gone? That part of the plan has serious problems. The rest seems reasonable at first glance.

Why isn't this all possible by reforming ICANN, or is the consensus that it is too far gone?

This would never work. I stopped counting the non-starters at 4:

"A free browser patch would install a virtual switch. Click on the switch, and you route your calls through the Inet": You honestly think you're going to teach the average user what this switch does and how to use it?

"Inet's Terms and Conditions would prohibit [...] spam, denial of service attacks, and zombie networks": Because the only thing that's preventing spam and DoS attacks today is a lack of terms and conditions? And: How is a DNS system going to detect and prevent these things?

"[DNS] could have been a simple automated service run for less than $1 million per year. Maybe Bob Kahn would prefer the Inet, too": But a "simple automated service" isn't what Inet is, if it's going to involve all these decisions about whether the rigid TOS has been violated and they should be cut off. Not to mention:

"The Inet's e-mail service would [...]": How is this a "simple automated DNS service" if it includes an *email service* as well?

Two things.

1. Isn't this centralization of a decentralized system? I don't see why centralizing anything makes it better or faster or safer or anything. Seems to introduce bottle necks and many constraints just to fix problems that can be solved with new protocols and regulations, which would still retain decentralization. Or am I missing something?

2. Protocols seem like a better solution. Fixing broken standards bodies like W3C, IEEE, ISO, etc seems to make more sense. For example, IAP is an alternative to HTTP. It fixes LOADS of problems and could really work to better the Internet. The problem is that without an open, free, and fast moving standards body, innovations like this just sit dead in the water. I think a better way to fix the world is not Inet, but OFFSO - The Open Free and Fast Standards Organization. It would have clearly defined rules, a board that isn't afraid to make choices and NO fees of any kind.

This feature of Inet creates a problem and will need to be changed:

"...any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated."

I own 4 domains that host websites that contain information of interest to me (and apparently almost no one else). What happens when I take my annual vacation in Tuscany? I certainly have no intention of remaining plugged into email and all the rest of it. That is not what vacations are for.

Sneering at someone for being 'just another Bush-hater' is just as bad as what you're complaining about. Congratulations on irony.

I don't know if I agree with the proposal above; in fact I see several problems on the face of it. But to casually dismiss the idea that no one but America could manage the DNS system is, frankly, a great example of one of America's more infamous traits. There are plenty of countries in the world who could do as good -- if not better -- a job.

Telling the rest of the world you're doing us a favour while clutching such a useful tool of internet power to your breast doesn't exactly ring of altruism; but it is typical of American policy.

If this is the cream of the crop, I can see why you haven't written about any of David's other ideas. Respond within 24 hours or lose your registration? Domain disputes settled in a week? So if I go on vacation I can come back to an empty virtual home?

A couple of points, leaving aside whether this is a good idea or not:

• No browser patch is needed.


• On client machines, set the DNS server to point to one that uses the new scheme.


• On DNS servers, replace the root servers list with one that points off to the new system.




• If systems want to optimize differently for clients that come from different DNS systems, they can hand out different IP addresses for each system; any participating server can check which IP address was used for access.


• With no browser patch required, the new system is not HTTP-specific.


• If you wanted to be able to switch back and forth easily, it's a straightforward OS-specific program to write.


• If you looked at the new system as a filter and new set of TLDs, rather than a whole new DNS, it might be less scary.


• OpenDNS might be an interesting starting point for exploring some of these concepts.

That's the worst idea I've ever heard. It's a utopian pipe dream with no regard for the real-world constraints faced by such a plan. Basically, Dave Harrison seems to think that a second ICANN could do everything the first ICANN does, only a thousand times better and with no glitches. How would it accomplish this monumental task? By not being American.

Why should I have to register with two DNS hosts when I want a new domain name? Why should Internet users have to worry about which DNS host they're using? How do you expect your grandmother to take the time to learn what DNS stands for, letalone become familiar with the difference between ICANN and EuroICANN?

How do you expect to be able to resolve all disputes over domains within 24 hours? Do you have a magic wand of justice that you will use to decide every case? There's a very good reasons why disputes of this nature take a while to resolve.

And why should corporations pay more for their domains than individuals? Do corporations' rights to property ownership differ from individual ownership rights? How do you differentiate between an individual and a corporation? Does a small business owned by one person count? How about a small business with five people? Or ten? What's the magic number where all of a sudden if you hire one more person your domain names will cost a hundred times more?

Dave Harrison should spend more time investing in a good idea (not this one), and less time deciding what the rest of us should be doing.

Bad idea. I have a couple of domains but haven't put up web sites - they're not needed yet for my business (word of mouth is much more valuable), and they cost money to build & maintain. As you can imagine, I pay little attention when a domain registrar demands my attention. Dr. Harrison's scheme would force me to either a) develop and maintain these domains, a money-wasting distraction, or 2) forfeit them.

Instead of looking at the enforcement side of the problem, how about working on the technical side (look out, here comes yet another PhD who's in over his head)? Maybe expand urls with more heirarchical terms to expand available domains?

So why can't the current DNS system adopt these rules? Is it just too politically entrenched for such an overhaul, or is the point also to change the political home of the system? Perhaps lobbying to get ICANN to do these things gradually would work just as well as rebooting the entire DNS.

Seems like you just wrote about OpenDNS.

So if I go on vacation, and the registrar has a question for me and can't reach me within 24 hours, I lose my domain?

Perhaps instead of spitting out an idea a week, Mr. Harrison should buckle down and (quietly) concentrate on perpetual motion.

That sounds very much like 'Stalin's Method for Operating the Internet.' With rules that strict there would be too many cases were innocent people were punished for the actions of a few bad seeds. We need to come up with better was to stop spam and phishing in the current system. Switching to a different system will only make the way they game it different, not stop the gaming.

Yes, That's the ticket, top-down bureaucratic control. That's the normal social democratic answer. If something goes wrong it is because of a lack of control by the authorities. In light of this and with all due respect, I am not surprised to learn David Harrison is european.

What about a free-market approach. There have been many proposals to get rid of spam by making it cost-prohibitive to send. If people get angry enough and the spammers, they will undoubtedlt adopt one or more of them.

Let's not adopt a cure that is worse than the disease.

Alternative DNS has been around probably since the inception of the concept. It suffers greatly from the chicken/egg problem and ISP support (yes, as long as ISPs can still determine where their DNS servers go for the root, there won't be much change).

Check out e.g. http://www.opennic.unrated.net/

All in all a pipe dream, and a disappointingly underresearched column today. Robert, you're getting sloppy.

As I read what you said this proposal would either cost me my domain name or result in a lawsuit that would bankrupt me.

Long ago I registered my last name as a .com domain. I did not register .net because I do not offer a network service and I did not register .org because I am not a nonprofit organization. (Anyone else remember when there were rules for who got to use .net and .org?) Well, it so happens that my last name is also a trademark for some woolen goods company up north who registered the .net and .org domains and threatens to sue me every few years to try to get the .com domain. Under the scheme you describe one of us would wind up getting all three. So, either I lose the .com or they lose .net and .org. If they lose I can count on being sued.

This proposal does takes a very naive view or the reality of domain name registration.

DNS is not a trademark mapping solution, and the insistence on using it as one is itself the problem.

There needs to be a separate protocol which maps from trademark names to resources (web sites, mail servers, whatever. It wouldn't be as simple and unambiguous as DNS (one name: one owner), but trademarks are NOT simple, and the protocol needs a certain level of complexity to satisfy the needs of corporations to be found by users, and the needs of consumers to find corporations.

If users can be trained to type "tm: coke", or "tm:hilton", or "tm: just do it" into their browsers, we can drastically reduce the number of domain-related disputes. The official trademark offices are the authorities on this matter, and giving them an office on the internet might bring some peace to DNS.

http://www.opendns.com/

no software to install, intercepts phishing attempts (customers will be warned if they attempt to visit a phishing site), works well for individual computers and large networks, several redundant network connections, use 208.67.222.220 or 208.67.222.222

I've been reading Cringley's blogs for more than a year and generally found them interesting and inciteful... except for the last dozen or so articles. Has he hired a ghost-writer or something? Maybe outsourced to a committee of semi-literate child slave labourers from random third world countries... not serious stuff at all, rather a bunch of ridiculously trivial and self-contradictory ideas full of factual errors and overhyped conspiracy theories.

At least his site isn't loaded up with obnoxious advertising... :-)

If telling bad usage from the good usage is so easy, why aren't we already doing it?



Domain squatter vs lazy domain user?

Spam vs legitimate advertizing or even newsletters?

Pornography vs art or education?



If you don't like or understand the traffic coming from my PC, is it a Zombie?

Domain owner responses within 24 hours? Can't I go on vacation anymore??

If DNS names are all trademarks, should all noncommercial uses be deleted? Or moved to where? Where do we draw the line between commercial and noncommercial?



I think we have plenty of tools for stopping traffic, I don't see that we have any workable tools for telling good traffic from bad traffic beyond estimated statistical volume counting.

Sorry, but absolutly none of this is "new". The ideas have been around for years.

The main sticking point, as always is who controls it. From the article it seems that David and Bob want it.

Replacing the current system is something that will not happen until it crashes and dies. Then there will be a consortium that takes it over and creates a new order, which we won't like either.

I've been reading Cringely for a couple years also now. Most of it great to read, and insightful. But this one was blah at best.

I agree with the comments above. People have tried to create a "secondary" name system to work with the internet, some have worked OK, others have already failed. Just Google for "Alternative DNS" there is plenty to read up on.
Wikipedia Alternative DNS lists a few of such attempts.


I personally do not trust any one company like OpenDNS to "filter" my requests. Especially with the amount of gTLD's we now have, it's bound to be quite difficult to police things. DNS should be Open, Distributed, and Automated. Anyone with a copy of ISC Bind, and some knowledge can start their own System. Brining users to the table is the hard part.


I also agree with Bob Kahn, that the name system shouldn't be generating as much money as it does. But if someone waved that much money in front of your nose, could you honestly say no?

And who would run this "inet"? Why would this group be any better than ICANN?

http://education.zdnet.com/index.php?page_id=18&id=1553226&tag=nl.e623

This is the page that intoduces your article.
It has 637 words, of which 107 belong to you.
Why didn't you say NO?

http://education.zdnet.com/index.php?page_id=18&id=1553226&tag=nl.e623

This is the page that intoduces your article.
It has 637 words, of which 107 belong to you.
Why didn't you say NO?

Inet sounds a lot like Big Brother to me? 24 hours to respond to any complaint or suffer the loss of your domain? You could spend that long on a JetBlue plane on the tarmac in New Jersey. Spam is crap, ads are crap, scams are crap, but loss of freedom is worse than any of those.

Inet sounds a lot like Big Brother to me. 24 hours to respond to any complaint or suffer the loss of your domain? You could spend that long on a JetBlue plane on the tarmac in New Jersey. Spam is crap, ads are crap, scams are crap, but loss of freedom is worse than any of those.

How exactly would you enforce these rules? With millions of sites and billions of pages, who's going to check them all? As the previous commenter mentioned, what about the grey areas, porno vs art, spam vs ads, etc? Who decides? We all have our own vision of how we wish the internet would be, and none of our visions are the same. This 'plan' would just give us two internets with problems, one is enough for me.

The internet by definition is international, so when the Inet organization makes decisions regarding which companies are "fraudulent" and which are not, which country's laws are they using to decide? Many cases are not very clear cut and require the ruling of a court, and the laws in different countries are sometimes wildly conflicting. How can Inet respect the laws of all the countries if the laws are conflicting? Would this result in a virtual country, with a set of citizens and laws that only exist in cyberspace - The United Servers of Inet?

David Harrison is most likely a brilliant and insightful person who can correctly identify problems with the current system.

History has shown time and again that systems based on preconceived rules will fail. There seems to be something about the concrete nature of structured policies that make them less efficient, so that they get easily surpassed by other systems or mired in their own bureaucracy. (Citations available.)

Starting a new internet based on fixed rules is doomed to failure.

A better plan, which *might* work, is to write down the intentions for the new system and then create a method for defining policies which address the intentions but which is flexible enough to be modified as needed.

Some companies are already doing this. They define a "vision" statement which defines their overall goals and reason for being. Any decisions or policies are made with these goals in mind, with the result that rules can change as problems are identified, or as people come up with better ways of serving the vision.

Everyone would agree with the new DNS system vision statement, and it would be clear how policies contribute to the vision. Public review/criticism would help tune the policies over time, ultimately resulting in a highly efficient service-oriented system.

Flexibility is key. No one has ever been able to predict the future with enough precision to make a lasting rules system. Flexibility is the only way it can be made to work.

Staying true to your PBS roots. The evil Americans and Bushies (very juvinile, btw) are controlling the internet. I supposed the holy United Nations would be a much better alternative.

With something like 80 - 90% of UN members being dictatorships (and probably a higher percentage were all countries considered) where are you going to find a truly "open" body?

Have to say, I usually read Bob's columns and immediately send copies to colleagues and friends.

This time I more apt to question what he (or friend David) was smoking prior to sitting down to the keyboard *G*...the user comments were actually a more realistic take by far.




But what makes you think that the "Inet" would be any easier to police or enforce than the Internet?

HEY! I got it!!




We can just let PBS run the Inet DNS system...then while the sitename is being resolved, they can run interstitial ads asking for donations.




"Help us by financing a root server and you'll get this lovely mug and tote bag..."

Your joking right.

I give it 1 year before
a: it fails under the weight of the volume of web sites, complaints, appeals, etc, etc
b: to avoid a: becomes what it it replaced with a complex series of rules, regulations, political posturing, etc, etc.

Besides most people i know are perfectly happy with how their personal web works, they can get the news, email, banking, and vpn to the office without much hassle.

I don't get that much spam after getting my own domain and being careful about where I post my email address unlike my poor old yahoo account circa 1995 which has died under the weight of 1000 spam a day.

Just say No to David Harrison.

I don't know where to begin with all the problems in his proposal. The nutshell version is that he wants to replace an inefficient, unresponsive bureaucracy with an allegedly benevolent dictatorship. He considers ICANN to be an imperialist tool, and his solution is to replace it with his own imperialist tool.

Their site would be checked to see that it complies with Inet's Terms and Conditions, and if so, they get it.

There are millions (billions?) of domain names. If someone spends just 1 minute looking at a site to approve it, what year do you expect they'll get around to your application? And who decides the T&C, which seems to control what content is allowed on the Inet?

When you register a trademark as a domain on Inet, you automatically get all of the global alternatives in one go.

Does David know that trademarks are regional? If two companies own the same trademark in countries on opposite sides of the globe, which one gets arbitrarily granted global ownership, and what compensation is offered to the other? And what if a company registers a trademark that matches a domain name I've owned for years? Do they automatically trump my prior ownership?

Domain dispute resolution would be rapid: one week for evidence presentation, 24 hours to decide, and 24 hours for appeals.

If it takes you more than a week to locate the necessary documentation, because perhaps you have other responsibilities, that's just too bad. And don't even think about taking a vacation. If anyone finds out, you won't have any domains when you get home.

All domain activity would operate through Inet, not be farmed out to resellers, since the system is too important, and has proved to be difficult to police on the Internet.

Network Solutions already taught us how efficient, responsive, and low cost a monopoly registrar is. I jumped ship to OpenSRS the moment it was possible, and I have no intention of going back.

Inet domain holders would be expected to maintain control over the content of their users on sites with Inet domain registrations.

Who polices this (see Mathematical Impossibility), and who decides what level of free speech we are or are not allowed to have?

any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated.

I was just planning an international flight, and the total travel time was over 24 hours. A few years ago, I drove to visit my sister. That was a 3 day drive. I've also been known to go camping for days at a time. But nevermind travel. Sometimes I actually turn off the computer and relax with the family over the weekend. But your brave new world would chain me to my computer. Thanks.

Pornography sites could only register using the .xxx top-level domain

Define 'pornography'. Get the world to agree with your definition. We'll wait. When you're done, let me know whether I'll be needing to register michaelangelo.xxx for an art appreciation site.

All known spammers or phishers would, where identified, be banned from the system for five years or life.

With what legal recourse? Lifetime banishment is a pretty stiff sentence, even from a governmental legal system. But in a system without checks and balances, it's unacceptable. Better that the system provide the information necessary to prosecute fraud, and let the legal system with jurisdiction handle it from there.

I like it.

I've been reading your column for years. This is the first time you've scared me.

Maybe Bob Kahn would prefer the Inet, too.

Why? It's not a simple automated service, either. It's worse than a multi-billion dollar industry. It thinks it knows what's good for you. Maybe Bob Kahn would prefer you avoid name-dropping without an explicit quote of support.

The 24-hour response requirement would make Inet a non-starter for most of the individuals and organisations that would otherwise find it attractive. Many laudable ideas there, but the crushing administrative and policing costs not thought through. Freedom from bandits, frauds and abusers does not come cheap. Check out the police department budget for Laguna Niguel, CA or Chestnut Hill, MA -- John W, Hong Kong

Where do I sing up? Sounds great! Most the the rules could be put in code and automated. I'm sure the good fellows who invented DNS would have done something like this if they could have seen where it was leading to today. People making so much money out of nothing is what the DNS has become. Its very nature makes it a crook's wet dream. Down with DNS now!

I see that Apple and Cisco Systems have settled and both will have Iphone.

Sounds too good to be true - because it is.

All the "would be/should be" stuff will be spoofed the same way they are now, or more clever workarounds easily invented. The money's too big to let trivial techniques like those mentioned become a permanent factor.

But we can dream though, and remember the good old days - 3 or 4 years ago.

you said "Certainly a different DNS with different rules would not be hard to build from a technical or even a financial standpoint, and it could exist on the current network right alongside the current DNS system," and i wonder where you got your facts? DNS is a universal namespace, and there is no way it can work if its internal inconsistency rules aren't followed. an incoherent naming system could be built, and it might look a little like DNS, but under the covers it would be almost completely new. moral of story: "wishing won't make it so."

How do the lawyers make any money on a 1 week trial?

How do the lawyers make any money on a 1 week trial?

Boring old idea. Surprised? Try looking for any mailing list or newsgroup that actually discusses DNS. Many people have ideas to "fix" DNS, including this unoriginal one. Not surprisingly, just as with this one, the "fix" is worse than what we currently have.

With the idea here, we partition the 'net into two separate internets. But as we saw with Protestantism, once you start splitting, it won't end with two. As soon as you have as many separate internets as there are ISPs or Protestant denominations on any given day, you will get more. And guess what. NOBODY CAN TALK TO EACH OTHER ANY MORE.

These separate DNS roots already exist, by people who hate the people currently trying to run DNS. Try to get to a URL such as <http://itnews.newspaper.joe.america/>. You can't, you say? Well, that's because it's not available from your DNS root!

*sigh*

'Spam is not a free speech issue, it is a digital pollution and fraud issue . . .'

Sums it up well. But I'd make sure privacy (or trespass), too, is covered.

Wendell Cochran
West Seattle

I recently left the anti-spam arena after working in it for over 5 years and we have a name for proposals like this one: a FUSSP. Please check to see how many of the items on this page this proposal triggers to see why this will not work:



">http://www.rhyolite.com/anti-spam/you-might-be.html



As well, think about this for a minute: "any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated".



What?



Are you kidding? What if I was on vacation and you cancel my personal domain? Come on...

I have to admit this is pretty bizarre. David Harrison needs to add some books on IP Law to his reading list; several other posters already pointed out the huge gap in trademark handling.

Also, like the others mentioned, this whole "respond in 24 hours or die!" kidnapper motif is pretty irresponsible. I can see hundreds of ways that innocent people get bombed either through the deliberate attacks or honest mistakes.

One of the things that I've always enjoyed about these columns is the amount of back-up provided -- you always explain why something can work or will work. Here, however, any backing data is noticably absent. It reads more like a marketing pitch than a review.

A lot of things are taken for granted in the article, as well; what "spam" means, what "pronography" means, what is considered "safe" content, etc. Oh, and you forget about any site that has a social element like facebook, myspace, youtube, slashdot, etc., unless the website can afford to hire hundreds of employees to sift through every scrap of data to make sure it "complies" with Inet's demands.

Oh, and I like the fact that everyone who owns an address is required by Inet law to list their name, address, and contact information. That'll make my stalker hobby much easier, thanks! I think I'll call up David Harrison every day to tell him what a great idea that is.

Finally, the article says absolutely nothing about who runs this thing. It also has the rather bizarre thought that just because it's not the U.S., it'll be corruption-free and the image of honesty. Hello, reference desk?

All in all, a really weak column. When even I, a non-techhead, can see the gaping holes big enough to run a Peterbilt through, that's a really bad sign.

A little background on my comments. I sell computers for a living, so I am on the front lines of computer use. I get to see people who've never touched a mouse before, and people who are looking for an obscure part available only in a single store in Tanzania. My question on Harrison's proposal is why would the average user switch over to Inet? Certainly, there are advantages, but to those who just want to buy some plane tickets and send photos, this is an unneeded service. It's one more thing to concentrate on. I can imagine the questions, "Now I need to figure out which Internet I use? There's supposed to be only one!" Don't laugh, I get those sort of questions. I suppose a starting point for such a change would be for this network to be aimed at users with a working knowledge of computers, and would branch out to the much larger and much less experienced portion of the population. But if it can't make that jump, and it might not, since most people can do basic things without major issue, this idea might be dead in the water. This is simply one peon's opinion; feel free to disregard.

Cringely has totally lost his knowledge of internet technology it seems. HTTP is not the only protocol to use DNS, and the web browser is not the only application either to use HTTP.

And to switch from DNS to another name resolution protocol just begs to have the evolution of DNS yet another time, and to resolve all the same problems all over again.

A few things strike me about this article. One is positive; there is an awful lot of dead wood and on the Internet as we know it today - abandoned sites, domain squatters, all that stuff. If DNS were to be replaced or even cold-started, we could maybe get rid of all that junk and speed the whole thing up.

However, you mention Internet Video as a particular cause of bandwidth saturation. Assuming all the content meets the 'legal' stipulations for joining the Inet, wouldn't that problem just follow us to the new structure?

And if David were to gather enough support to launch the Inet and it became even a moderate success, what's to stop anyone else doing exactly the same thing? If we start down this road we could end up with hundreds of 'little internets' and possibly lose the global nature of the 'big Internet'. It wouldn't really be the World-Wide Web anymore.

I also agree with the earlier poster who commented that the '24 hours or die!' rules were too strict. What if someone hijacks your site for nefairous purposes while you're on a backwoods fishing trip 300 miles from a web browser?

I'd be in favour of an overhaul of the current system - maybe ICANN could be reformed in some way - but I think any new version of the internet that starts up (and I'm sure it will happen in time) will be a stunted version of the real thing for an awfully long time.

And to switch from DNS to another name resolution protocol just begs to have the evolution of DNS yet another time, and to resolve all the same problems all over again.

this Inet idea looks, sounds and tastes like a lateral way for taking control of the Internet, something the EU (and not only) governments have tried to do for a long time. socialism applied to internet.
I dont think internet needs a marxist-like revolution to give back the people ownership of the domain names. and Icann, with all its defects, has allowed a huge and more or less free growth of the Internet. It won't solve anything, just hand over control of the Internet from the hated americans to the smart, snobby europeans. who dont have a clue whatsoever about freedom or free market. no thanks. ill keep my old internet.

http://new.net/ tried/is trying a partial step in that direction. Not sure it is any cheaper and given the amount of time that it has been around it doesn't play well to the willingness of people to switch their DNS. Most don't 'get it' enough to understand -- even with the simple plugin.

Just a thought.

Steve

hijacking DNS is not remotely "replacing the Internet", nor does it fix any of the particularly difficult problems.

the biggest challenge in any scheme is the need to infer the recipient's interest in receiving a given packet without them having to pre-arrange that reception. implicit in this is the ability to accurately identify the human originator of an activity as well as a priori impute their intent in real-time.

if we could do either of these sufficiently well in the real world, a great many things besides "the Internet" could be addressed with much greater efficacy.

I don't expect it would be quite so long before this new Inet too would be compromised by an entire host of clever workarounds.

I much prefer the idea of a 1-penny per email and/or video transfer international bandwidth tax direct-billed to the originator or the URL port-of-entry at national boarders.

Spam would quickly become un-economical at even this ultra-low rate. This would cost me personally about $10 a year. I would pay it and gladly.

While the idea is good-intentioned, it is evident that it was put together by someone with a rather dull life who never goes anywhere and does little. There are some draconian measures proposed here that could destroy a person's livelihood with no warning. Most disturbing is the notion of any communication from Inet must be answered within 24 hours. So, an independent businessperson (photographer, writer, even a freelance IT guy) can get on a non-stop Continental jet in Newark, land in Singapore 22 hours later, go through immigration, collect his luggage, get to his hotel and discover that his domain has been deleted because he wasn't in constant contact for 24-hours? What about weekends? What about holidays? What if some mom-and-pop store's computer breaks and it takes a couple of days or even a week to get fixed? What if a person has their own domain for their family pictures and they go on vacation for a week? Does it not occurr to the inventer that some people do more with their lives than stay chained to a desk? Can't wait to see millions of European domains get poached when they avail themselves of their 7+ weeks of vacation!

It seems this Inet idea is business oriented. Some of the implications of this setup fly in the face of the internet and attempt to eliminate some of the core ideals that create the value of the internet in the first place.

I posted a much longer reply on my blog found at http://www.foreverdean.info/kevin/?p=17 .

Good but old idea. Many of us are already using this tactic via our own private URL lookup engines that are not part of the regular DNS hierarchy. This is being done both in the open and by using secure VPN architecture. There are several fair sized business and hobby entities like this already in existence.

Open Source DNS servers and open Source Browsers offer the opportunity to customize these such that they can be either prioritized or walled to build virtually (!) any private network you desire, all within the existing Internet backbone structure.

Arv
_._

Sorry, I don't buy it-- we'd be trading a mixed imperialist/capitalist system for a monopoly, and for what reason, how does it benefit me?

• 
  Choice and cost: I've owned a domain for several years. The capitalist system allows me pick from a plethora of registrars according to the terms I desire, and I've always paid a low-low price. Why would I want to give that up for a monopoly?
  
• 
  Domain name dispute: I have never had to initiate a dispute, and I would suspect 99% of individual owners are in the same boat; there is no pain point to incent us to switch. Ditto for corporations, since they like the current dispute rules.
  
• 
  Centralized anti-phishing and anti-spam:
  Every ISP offers email with these services that works pretty well. If I'm not happy with the quality offered by my ISP, there are a plethora of products (both fee and free) that I can install myself to get enhanced protection. Corporations have email systems with this built in and wouldn't be keen on having a monopoly peek inside their email.
  

• Browser patch: Browsers are but one application that makes DNS calls. You're forgetting about Quicken, Windows update, anti-virus, email programs, corporate VPN, etc. The only feasible way to implement would be to change the DNS code in the operating systems. You can probably drive a patch into Linux distros, but Windows would be more difficult (perhaps with an add-on device driver)? Point is, it's not as easy as just "a free browser patch".
  
• Less spam: Spam can't be fixed with an alternative DNS because you don't need a domain name to send email. Spammers often send spam from short-lived ISP accounts or networks of zombie machines, and always include bogus "from" lines. The only way to stop a spammer is for the ISP to pull his plug.
  
• Centralized anti-phishing and anti-spam: Just how do you accomplish this? Today email is sent directly from sender to receiver through (hopefully) the shortest path of intervening systems. Would all email be expected to go through the central Inet servers to be scanned and forwarded to their destination? That would quickly turn costly to pay for the necessary bandwidth and MIPS to ensure timely delivery. Would anti-phishing and anti-spam be offered as a service to email providers? What incents them to switch from what they offer today (which they can use as a differentiator vs other ISPs)?
  
• 
  Political: Who controls Inet and why should I trust them any more than ICANN? If it's controlled by any nation's government, it defeats the purpose of the proposal. If its controlled by the UN, it would be politically viable, but when was the last time the UN carried out a complex program that changed the world?
  
• 
  Timeframes: Almost seems trivial to point out, but the durations listed for the various parts of the resolution process are ridiculously short. 24 hours and your registration is terminated? 1 week for resolution? I think I'm in the norm when I say I don't read email every day and have a habit of taking vacations that last > 1 week.
  

Bob:

I think you had some sort of breakdown there at NerdCentral and were grasping for a pulpit subject.

So you cut and pasted something from David Harrison..

Don't be so uptight. If you don't have something for a pulpit, just don't publish.

The folks who read your notes want substance. Not to waste their (my) time with half-baked concepts.

David Harrison's proposal is basically sound. The 107 comments so far show that there is no consensus on the details, but those arguments need not take away from the attractiveness of his central idea.

Many of the problems that he addresses are fundamentally political rather than technical. To the best of my (admittedly amateur) knowledge, this applies to spam more than anything else. See, for example, this comment from 2003 (!): Fight Spam With the DNS, Not the CIA (http://www.circleid.com/posts/fight_spam_with_the_dns_not_the_cia/
)

Other problems such as funding the proposed "Inet" are open to solution. We are all either providers of backbone infrastructure or purchasers of it. It is surely not beyond human ingenuity to devise an efficient payment system for backbone providers to fund the Inet and to recover those increased costs from downstream purchasers (individual ISPs and eventually you and me).

I agree with the posts I've read: changing the DNS system is not the silver bullet that will stop spam or cybersquatting. You don't spot spammers and cyber squatters like that, and the sheer number of domain names in use worldwide makes the task even harder.

Yes, the world of DNS is not well. ICANN is not the greatest manager for domain manager, but excuse me, another layer of bureaucracy on top? "...hijack the current Domain Name System and replace it with something better." Please define for us this "something better". Excuse me while I watch this utopian socialism develop. After a couple of years listening to the dysfunctional organization called the United Nations declaring that it should provide Internet governance, we now have the true answer -- Inet. One more hierarchical layer, abet a layer controlled by some organization that will have the same success as attempts to ratify the European Constitution. Oh yeah, don't forget the centralized anti-phishing, anti-span, anti-free dissemination of ideas and speech that this layer will offer. Just what we need, more good intentions paving that road to hell. This is a good idea? Cring, I'm disappointed, I've always thought better of you.

Inet is a clever name?

Unless I'm missing something, this opens up the whole DNS system to competition. If the browser plugin itself is flexible enough to let Inet-alternatives work too, then any number of services could be launched - free, for-pay, ad-supported or private - each with their own set of rules and offerings.


There must be something in all this I'm not thinking through, because from here it seems to be an idea with no downsides so it should have been done years ago... Forget the "plugin" (except maybe for IE), just get Firefox/Safari alone to support it and the ball would be rolling. Or would you even need a plugin - perhaps a quick tweak to the hosts file or similar bit of voodoo might suffice.


I'm already liking the idea of a child-friendly Internet proxy run to my own decency standards rather than some Puritanical corporation's, using collaborative filtering perhaps combined with a sort of Slashdot-esque moderation. Each parent can answer a survey about what they think is cool, and then you rate a few sites every week - or even in realtime, as kids whose parents expressed similar preferences are attempting to access new sites, you help decide what they can see and get credits for your own kids' online time later on.

I always thought it was a damn shame they vetoed the .xxx thing: the porn people could have gone their way and the rest of us could go ours, without hypocricy or rancour. And I think that treating spam as a fraud issue rather than a free speech issue is a long-overdue distinction. Getting 5 or 10 pounds of physical junk mail a week on one's front porch is bad enough, but you know who's doing it, and if you get angry enough at them there is recourse. But the way spammers can abuse cyberspace with impunity is a disgrace. Personally I think it all started with free email adresses at Hotmail.
Lou

This system is flawed in so many ways...

What about the fact that you can have a company name / trademake in one country but there can be a different company in another country with the same name?
In this crossover period who gets a generic or even possibly trademark infringin name? He who got their first (in this case the most tach savvy) or the big corporation?

What about all the other protocols in addition to the browser ones?

Who is iNet to decide on what content is allowed for decency regulation? Different countries and cultures are offended by different things.

24 hours to present evidence and then a 5 year ban? Has he actually ever dealt with the problems of spam in the real world? Hacked servers or drone PCs are often the cause and for helping the guy the unknowing and non tech savvy PC user is supposed to gather evideence in his defence and get a judegement? As per users today, they probably won't even know it's going on, what the procedure is or anything.

How does this centralised DNS systm scale?

With all these quick repsonses and decisions who is going to pay the army of people required to keep this thing going?

Nominet in the UK is a not for profit organisation that is months behind in its work.

Trusting one central organisation (in this case Inet) is from a non Americans point of view the same as trusting ICANN and the US government. An act of faith that is probably misplaced.

I'm a pornographer and thus speak from a position of some understanding regarding .xxx (I've also been involved in a number of Oscar winning movies if you find the porn-connection hard to respect.)

.xxx is a bad idea which won't protect children or ensure non .xxx domains are free from pornography for a number of reasons:

1. Who decides what's porn and what's art. Can I show nudes from the Louvre on my domain? Nudes by Terry Richardson? Max Hardcore? There's no logical way to determine what pornography is, it's a matter of widely differing opinion. Many news magazines in Europe publish images which in the US are considered pornographic. Whose morals will INET inherit?
2. Visiting a .xxx domain will be impossible without circumnavigating the tools in place to protect children. Many people won't want to have to declare their interest in sex and thus be less inclined to visit adult websites bearing the .xxx 'scarlet letter'. That will mean adult sites outside .xxx will be more profitable than those within it as they can be accessed more discretly and that pressure will mean people involved in pornography will have a strong financial incentive to pevert the INET system. How much porn needs to exisit outside .xxx to make the idea of safe-harbour ridiculous and in the history of the web how many times have rules won out over hackers?
3. .kids is a better way to protect children than .xxx. It will be possible to limit some browsers to .kids domains and these can be given to any company which a website which isn't considered age-innapropriate. As pornographer want to make money, there's no incentive to penetrate a .kids enclave as there'll be no money to be made. While definitions will still be a personal issue, adults and older kids will be able to access an unfiltered internet free of such politics, and minors will be safe in an area which offers positive advantages to the people within it. Freedom and children are protected along with an adults right not to feel that an interest in sex is somehow deviant.

If INET is to improve on DNS, it should start by rejecting ideas so bad even ICANN's reluctant to embrace them. I'd be very happy to continue this conversation with anyone interested in understanding the implications (you'll find me at my blog).

Gan's suggestion (which I believe you have made before in your column, Bob) is the only thing that stands a chance to stop spam...but with the botnets and their sophistication, the cost for email would simply be passed back to the slave machine's owner, wouldn't it?

Mind, that's not necessarily a bad thing, as receiving a bill for $1000 might actually make the owner of a compromised machine do something about it.

As a fan of many years, frankly, the rest of the article was rubbish. I mean that in the nicest possible way.

The biggest problem with any technology comes from the brilliant dudes with neck beards who sit around tables eating pizza and dreaming this stuff up. They never, ever consider the bad things which might be done with their shiny new invention, or dismiss it out of hand. If you were to rewrite the way email was sent and handled from scratch today, I'd be willing to bet it could be made resistant to almost all spam without much trouble.

I assume this has been said in another comment, but this seems like something Google could / will do with their shadow net.

Although some of this sounds like a good idea it will never happen. The idea that ICANN is going away is simply not going to happen. In addition, the idea of getting world wide acceptance of anything is not likely. Even if you did manage to get this started I think you would end up with two Internet's for a very long period of time. I think this would increase network traffic and make things worse. I think no matter what you change spammers, and other scam artists will follow.

The second DNS Solution already exist and has proved to not work all that well already. The first and largest issue with the idea of replacing ICANN with Inet is that Inet will still be pushing data over the same clogged data lines as ICANN. WWW2 has become popular with military, government and educational institutions is because it really does solve the performance problem. Does it solve the spam issues not even close. Will Inet solve it, nope. I can already think of at least 1 quick way to beat Inet protections. Use a DNS scrapper from a legit domain or better a legit user to scrap all current Inet DNS entries. Now I have every Inet DNS name I need to spam up the new Inet network.

I will say for the openDNS it has been around for several years now. Though it only recently appears to have turned itself in to a business.





Reference:

http://opendns.com/

http://en.wikipedia.org/wiki/OpenDNS

http://www.techcrunch.com/2006/07/18/opendns-wants-to-watch-the-web-for-you/

When is Bob going to come into the hypertext era and place "this week's links" inline in his column? I shouldn't have to scroll up to click on something mentioned on the bottom of the column. It's really ridiculous ignoring how the web has always worked.

You mentioned replacements for TCP lately, and I seem to recall one annoucements that researchers had developed a new protocol that made much more efficient use of bandwidth for streaming content. Let's hear more about it!

A lot of good ideas but 24 hours is way too short a time to respond to an official email from init. I do have a life and am not glued to my computer 24/7. What if i want to go on holiday? Likewise the process for resolving disputes seems brutaly short i can see a lot of unhappy people but they won't be the nerds.

[quote] "A free browser patch would install a virtual switch." [/quote]

Tying the use of DNS, or any suitable replacement, to web-browsing, with the virtual switch happening at the browser level seems overly simplistic at best, and possibly quite absurd. Plus maintaining plugins for every known browser would be nigh impossible. Seems that a change like this would need to occur at the resolver level.

Oh boy. He sounds like one of the guys with the famous green ink.

Where do I start?
The Internet was specifically designed NOT to be centralised (yes I am English and that is how I spell it). This is centralisation with a capital C. All the power in a single set of hands - no thank you.

The minor problem with the required 24 hour response has been mentioned - this would kill all small businesses dependant on their webs sites. If they became successful someone would complain and shut them down - and they're dead. Great. Same for small news sites and protest sites - almost the opposite of what we want, don't you think?

There's no doubt email, in particular, is in trouble and needs fixing. And these are interesting ideas, but impractical in the extreme. I think you are falling for the "I hate all dicatorships unless the dicatator is me" trap ...

Inet = Fascist Net.

I think this is a nice idea, but far to impractical to implement.

Why doesn't everyone pull together to deal with each issue at a time. Most of the traffic on the internet is email spam, so why not deal with that issue first?

Then other issues can be dealt with after that, such as DDoS attacks, porn, domain name parking etc. I think the registering of one company name across all domain variants of that name is a good idea, and efforts to implement that should be made now, with the current DNS system.

Your "David Harrison" seems to think that ICANN is the source of all the evils in the (online) world, and that the magic wand to make them all vanish is to replace ICANN with an organisation that would be more intrusive, more judgemental and show less regard for procedural fairness or other countries' laws than ICANN does.

That sounds a dubious proposition to me.

Almost every one of the things this "Inet" would do differently is perfectly "possible" under the current system, it's just that there are strong reasons not to do such things.

• Having the Registry manually check web site content is hideously labour intensive, error prone, and subjective.
• Registering "all global variants" automatically eliminates the diversity, and with it the point of having global variants.
• Eliminating domain resellers simply enlarges the centralised bureaucracy, eliminating the cost, service and efficiency benefits that derive from market competition.
• The proposed Dispute Resolution Process would be decidedly unpopular with IPR lawyers - and probably everyone else. The strict timeframes you imagine are procedurally unfair (have you never had a two week holiday?) and an invitation to abuse of process.
• The proposal that real world postal addresses should be published for all domains is a breach of privacy - an unlawful one in many countries arguably including the whole EU - and an invitation to exploitation by criminals and crazies.
• The content controls would be impossible to police, terribly subjective, and would run into irreconcilable conflict about varying 'community standards' around the world (do you expect San Francisco and Tehran to ever agree?)
• The mixing of DNS resolution with content-analysis in many different application protocols is a bad idea for so many different technical and policy reasons I hardly know where to start.
• Splitting the root is itself a dubious prospect when so many users barely understand the complexity of what they've got, let alone why they can't reach the advertised site because the "Inet" light isn't on.
  

I'm sure there's much more wrong with this proposal that I haven't mentioned. But it's fundamentally not a proposal for a new system: it's just a set of warmed-over, passed-over ideas barely disguised as a proposal to "start again".

Jabez says "I think this is a nice idea, but far to impractical to implement". He's being too kind.

Spam and porn both are ambiguous concept: I consider my sister's "news e-mail" spam, and some people are happy to buy prescription medicine from Canada; an e-nanny service has labeled Boing Boing a porn site.

Toppling everything down is not an option, unless you want DNS to look like Baghdad; having more international voices heard at Icann might be the direction you are considering.

China expects to launch IPv6 by BeiJin Olympics, replacing a system extremely unfair to Asia: wouldn't that challenge most of the situation you describe?

That David seems to be an Utopian, disconnected from reality: great SF novelist, but none of his ideas will go anywhere unless he gets out of his cave and realize Icann is not populated by monsters with five rows of sharp teeth, three heads and a tail of fire.

In the comments, several people have pointed out that

1. this is not a new idea


2. this takes many of the problems we already have with ICANN, and makes them worse


3. you can already do some of this using alternative DNS systems or alternate DNS roots


4. the proposal as it stands is sloppy work that fails to address a great number of technical issues


5. it seems to have a number of half-baked "gee that's easy to solve" ideas that seem like a good idea for running a comittee of 12 people but that are a really bad idea for running a network of, say, a billion users, some of whom are only connected one day a week (that's "connected", not "at a computer". If you live somewhere where electricity doesn't happen every day, 24 hours is going to be an incredibly hard limit for you to meet).

What I have not seen so far in the comments is the point that this scheme, as many of these sorts of plans, automatically give up several of the features that have made the Internet work. For instance, the loose coupling of names and numbers that allows the DNS to work at all is the source of phishing attacks; but it also makes it possible to use the Internet when several pieces are broken, and allows changes to be reflected on the network in very short times. That end points don't need to be intelligent about how the network works, because the DNS provides all that, means that new services can be launched quickly: do you really think that Really Complicated Syndication would have taken off the way RSS did? And the existential threat over ICANN's head -- that if they do get too uppity, people will just set up another root, and let the ICANN root rot -- means that there's an automatic check on ICANN, even if the Bush administration or, for that matter, the ITU doesn't want to admit it.

There's another way of looking at this. Some of the best arguments for "Net neutrality" come from the observation that close to none of the really cool things on the Internet have happened because of people who already owned a significant part of the Internet before each wave of invention. Setting up a Central Overlord Gatekeeper Guy (and his mum) will ensure that such innovations aren't allowed any more. I think that's a pretty bad trade, just so we can solve the problems with ICANN, save some bandwidth, and make spamming harder.

I think we are at a crossroads. The internet is like the wild west. From here, we can do two things.

We can keep doing what we've been doing. That is, slowly injecting law and order into the wild. This is a hard and long process.

Or, we can pull out our stakes, and essentially "pave over all of the desert." It would solve many problems, but is it right?

I don't think there is a quick and easy solution that works well. This is a quick solution. It might be easy. It will, however, cause many problems.

I like what we're doing now. The world has gone to the brink in many things. The Internet. The Environment. And now we're stepping back slowly. The Internet needs to be a bit more regulated. Sure. But I don't think hitting the "reset" button will help.

In the end, we will have traded known problems for unknown problems.

The analogy I like to use with classes on the Internet is it is like going to a megastore that sells everything, Think Wal-Mart.
The isles are crowed; there are busy times and slow times. Because the store gets money from vendors to place racks of cans, or dishtowels in the middle of the isle they are going to continue to be there. People are kind or rude if you bump into them. If you know what you what, and where it is, you can get in and out quickly. If are searching for something, help is hard to find, and the help may send you to the wrong part of the store.

Well, you get the point. Changing the name of the Internet will not change it. It is still too new to give up on; remember half the people on the net are newbies.

"Or to learn from our mistakes and get it right, this time around?"

World history teaches us that this seldom, if ever, happens on a global scale.. with global being literal, or figurative inside any large organization. I just chose that one specific comment to remark on. All in all, the entire idea espoused by this article is rubbish for the many reasons listed by other commentators responding to the posting. I have read no other articles by the author, but I hope his other work is much better researched than this article.

I see a few problems:

1) 24 hour response time or your registration is terminated? What if you're on a trip abroad at the time?

2) Who gets to decide what constitutes pornography?

3) An Inet DNS bit in the request? What's to stop such a request being forged following a non-Inet DNS request?

4) gmail -- a trademark of reputable (but different) companies in both the USA and UK. Which one gets the domain in the new system? There are *many* more examples of this.

5) Who decides what constitutes a denial of service attack? There's a significant chance that legitimate network research would be banned here.

6) Fraudulent commercial services? Fraudulent according to the laws of which country? Some are obvious, and are probably covered no matter which country you're in. But there are plenty of activities that may be considered fraud in one country but not another.

7) The big one -- what real advantages does it provide over the existing setup, other that just being different?

24 hour response time? What about weekends?

I agree with the idea (replace the internet). But who could, or should, be trusted with such power? If ICANN is bad we don't want to replace it with something worse.

I liked it better the way your column used to be. Give me industry espionage, ahem, intelligence any day. Put the trenchcoat back on.

Oh, and put the soapbox away, please.

I'm not sure which is more amusing — Bob's article, or the urgency and seriousness with which people have responded to make Bob see the error of his ways!

You know, the world has more than its share of naysayers and poo-pooers — people who are keen to point out the faults in any idea that's a little bit 'out there'. These were the same kids who couldn't build anything very creative in play group, but took satisfaction from knocking over other kids' block towers. These were the teachers who insisted that you colored the grass in green, and not red. And guess what? Great ideas don't come from these people.

Now people like David Harrison are a rarer variety — people whose minds just can't help but churn out ideas. Many of them probably crazy, but hey, occasionally, out comes an idea that is so crazy it might just work!

As for Bob, well, you gotta love that line, 'I like it.' Thrown out there like a piece of raw meat into a ravenous circle of great white sharks! Thank goodness Bob's skin is thick enough to have some fun with this and digress into a bit of dreaming about a world without Internet spam and marketing scams. Can't Bob be allowed to occasionally indulge the suspension of disbelief?

Inet as proposed sounds both utopian and draconian. It seems like the resources to regulate as is proposed in this article would be prohibitive. The real gem here is the idea of a transitional browser plugin. Why restrict dns to one additional provider? Why not open source it, and allow as many dns providers as is possible to make use of the plugin for dns redirects? That way people can use ICANN,INet,IFoo or IBar, and thats only 2 flag bits, imagine if we had 32 or 64 bits to signal the appropriate dns. Then we could really leverage the power of our networks to promote freedom of information.

Those of you that are telling Mr. Cringely to "put the soapbox away" and get back to bringing the espionage or dirt should recognize that the title of this blog is The Pulpit. Any opinion that Bob decides to dish out on these pages is viewed by you with a proper warning. Not to mention you can't expect there to always be something undercover to talk about. Sometimes the hot topics are right out in the open just waiting to be talked about.

"That way people can use ICANN,INet,IFoo or IBar, and thats only 2 flag bits, imagine if we had 32 or 64 bits to signal the appropriate dns."

I imagine that what was meant was an HTTP header; stuffing it into the TCP packet would be pretty painful to pry out with some languages/frameworks. Since HTTP headers are strings they can be anything.

Bob, as a longtime and very respectful reader, could I ask you to find a way to better automate the archive page, which is always about a month behind, and please write about anything besides the Internet for just one week?

Okay, only the first really needs doing, but you must admit that 6 articles in a row on the internet (if you count the second half of the DRM article) is pushing it for a column about the whole breadth of the IT industry.

It would be tempting to give Inet a shot. Who has not cursed the spam, viruses, phishing and other security threats? This kind of stuff has even become common fare in non-tech news media.

But the rules needed to keep everything clean and legit, would stifle the germination of new ideas and new ways in which this medium is being used.

Rules like "...any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated" would make it daunting for anyone trying to get something going on the side - i.e. not as a main job.

Wouldn't Inet be too sterile for any new "seeds" to grow organically? Without complicated and expensive life support?

Isn't it the chaos in today's internet which makes it a seemingly bottomless source of new "things" - good as well as bad?

We should be patient and have confidence that the self regulating forces of an open market will function well enough to keep the internet an interesting and enjoyable place to visit.

The grass seems always greener on the other side. Until you get close enough to see there are also brown dried up spots where the dog left it's business.

Inet sound like too much work. Try this instead: use IPv6, which is here. Attach yourself to an address permanently, and incorporate a reputation system. Instead of going to unknown places, operate it on a whitelist-only, but if you want to go to a new address (be it a news site or a new acquaintance), check their reputation before you add them.

I call it TribeNet :P

You can increase it's security and robustness by making it encrypted and mesh networking-only.

There are some flawed assumptions in this plan:

How many of the negative aspects of the current internet are facillitated by the botnets, which are essentially the computers of innocent people (mostly running Windows, I think) which have been hacked remotely and turned to the evil tasks of their remote masters? And how does this plan solve that problem? It would see that huge numbers of Windows Users would end up being banned from the new INET shortly after they hooked up.

Furthermore, with the oncoming internet 2, we have enough potential issues of control and discrimination against the little guy, without needing yet another system where email is centralized and any unconventional views would be eliminated.

We need some kind of reset, but hopefully it will preserve one of the central great design principles of the internet: decentralization, and "interpret censorship as blockage and route around it"

Your solution advocates a

(x) technical ( ) legislative ( ) market-based ( ) vigilante

approach to fighting spam. Your idea will not work. Here is why it won't work. (One or more of the following may apply to your particular idea, and it may have other flaws which used to vary from state to state before a bad federal law was passed.)

( ) Spammers can easily use it to harvest email addresses
( ) Mailing lists and other legitimate email uses would be affected
( ) No one will be able to find the guy or collect the money
( ) It is defenseless against brute force attacks
(x) It will stop spam for two weeks and then we'll be stuck with it
(x) Users of email will not put up with it
(x) Microsoft will not put up with it
( ) The police will not put up with it
( ) Requires too much cooperation from spammers
(x) Requires immediate total cooperation from everybody at once
( ) Many email users cannot afford to lose business or alienate potential employers
( ) Spammers don't care about invalid addresses in their lists
( ) Anyone could anonymously destroy anyone else's career or business

Specifically, your plan fails to account for

( ) Laws expressly prohibiting it
( ) Lack of centrally controlling authority for email
( ) Open relays in foreign countries
( ) Ease of searching tiny alphanumeric address space of all email addresses
(x) Asshats
( ) Jurisdictional problems
(x) Unpopularity of weird new taxes
( ) Public reluctance to accept weird new forms of money
( ) Huge existing software investment in SMTP
( ) Susceptibility of protocols other than SMTP to attack
( ) Willingness of users to install OS patches received by email
( ) Armies of worm riddled broadband-connected Windows boxes
( ) Eternal arms race involved in all filtering approaches
( ) Extreme profitability of spam
( ) Joe jobs and/or identity theft
( ) Technically illiterate politicians
( ) Extreme stupidity on the part of people who do business with spammers
( ) Dishonesty on the part of spammers themselves
( ) Bandwidth costs that are unaffected by client filtering
( ) Outlook

and the following philosophical objections may also apply:

(x) Ideas similar to yours are easy to come up with, yet none have ever been shown practical
( ) Any scheme based on opt-out is unacceptable
( ) SMTP headers should not be the subject of legislation
( ) Blacklists suck
( ) Whitelists suck
( ) We should be able to talk about Viagra without being censored
( ) Countermeasures should not involve wire fraud or credit card fraud
( ) Countermeasures should not involve sabotage of public networks
( ) Countermeasures must work if phased in gradually
( ) Sending email should be free
(x) Why should we have to trust you and your servers?
( ) Incompatiblity with open source or open source licenses
( ) Feel-good measures do nothing to solve the problem
( ) Temporary/one-time email addresses are cumbersome
( ) I don't want the government reading my email
( ) Killing them that way is not slow and painful enough

Furthermore, this is what I think about you:

(x) Sorry dude, but I don't think it would work.
( ) This is a stupid idea, and you're a stupid person for suggesting it.
( ) Nice try, assh0le! I'm going to find out where you live and burn your house down!

Your solution advocates a

(x) technical ( ) legislative ( ) market-based ( ) vigilante

approach to fighting spam. Your idea will not work. Here is why it won't work. (One or more of the following may apply to your particular idea, and it may have other flaws which used to vary from state to state before a bad federal law was passed.)

( ) Spammers can easily use it to harvest email addresses
( ) Mailing lists and other legitimate email uses would be affected
( ) No one will be able to find the guy or collect the money
( ) It is defenseless against brute force attacks
(x) It will stop spam for two weeks and then we'll be stuck with it
(x) Users of email will not put up with it
(x) Microsoft will not put up with it
( ) The police will not put up with it
( ) Requires too much cooperation from spammers
(x) Requires immediate total cooperation from everybody at once
( ) Many email users cannot afford to lose business or alienate potential employers
( ) Spammers don't care about invalid addresses in their lists
( ) Anyone could anonymously destroy anyone else's career or business

Specifically, your plan fails to account for

( ) Laws expressly prohibiting it
( ) Lack of centrally controlling authority for email
( ) Open relays in foreign countries
( ) Ease of searching tiny alphanumeric address space of all email addresses
(x) Asshats
( ) Jurisdictional problems
(x) Unpopularity of weird new taxes
( ) Public reluctance to accept weird new forms of money
( ) Huge existing software investment in SMTP
( ) Susceptibility of protocols other than SMTP to attack
( ) Willingness of users to install OS patches received by email
( ) Armies of worm riddled broadband-connected Windows boxes
( ) Eternal arms race involved in all filtering approaches
( ) Extreme profitability of spam
( ) Joe jobs and/or identity theft
( ) Technically illiterate politicians
( ) Extreme stupidity on the part of people who do business with spammers
( ) Dishonesty on the part of spammers themselves
( ) Bandwidth costs that are unaffected by client filtering
( ) Outlook

and the following philosophical objections may also apply:

(x) Ideas similar to yours are easy to come up with, yet none have ever been shown practical
( ) Any scheme based on opt-out is unacceptable
( ) SMTP headers should not be the subject of legislation
( ) Blacklists suck
( ) Whitelists suck
( ) We should be able to talk about Viagra without being censored
( ) Countermeasures should not involve wire fraud or credit card fraud
( ) Countermeasures should not involve sabotage of public networks
( ) Countermeasures must work if phased in gradually
( ) Sending email should be free
(x) Why should we have to trust you and your servers?
( ) Incompatiblity with open source or open source licenses
( ) Feel-good measures do nothing to solve the problem
( ) Temporary/one-time email addresses are cumbersome
( ) I don't want the government reading my email
( ) Killing them that way is not slow and painful enough

Furthermore, this is what I think about you:

(x) Sorry dude, but I don't think it would work.
( ) This is a stupid idea, and you're a stupid person for suggesting it.
( ) Nice try, assh0le! I'm going to find out where you live and burn your house down!

Someone has probably already pointed this out, but what your proposing is basically an alternate root DNS service.
This has been done before, and in fact there are already many in place.
Have a look at Wiki http://en.wikipedia.org/wiki/Alternative_DNS_root

Its not a bad idea, but it isn't original. And it has the problem of having to get a good percentage of users to buy into it before it becomes viable.

If SPAM is a problem, we need to look at the protocol. Spamming is easy not because the root dns is bad, spamming is easy because the SMTP protocol is too open.
I have been fighting spam as an administrator of mail servers for about 10 years now, first at an ISP with 40,000+ dialup users, and more recently for a large corperation.
We are using kludges to try and correct the shortcoming of the SMTP protocol.
Maybe SMTP needs to go the way of GOPHER and GEnie.

My head hurts.
This idea sounds like the possible beginnings of another new.net tragedy. Count me out.

>>The Inet's e-mail service would incorporate centralized anti-phishing and anti-spam techniques, and would block known spambots.

So in addition to replacing DNS, this also requires replacing SMTP! If not, then spam will go back to spoofing the sender, appearing from walmart.com[.inet], with links to legacy, non-inet domains. If so, then we're back to well-worn issues of replacing hundreds of millions of existing email clients.

Just on the DNS side, there are over 100 million domains used for active web sites and at least that many used solely for email addresses. Thus any scheme like this would have to support legacy DNS for years, which would completely undermine the point of this exercise.

Sorry Bob, but this clearly isn't a fully baked idea.

The whole .xxx tld is ridiculous. Obviously the .blah hasn't really maintained a meaning. For instance, other than .gov and .mil sites the .com, .org and .net variety are all the same--there's little meaning to the tld name. You could argue that starting over in a new dns system would allow for this to be fixed, but who's going to decide when someone gets a .xxx and when someone gets a .com etc. It would only replicate bureaucracy. If it's open (Independent) net you can't filter sites by content.

ChrisD is right regarding SMTP. Email just simply started out extremely open and it's coming back to hurt us. We can only hope that eventually spam filters will eliminate spam and seriously reduce the interest in using spam "marketing".

Wow, I've read a lot of the comments, and they do a hell of a job shooting down this idea. However one thing that I have not seen in the comments is anyone that has any idea about how complex and secret the root servers are. From Wiki: "each DNS server in a given site is actually a cluster of servers behind a load-balancing set of routers." That's just for starters. There are 13 main servers (sites), and they are often under attack. The fact that the DNS system is a single point of failure for the internet, and it hasn't failed makes me think if it ain't broke don't fix it.

One other comment, the main reason that a huge part of the internet sucks is because it is cheap. If you make it more expensive a lot of those things go away, but then you also make it more expensive for that guy with his little hobby site who gets 20 hits a week. Only when you decide you want to take the internet away from the little guy, do you get rid of a lot of the problems. Those problems, by the way, will be every bit as prevalent in this new Inet that this guy is proposing.

@ "When you register a trademark as a domain on Inet, you automatically get all of the global alternatives in one go. So when Wal-Mart registers walmart.com, they'd get all the similar domains automatically."

What happens if one person/company owns trademark X in one country and another person company owns trademark X in a different country? I assume each would get the TM for their respective countries, but what about "all the other similar domain names automatically." Would it be a race to the registrar? A coin toss? Who has the most money to spend on lawyers? Something else??

Cheers,
Chrisco

Bob, are you feeling okay? Proposing placing a "czar" in-charge of censoring the net worldwide isn't like you at all.

I switched to using this for my DNS service, works quite well compared to my ISPs DNS.

http://www.opendns.com/

Good article Bob. Everything sounds good except the mandetory .xxx domains for porn sites. Imidietly, this strikes me as censorship. And secondarily, I can imagine some practical problems. For example, what about an art website with a .org domain which happens to display images that could be construed as "pornography"?

There's no doubt that the current DNS system is screwed up. This looks like a good start.

In addition to all the other reasons given why this wouldn't work, I would note that spammers would be able to continue to send spam by connecting directly to machines' IP addresses. Even today this occurs -- I've seen multiple reports, and have experienced it myself, that spammers continue to connect to SMTP servers via IP address even long after all DNS records pointing to that server have been removed (and cache times have expired). They'd still need to harvest the addresses, but they could continue to use the same illicit techniques they use for this today, and very few of them would be significantly impacted by the enhanced ability to turn off people's DNS service.

And in cases where the spammers are advertising websites as opposed to phone numbers, mailing addresses, or pump & dump stock names, they could continue to advertise sites on the normal Internet, and only those few people who had turned off all lookup of addresses on the regular Internet would be unable to get to the sites. Even if we got to a point where the majority of users had done a hard switchover, URLs of such sites could be given in the form of IP addresses. Users running special software to block access to IP-only URLs would be cutting themselves off from legitimate sites including Google Cache, some home routers, etc.

"Inet DNS registrants would have a real name, address, and contact details (not a PO Box), and any communication from the Inet DNS system to the named registrar must be answered within 24 hours or the registration would be terminated."

Yeah. You had me until this part. I personally only get mail at my PO Box. Large parts of this part of the state only get email at a PO Box, including a tiny town named Page, AZ. Thousands of residents on the reservations only get mail at their PO Box. The post office doesn't deliver to 1/4 mile past the big outcropping near Bitter Springs. It's the Bitter Springs contract Post Office in the back of the convenience store or nothing. The "Patriot Act" has not only not found Ossama, it's made things worse for residents whose physical address really is something like third house from mile marker 326, Highway 89.

Then 24 hours to respond? This might work in densely populated areas where you don't ever go far from home. But again looking at my area, I travel constantly. I have a domain, but I don't carry a crackberry or anything like that - and truthfully I like getting a break from the 'net. This would mean I have to constantly monitor email or something regardless of what's going on with my life. I'm not big enough to have a staff - really for me it's part time. This would make a full time job of being "available".