Creating the New Public Network 104
Codeine writes: "Tom Lyons argues persuasively that the incumbent competitors might be incapable of delivering an utility IP network. Competition in such commodity markets encourages the breaking of connectivity, ``Connectivity is the fundamental service of the Internet, yet it is connectivity that suffers first when network providers compete for users and services.'' Thus he proposes the Institute for the Promotion of the Internet Protocol Utility."
conected (Score:1, Interesting)
Uhh, what should it be solved with? Good will and intentions? Without the blocks, you cannot build a skyscraper.
As long as I have conectivity where I am and where I will be going, I'm fine.
Internet Building Blocks? or House of Cards? (Score:1)
With UUNet dark, the remaining network lacks the switching capacity to handle all of today's traffic (it barely can handle today's traffic without packet loss monitored here [internettr...report.com]), much less short term growth as the world economy recovers from the dire recession. The resulting high packet loss would take us back 5 years where many DNS lookups timed out and simply failed due to high packet loss, and the network loading is dominated by 100% to 300% retries cascading into congestive failure (RFC896 [rfc-editor.org] Congestion Control in IP/TCP Internetworks. J. Nagle. January 1984).
There have been many people explore this issue, some very excellent papers (Quality of Service in the Internet: Fact, Fiction, or Compromise? [employees.org] by Paul Ferguson and Geoff Huston) - but largely missed are very basic architectural issues like NTP time syncronization network wide for packet loss retransmission that CREATES well synchronized additional packet loss. This happens because the retranmissions are all timed to arrive at the same time in overloaded switches just to be dropped again due to servers having their scheduling clocks syncronized at a very low rate of 50/60/100/1K Hertz.
A study I did in 1997 of peering point packet loss showed that 90% of packet loss observed correlated to retransmit clock boundries. Changes in traffic flow from primarily mail and ftp in the early 90's, to web traffic where browsers launch 4-20 concurrent small file lookups changed the nature and ability for Slow Start to be effective in throttling loads causing packet loss (web browser designers flood requests to mask packet loss timeouts) and the short files which are often only a couple packets in length do not throttle with TCP window size controls.
Nothing in the next generation design of the internet (IPv6, VoIP, Streaming UDP MP3's, FPS games which flood packets, or any other new protocol) addresses these critical failings
Tom's suggestions largely miss the boat, for all the wrong reasons - but the end conclusion is correct - the biggest problems tomarrow are not going to be solved by the solutions being offered.
it's the age old question (Score:3, Interesting)
Re:it's the age old question (Score:2, Insightful)
"Public involvement for the public good" can and does work, as was proven by John Nash's work on Game Theory that earned him a Nobel Prize for his paper on, if I recall corrctly, co-operative endeavour in game thoery.
IANA Mathematician, however..
Re:it's the age old question (Score:2, Insightful)
John Nash's work on Game Theory did not, IMO, have much to do with "public involvement for the public good". It basically attacked John Smith's notion that in general, the best outcome is achieved when all participants do what is in their best interest. Nash basically demonstrated that the best outcome actually MAY occur when all participants work together. This is hardly, I think, blanket support for "public involvement", rather, it supports the notion that if all the Tier-1 ISP's worked together instead of bloodthirstily competing with each other, the best outcome could result.
However, as most Tier-1 ISP's are publicly held, the shareholders do not really care about "the best outcome" for all involved, they want THEIR Tier-1 ISP to WIN COMPLETELY, and, obviously, have not read any of Nash's theories
Any links to both works? (on thread/ offtopic) (Score:1)
Most people don't know what is in there best interest.
Mr A reads John Smith and decides to do things that are good for 'HIM'
Mr B reads John Nash and decides to do things that are good for every one.
Mr C reads John Nash and John Smith and decides what is good for everyone is good for him and dies in a paradox of madness selfishly helping everyone to a better world (am I really helping 'everyone' or am I just helping my self )
Re:Any links to both works? (on thread/ offtopic) (Score:1)
Short distillation [lucidcafe.com] of his life on Licidcafe, along with some linkage.
biz/ed has topics on both Economic Systems [bized.ac.uk] and Game Theory [bized.ac.uk] which include lots of relevant material.
but i agree with your synopsis. Mr C is screwed.
Re:Any links to both works? (on thread/ offtopic) (Score:1)
Makes sense to me.
I'd suggest that Mr C should try to improve as much of the world as he can, while not killing himself at the world's expense, and take occasional days off for good behaviour.
An example from Nash himself (Score:1)
Imagine you had 3 radio formats with different percentages for listener base:
Pop - 50%
News - 30%
Country - 20%
With only one radio station the best choice is obvious run pop and grab the 50% market share.
With two radio stations the best choice is more difficult: they can both go after the pop market and get 25% each or one can agree to do news while the other does pop. This allows them to collectively do better. That isn't a huge problem yet, since the guy dropping half of pop (25%) to news (30%) is still gaining market share.
With three stations the situation becomes even more interesting. The first two get pop and news. However the third station still does better getting half of pop (25%) then running country music (20%). Only via. some sort of cooperative agreement can the three maximize their common profits. That is cooperative agreements can outperform individual maximization.
Re:An example from Nash himself (Score:1)
did make some kind of agreement to split the
market like that, they would be shut down for
anti-trust violations
Stable enough? (Score:4, Interesting)
Re:Stable enough? (Score:2, Interesting)
that's just my guess anyways.
Re:Stable enough? (Score:1)
No. Since we are at the moment in the transition phase between version 4 and 6 it should not be considered stabilized yet. Some time in the future when IPv6 is being used on most of the Internet, and has been stable for some time it will be time to ask that question again.
Re:Stable enough? (Score:3, Interesting)
are we really ready for it to be a utility?
I'm ready for it to be a utility.
aren't there some benefits to instability, such as innovation?
Indeed there are, but you don't need to dissolve all standards and semblance of order to create a nurturing environment for creativity.
Roads are an example. They were used for foot traffic and for horse and buggy and yet the innovation of the automobile was able to build upon the network of roads quite handily.
And, that innovation of the automobile still doesn't preclude other innovations (robotic aerial drones for package delivery is my favorite) that do not make use of the road network.
Likewise with digitial communications. I think the IP utility is a good idea with plenty of mileage left in it for new ideas. If someone does comes up with a dramatically better idea for digital communications networks, then I suspect that one of the hallmarks of its success will be the ability to easily piggyback legacy IP on it.
I'm not sure I understand the basis for the authors criticism of NAT. I think it's a matter of degree; that NAT can be a sufficiently good multiplexer of services that my view of the rest of the network is not overly inhibited. OTOH, security considerations for firewall NAT tend to start with the assumption of "start with it disabled", so possibly inhibiting new services.
The way the world works (Score:1, Insightful)
I think it is telling that so few today believe anything other than greed and threats to personal power and prestiege can be motivators.
Re:The way the world works (Score:2)
And competition is, pretty much, a race for self-improvement. Chasing after others is largely fruitless, and the obvious measures against them, such as predatory pricing, are frowned upon by the "referees". Violence as per the old railroad wars is even more of a no-no. Hence, well-run businesses try to increase efficiency, expand their product lines, and so forth, instead of, oh, property destruction and sabotage; today, that's the job of the unions as they strike.
Re:The way the world works (Score:3, Interesting)
(For those unfamiliy with the term "Sin Nature," it is a Christian theological term. It refers to human being's innate tendency to do that which is evil - to tend to lie, cheat, steal, destroy, etc. in pursuit of their own desires at the cost of the desires of their fellows. In this belief system, it is a fundamental building block of human nature. I think most (though not all) modern thinkers agree with some mutation of this idea, though they may decide to call it by a different name.)
Both of these systems, which dominate the world scene now due to their success, work because they channel our natural tendency to do what is good for me into doing what is good for all of us. The capitalist wants money or power or recognition. He/She does this by building a company (which provides jobs which benefits other) which produces a product (which meets a need of others) and brings in money (which pays for the other stuff and is also collected in taxes to support the many). The key is to make sure the structure is put together correctly so that this works. It is the strengths which lead to the unprecedented wealth and prosperity seen in First-World countries. It is the weaknesses which lead to problems like Enron, etc.
This post is a bit long. The basic point is that competition is not bad. It is simply a force which our society has harnessed for good. Don't try to end the competition. Try to put a structure in place which will make the compeition work for society to produce the Good you are looking for.
Re:The way the world works (Score:1)
Re:slashbot found dead (Score:1, Offtopic)
Re:Do you Opera users love it wide? I know I do! (Score:1)
Capitilism comes to the comunity (Score:2, Flamebait)
As soon as someone realises that there's money or power to be made(and lots of it) a once free (as in speech) technology becomes market controlled and regulated, in general the overall network gets dumbed down, and all but obsolete as new technologies come along.
Here are a few examples from the past few hundred years or so..
Science-Art, initially sciences and arts were free (thousands of years ago), they go locked down and made illegal in part for a hell of a long time.
snail-mail, this is a very good example of a vanishing technology, currently being opened up to full competition in the UK with the possibility of increased prices and a poorer service(YMMV).
Good old radio,
In the beginning anyone could put together an AM radio and broadcast anything, now a days they even control what you can listen to, and the airwaves are all sold out.
The Telephone,the common telephone has started to vanish in the UK, cell phones and broad-band are replacing standard telephone technologies almost to the point where telephone networks no longer operate for voice communications.
There's a very long story here especially in the US with bell labs and all that,
I believe there are lots of regulations in place if you ever wanted to start up a Telco. The selling off of the air waves regulation of other networks by governments makes telephone networks highly capitalised.
And now the internet,
Getting less private and more dictated, the market is not yet saturated enough for it to make a big difference to Joe public and corporations and governments are looking at taking over before Joe public realises what there going to miss.
"market economy"? (Score:1)
There are some services that become so called utilities,
In the UK they are/used to be
Water, electricity, gas, the telephone, mail delivery, public transport, air traffic controle, the police, fire service, hospitals, schools etc...
Most of which were sold off during the eighties, would you like Law to be controled by a market economy? if so would you accept the Law makers tieing you into there variant,
we'll shoot anyone that looks like they might be a criminal, make up new Laws when we think were getting ripped off, have an elegable EULA and driving out everyone else out
Re:"market economy"? (Score:1)
Identifying utilities (Score:2)
At least 80% of utilities could have been recognised as Utilities within a few years of startup (or less).
Utilities usually place a levi on Businesses to reduce the cost the the consumer, an example could be a 1% tax on revinue earned from pesonal data, going more-or-less directly to network infrastructor providers.
Where has he been? (Score:5, Insightful)
Re:Where has he been? (Score:2)
So is every (other) public utility (e.g. postal service, telephone, etc.). Do you think mail goes to and from China uncensored? I don't see how making IP a utility will be so much different.
Ah, spam. Maybe this could actually help fix the spam problem to a degree. After all, there will certainly be more regulation introduced worldwide if it's considered a utility. Other than that it's not clear what you are trying to say with this. You mean, that blackholes will still exist? I imagine that if delivery is regulated this could very well change, but in spite of that there are still "blackholes" where you cannot get some current utilities already. E.g. parts of some major cities often have no 911 service. So, again, no difference between current utilities and IP.
This is why the customs department, NSA, CIA, etc. exist in the first place. While they may also be dealing with things that are none of their business, they also handle security and threats from rogue nations. Thankfully I can still receive snail mail from a rogue nation even though the sender might have ill intentions. Customs regulations ensures that it probably will be carefully inspected before coming to my door. Likewise with telephone calls. Yes, I know, invasion of privacy, etc., etc. but we have been living with this type of thing for many years WRT our postal and telephone services and I don't see that changing anytime soon.
Re:Where has he been? (Score:2)
If IP becomes a utility - then Customs will probably gain control over a huge quantity of international bandwidth...
Public Utility has one very negative effect: Incompleteness. Companies (usually) attempt to reduce the number of workers/amount of work by improving the reliability and self-healing nature of their systems. Public works are interested in providing the service, but also continuing to provide jobs. Take the Dept. of Transportation...
The DOT is constantly repairing, replacing, reconstructing roads. In high-traffic areas, they try to do a very good job and ensure that the road lasts for some time (you don't like to tie up huge numbers of people with construction). In lower traffic areas, the roads are given far less care during building (I do not expect secondary roads to receive the same amount of maintenance/care as primary roads). If more effort was taken during building, the secondary roads would last much longer, age better, and ultimately be a much better investment. But low-maintenance roads don't require as many workers. They don't spend the budget, ensuring you will get the same amount of money next year (I won't even start on that travesty of good sense).
In short, public works provide decent service, but they sacrifice quality of continuum... Fiber wouldn't be run when copper would do... Sure an OC-3 is slower, but the equipment is cheaper, and if we buy several of them, we will eventually have the same bandwidth.
Spending less money each year and more money overall may be fine for buying a new car, but not for buying a new Internet.
Current environment and history (Score:2)
On the more historic side, apparently nobody remembers The Source or any of the early highly-proprietary online services. Even CompuServe, AOL, and Prodigy only survived by becoming Internet portals, though all but AOL are all but gone. People simply didn't want to be locked in. They wanted the "IP Utility" that the Internet originally offered. Ever since the Internet was privatized, there's been a tug-of-war to turn it back into a proprietary cash-cow, despite the teachings of even recent history.
But then again, we went to the Moon, and threw all of that away.
IP Fun! (Score:1)
'Ip 'Ip 'Ooray
Balancing rights (Score:3, Insightful)
It seems that the key arguement towards making internet access a utility is to remove onerous clauses from the contract, similar to consumer protection laws, existing utility legislation, and tenant rights laws.
I think this is good, charge for speed or data transfer.
But what about spammers flooding and other hostile attacks?
Removing the ability for the corporation to limit user behaviour would requite the government to limit user behaviour, with the current situation (MPAA, RIAA, DMCA, and others issues of course) we may want to be careful what we wish for.
Re:Balancing rights (Score:2)
plus spam is generally a few k of text, not much data at all
heh (Score:4, Funny)
How do you pronounce IPIPU? I-Pee-Poo?
</joke>
"AN utility" (Score:1)
How does one get to be an "editor" if one is illiterate and/or too lazy/arrogant to use a spell checker?
Public Utility (Score:5, Interesting)
Any such regulation, will also have to regulate things that are not in the public interest, because the public utility is for the benefit of the public.
If this were to happen, how are we going do to decide what is in the public interest? We have a real hard time even with the sample of people that is slashdot deciding what is in the public interest. We could find that many things we enjoy about the internet (its anonymity, its freedom, its ability to share information) might become regulated for the public interest. We have all heard this argument before and what is happening in Australia [slashdot.org] is a perfect example [slashdot.org]
This may sound like a paranoid rant, but I think its is something people should consider, before we make this kind of decision. Many bad ideas in the world started out as good ideas....
Re:Public Utility (Score:2)
If people believe that "mean, evil ISPs" which exercise "undue influence" over politicians who allow themselves to be bought (e.g., Senator Disney) will be remedied by some kind of regulation, they are wrong. If the issue of regulation even comes up for a vote, these same corporations are going to simply buy enough senators and congressmen so that the regulations are written to favor them. The end result will be regulations which effectively keep the formerly unregulated companies unregulated still, but now legally bar any new competition from springing forward. That new bill that is going to outlaw fair-use is a perfect example of this.
Now I don't blame the corporations for this, they are always out to protect their own interests, just as we all do. I blame the legislators who allow themselves to be bought for this problem. If something isn't for sale, there is no way anyone can buy it.
The Social Contract (Score:2)
This is the whole problem addressed by the concept of "The Social Contract"
Problem being, people have gotten so used to the system that they treat it like a video game, trying to get as many trinkets out of the system as possible, instead of working together for the greater community. Thus we have things like the preamble of the US Constitution:
- We the People of the United States, in Order to form a more perfect Union, establish Justice, insure domestic Tranquility, provide for the common defense, promote the general Welfare, and secure the Blessings of Liberty to ourselves and our Posterity, do ordain and establish this Constitution for the United States of America
This is an attempt at a social contract. This was hot and radical political thought at the time it was written.These days people some people might not relate to this. but the issues are very relevant.
Just take a look at your question.
Re:The Social Contract (Score:2)
It is not very often that people even think of "working together for the greater community" as you have put it, and this is too bad.
Re:The Social Contract (Score:2)
Often people who believe in "working together for the greater community" really mean "working to put my particular centrally-planned socialist idea in place, even if it reduces 'the blessings of liberty' or 'the general welfare'."
Need IPv6 anyway... (Score:1, Insightful)
If it's not broken .. (Score:2, Interesting)
danger of fragmenting into parts controlled by
separate companies that are unable to communicate
with each other, and that the solution to this
'problem' is a single centrally controlled IP
utility. Yet he provides zero evidence that this
is actually happening!
Because there are so many ISPs and carriers, none
of them would dare to cut off connectivity to
each other. Maybe if there was some mega-ISP that
controlled 90% of the market then it would make
sense for it cut off competitors
not the case today.
So what exactly is the current real-world problem
that this 'IP utility' is supposed to solve?
Re:If it's not broken .. (Score:1)
The reasons for ISP's not providing a purely raw IP service are many ranging from bandwidth constraints (and profit) to security, however, unless there is some supervision (either by consumer demand or govt regulation), I see no reason for this trend to improve for the consumer.
Re:If it's not broken .. (Score:1)
customers from over-using 'unlimited' broadband
services. If they didn't exist, ISPs would have
to start charging by the byte instead
be a good thing.
Re:If it's not broken .. (Score:1)
In general this is due to physical limitations of the infrastructure. In particular in the case of broadband the system is piggybacking on the cable network and thus can handle much more downstream traffic than upstream. So http requests work fine but web servers don't. An easy way to see this is note that DSL which is symmetrical doesn't have there types of limitations. I don't see how regulation is going to resolve a fundamental design issue like this.
Cable modems are cheap because they used existing infrastructure which was "close enough" to what was needed for the average home internet user. If the average home internet user becomes a user who requires high speed uploads and downloads, that is similar to the average business user then the nation is going to have to implement an entire infrastructure to support this which will be expensive.
Universal Satellite Network (Score:2)
It would offer universal connectivity. Ideally, it would use IPv6 as the network bases. It would be a separate network but still have built in gateways to the old internet (IPv4). That way you could preserve the function of old network while building the new network.
Everyone will simply apply for a free bank of IP#'s that follows them anywhere they go(111.222.33.44.XXX.XXX). XM radio already can transmit to terrestrial sources. Digital cell phone technology could be modified to transmit to the same celestial sources. Ideally a unique biological identifier would be used to associate your bank of IP number's with your identity.
One word (Score:3, Insightful)
You don't notice it on your XM radio, because it's all one way. The various satellite IP systems I've seen have played rather scary games with the network stack in order to get some semblance of performance (and even then, not nearly as good as cable or DSL).
Re:Universal Satellite Network (Score:1)
I can see it now. I'd hate to be the first poor victim of an IP spoof... talk about identity theft!
By the way, satellite phones exist. They are hella expensive, and most are seriously clunky. I prefer the convenience of my tiny Nokia 8260 to the satellite phone I carry in my truck...
Re:Grammar (Score:1)
As anyone who plays scrabble knows, that's because L is actually spelled 'el' or 'ell'.
See here. [dictionary.com]
Pure rubbish. (Score:2)
Anyone else see something grievously wrong with that? The way to compete for users is to deny them the product/service they're seeking? Preposterous. No one, not DSLcos, cable companies, other ISPs, is going to abridge your connectivity and get away with it. Not in the long run, and not without the aid of the force of law.
So Roadrunner has decided to block Kazaa. Any of their customers that really care about it are going to jump ship. But the real culprit there is not business but government, since if there were no potential legal trouble looming (trouble which is brought on by bad law, not bad business), there would be no incentive to block Kazaa or any other service. Some will point to the "IM wars" as an example of broken connectivity. Bogus. In IM, the nodes connecting aren't
Other than that, I can't think of a single example of connectivity-breaking. On the most basic level, the more a service provider limits the usefulness of the internet, the less value they provide their customers, giving their customers an incentive to switch to a competitor or do without.
This guy failed Econ 101.
Re:Pure rubbish. (Score:2)
I'm receiving broadband in my home that would have cost thousands of dollars per month not that long ago, for $40/month! I don't see why everyone is complainng (except, of course, for the reason that everyone loves to complain.) You don't get something for nothing.
Obviously home broadband has come a long way, and there will be stumbling blocks to determine how to provide desired service for reasonable prices. If you drop cable and pick up DSL, that will be a clear market signal
The Kazaa situation is bad - those that are power users of the service are using much more broadband than "surf/email" users. Someone has to pay for the capital investment of thousands of routers and the operating costs of sysadmins, help desk, and install people. Every broadband home takes years to recoup the investment, even for "surf/email" homes.
Moreover, Kazaa primarilly deals in the illegal exchange of copyrighted works, which is a serious liability problem for the provider (in these post Enron days, I supposed we're all concerned with corporate liability).
Until YOU elect a Congress that will make piracy legal, Kazaa is a threat to the provider.
Re:Which utility wins? (Score:2)
Who's your power company? I live in the City of Los Angeles (at least until 11/5/2002) and luckily I have the DWP as my power company. However, if I lived a couple of miles north of here, in the City of San Fernando, I would have Southern California Edison as my power company, and SCE is still recovering from the manufactured power crisis of last year.
I would trust the DWP with Last Mile, but god forbid I'm stuck in SCE or Pacific Gas and Electric territory...I'd be up a creek then, wouldn't I?
However, I agree with the basic premise of the article: the current Internet infrastructure is fuxored and it needs to be taken out of the hands of for-profit entities. I think this should happen sooner than later.
Name? (Score:1)
Institute for the Promotion of the Internet Protocol Utility.
Um, the IPIPU? Personally, I'd never let myself be associated with an organization pronounced "eye pee eye poo."
Re:"an utility network" (Score:2)
"A" preceeds a word beginning with a vowel.
Then how come you said "a word", "a consonant", and "a vowel"? Shouldn't that be "an word", "an consonant", and "an vowel"??
As an AC pointed out elsewhere, "Use of "a utility" is favored over "an utility" because, despite the spelling, the word begins with a consonant sound". To see this in action, compare e.g. "a utility" to "an understanding".
QoS (Score:1)
Some confusion over services (Score:1)
The author makes the assumption that the service that cable modem and other broadband providers should be selling is interconnection of networks. In that case, he is right that "basic" service consists of (a) giving the subscriber a network address and netmask, and (b) routing their packets. Presumably it would be up to the subscriber to handle domain hosting, mail service, etc.
In practice, there are relatively few people out there who want to buy this type of service. As a result, current broadband service focuses on operating a subnet to which subscribers connect individual hosts. Even in those few situations where the cable operator allows multiple devices, the model is almost always one with multiple independent hosts, not a true subnet.
If the market eventually demands (and is willing to pay for) network connectivity instead of simple host support, the broadband operators will undoubtedly offer it as an option.
We need competition (Score:1)