Become a fan of Slashdot on Facebook

 



Forgot your password?
typodupeerror
×
The Internet IT

Japanese Researchers Aim to Replace the Internet 214

Gary writes "Japanese communications minister Yoshihide Suga said Friday that Japan will start research and development on technology for a new generation of network that would replace the Internet, eyeing bringing the technology into commercial use in 2020. The envisaged network is expected to ensure faster and more reliable data transmission, and have more resilience against computer virus attacks and breakdowns."
This discussion has been archived. No new comments can be posted.

Japanese Researchers Aim to Replace the Internet

Comments Filter:
  • by morgan_greywolf ( 835522 ) on Monday August 20, 2007 @08:31AM (#20291851) Homepage Journal
    Doesn't this already exist [internet2.edu]? I mean, seriously, how many parallel projects do we need to do the same thing?
    • by Anonymous Coward on Monday August 20, 2007 @08:38AM (#20291889)
      Agreed. This is totally redundant, and there's no reason to do it - just like other companies writing operating systems when Windows Vista was being developed.

      Wait, bad example...
      • by Oswald ( 235719 ) on Monday August 20, 2007 @09:03AM (#20292033)
        My fear is that it's a perfect example. By 2020, the current internet will have a level of lock-in that makes Windows look disposable. "Faster" and "safer" will have a tough time overcoming "empty".
        • by NickFortune ( 613926 ) on Monday August 20, 2007 @10:00AM (#20292445) Homepage Journal

          By 2020, the current internet will have a level of lock-in that makes Windows look disposable.

          You're going to have to explain that one a little, I'm afraid. "Lock in" doesn't just mean "used by a lot of people". The term, "vendor lock in" to use the full term, is where a single company controls a protocol and abuses that control to force price hikes, unnecessary upgrades and arbitrary restrictions upon its customers.

          But I don't think TCP/IP (the protocol that underlies the Internet) is owned by anyone as such, so it's not like we're going to get forced to pay more for a protocol "upgrade". Nor could some hypothetical owner force us all to use any such upgrade - so there's no fear there.

          As for arbitrary restrictions, the Internet already lets you run any protocol you can devise over TCP/IP without the need for permission or approval. That may change if the anti-net-neutrality crowd start a program of aggressive traffic shaping, but that's hardly likely to be improved by a new proprietary Internet; more likely we'll see DRM on every hop, and no new usages permitted without a five year committee process.

          So, to summarise: please explain how can we have any meaningful lock in on the internet, and (assuming this to be possible), please also explain how this would be bad.

          • by teh_chrizzle ( 963897 ) <kill-9@@@hobbiton...org> on Monday August 20, 2007 @10:47AM (#20292855) Homepage

            please explain how can we have any meaningful lock in on the internet, and (assuming this to be possible), please also explain how this would be bad.

            i think the parent post is referring vendor lock-in, specifically provider lock in.

            if you have no real choice in who provides your internet access you have take what they give you or choose to live without internet access. with all of the shenanigans (filtering, capping, throttling, etc.) that american telcos and cablecos have threatened to pull (or are actively pulling) thanks to the lack of competition in the residential broadband market, perhaps a non-american competitor to the internet as most americans know it is just what the doctor ordered.

            with that said, if they really wanted to impress me they would make such a network accessible from the US.

            • Re: (Score:3, Insightful)

              by NickFortune ( 613926 )

              if you have no real choice in who provides your internet access you have take what they give you or choose to live without internet access.

              Point taken. However, that's not a shortcoming of the way the internet is currently designed. If I wanted to get a second phone line, I could open an account with a second ISP and have two gateways into my home LAN. That would take a little more vendor support for the average user, but there's nothing in the current implementation preventing such a usage.

              I do appre

              • Re: (Score:3, Interesting)

                Reinventing the infrastructure is not going to solve the anti competitive nature plans of some large carriers, and at best it will only provide a feature that we already have.

                i'm certainly quite skeptical of any research project with such a lofty goal, and the point of my post was to clarify what i took to be the parent post's idea of vendor lock in, which you identified (correctly so, in my opinion) as a business/implementation problem rather than a technical one. i am certainly not a nascent-japanese-r

          • Re: (Score:3, Interesting)

            by FireFury03 ( 653718 )
            The term, "vendor lock in" to use the full term, is where a single company controls a protocol and abuses that control to force price hikes, unnecessary upgrades and arbitrary restrictions upon its customers.

            I think you've gone a bit too far with that definition. Vendor lock-in is just where a single company controls a protocol and no third parties can use it in an unrestricted way.

            The company doesn't have to abuse this position - the mere fact that you _have_ to use that company's services constitutes ven
            • I think you've gone a bit too far with that definition. Vendor lock-in is just where a single company controls a protocol and no third parties can use it in an unrestricted way.

              I think I should have had a "for example" between "and" and "abuses". No argument from me :)

              Well, you might be inherently forced to upgrade. When content moves to IPv6, you will need to upgrade to IPv6.

              Well, you will if you want to access that content. But IPv4 addresses are a subset of IPv6, and it's likely that you'll be

              • But IPv4 addresses are a subset of IPv6, and it's likely that you'll be able to access most if not all of the current web for a long time to come.

                Not entirely accurate - IPv4 addresses can be represented as IPv6 addresses, but the actual protocols are still different (even if that's hidden from the application).

                it's likely that you'll be able to access most if not all of the current web for a long time to come.

                When we run out of IPv4 addresses you're going to struggle to make services available through IPv4
                • There's a lot of points I could take issue with there, but I'm going to stick to the main one:

                  IPv4 has lots of problems, not least the lack of IP addresses, which forces you to use nasty hacks like NAT. However, you can't ditch IPv4 in favor of a protocol that solves these problems because you are locked into using services which are only provided over IPv4. This is why lock-in is bad - it prevents you from switching to a better technology (or at least - a technology that's better for _you_).

                  The prob

      • Re: (Score:3, Funny)

        by marcello_dl ( 667940 )
        This is totally redundant, and there's no reason to do it - just like other companies writing operating systems when Windows Vista was being developed...

        Wait, bad example...


        Of course it is. You called Vista an operating system.

        Operating systems make computers work, vista makes Gates rich.
    • Re: (Score:3, Insightful)

      by MightyYar ( 622222 )
      I don't see anything wrong with trying different approaches to the same end. Perhaps they disagree with compromises or design decisions that were made with Internet 2.0?
      • by mikael ( 484 ) on Monday August 20, 2007 @09:03AM (#20292035)
        The internet is built over a series of seven layers - the . [wikipedia.org]

        The idea of splitting everything into layers is so that any one system could be changed without having to totally rewrite everything else - if you want to replace your dial-up modem with a wi-fi card, all you have to do is replace the drivers. If your ISP wants to replace their router network with an ATM network that's easily done without affecting you. If someone came along with a better router management protocol, that's easily done.

        The original Internet did have redundancy and resistance against breakdown built in. Unfortunately, many network companies found it cheaper simply to route separate logical networks along one connection, rather than have multiple and completely separate connections. Thus, we end up with a hard-wired minimum spanning tree network, that fails as soon as one link goes down.

        Let them go ahead with this idea, but by the time they complete their literature survey, they will probably find out there is very little that they need to change.
        • by Retric ( 704075 ) on Monday August 20, 2007 @09:42AM (#20292317)
          I know this is /. but did you read the article you linked?

          Noting actually uses the OSI model it's just an abstraction to help people understand how networking works. The Internet uses the TCP/IP http://en.wikipedia.org/wiki/TCP/IP_model [wikipedia.org] model of Application, Transport, Network, and Data link layers.

          PS: The internet has redundancy as a mesh of networks even if many of those networks have single point's of failure. On second as you speak with such conviction on subjects you know little about you might belong on /. Welcome to the party.
          • by larkost ( 79011 )
            It may seem like a minor difference, but the Internet at large is not "TCP/IP" only but rather IP only. You can layer TCP on top of it quite easily, and most of the routers out there have special optimization for TCP, but it is not required by any stretch of the imagination. TCP is a very common layer, but many connections are UDP, and you can technically have any other packet type you want, as long as it fits on top of IP.

            Now it may be that firewalls will not let packets other than TCP or UDP in, but that
            • There are a couple of protocols out there that beat TCP for almost every use-case, and in 10 years they might be more popularly used. The internet at large does not have to change a bit to handle these protocols.

              A sincere question for you larkhost (how many of those have you seen on /. this year? :) - could you expand on this a little please, which protocols and why? My knowledge of underlying network protocols is not as good as it should be.

              Thanks.
              • udp beats the hell out of tcp for tunnelling tcp.
              • Re: (Score:3, Insightful)

                by Kadin2048 ( 468275 ) *
                Not to put words in the GP's mouth, but there are a few transport-layer protocols that I've come across which go in different directions from TCP and UDP.

                There is a list over at Wikipedia [wikipedia.org], although I don't know if it's really close to exhaustive.

                A lot of them are aiming for some sort of [wikipedia.org] middle ground [wikipedia.org] between TCP and UDP. They want the statelessness of UDP but some of the congestion-control and error correction of TCP, but without having to reinvent the wheel by building their own error-correction on top of
            • by Retric ( 704075 )
              I was talking about the TCP/IP model which oddly enough works just as well with IRC/UDP/IP/Ethernet/Ethernet physical layer as HTTP/TCP/IP/Frame Relay/Optical Fiber.

              "The TCP/IP model or Internet reference model, sometimes called the DoD model (DoD, Department of Defense), ARPANET reference model, is a layered abstract description for communications and computer network protocol design."(http://en.wikipedia.org/wiki/OSI_model)

              Anyway, I was trying to walk the fine line between roasting someone with a reall
              • You are correct about the Internet being based on the TCP/IP model. This makes sense due to that fact that the OSI model was not created until 1977.

                The OSI Model added the presentation and session layers, and renamed the "Internet Layer" to the "Network" to make the model more generic.

                The TCP/IP Model can be mapped to the OSI Model, since TCP/IP Model defines all the layers except for the presentation and session layers.

                The two are not mutually exclusive. The wikipedia article that you linked for the

          • Re: (Score:3, Interesting)

            by mikael ( 484 )
            Yes, I did - I've also worked on networking products (network probes and LAN analyzers). Even had one of those protocol charts above my desk, so I've got a good idea of how "the tubes work".

            The Japanese have always had these grand computer initiatives (the last couple were "The TRON project" [super-nova.co.jp], and Fifth Generation computing (AI, Expert Systems, Automated Learning).

            The TRON project was an attempt to have computers be able to have a standard communication protocol:


            First, there is the problem of reliability. Ha
      • by bockelboy ( 824282 ) on Monday August 20, 2007 @09:20AM (#20292151)
        The Internet 2 is a consortium maintains a high-speed backbone across the US; the costs are subsidized by the government so universities can communicate each other at 10Gbps rates without having to go out to the commercial Internet. A small portion of the funding goes to some middleware projects.

        However, most NSF-funded networking projects use the I2 as their testbed, but they're not necessarily a part of the I2. For example, GENI - the US effort to redesign internet protocols from the ground up - will run in parallel with I2. GENI is the US counterpart to this Japanese effort (although it's hard to tell from the light-on-details article).
    • As many as it takes to get one that's free of private interests.
    • by lheal ( 86013 ) <lheal1999NO@SPAMyahoo.com> on Monday August 20, 2007 @08:57AM (#20292001) Journal
      Unless they figure out how to ensure redundancy, they will have reinvented the wheel. The reason the Internet is unreliable is that the last two nodes on the graph have only one connection. Why do we have only one ISP, and why do ISPs only have one upstream provider? Economics. Let's see them solve that one.

      Furthermore, we've been about to implement IPv6 for years now.

      Even furthermore, their ultra-secure shiny modern internetwork will still have to connect to the kludgy 1980's rustbucket the rest of us use on our Windows-based computers, which means it will be pwned in a few minutes just like the original.

      It's the Silver Bullet Syndrome. They think they'll invent a secure network, when all they'll be doing is achieving a bit of obsecurity.
    • by cp.tar ( 871488 )

      From what I'd understood, Internet2 is a fundamentally different network - first of all, it is not de-centralized like the Internet; add the fact that AFAICT it is not (yet) open to general public, and despite the name, it doesn't seem to be a replacement.

      Granted, I may be missing some important points, as Internet2 doesn't exist in Croatia, but whatever... I don't mind the development of different networks; may the best and fastest one (both in transmission and wide deployment) win.

    • I mean, seriously, how many parallel projects do we need to do the same thing?
      Well. One is American, one is Japanese...

      It's called competition. At some point someone makes a bundle of money that the others don't make.

       
    • Look, the article's content-free hype from a high-level bureaucrat reported by a non-technical newspaper writer, saying that he's obtaining funding for committees to develop future projects that'll be Really Cool. Nothing to see here, move along....

      ...

      Oh, still here? Presumably the research he's talking about isn't just IPv6, because that's starting way too late, and lots of good work was actually done in Japan. Maybe it's something transport-related or router-related or content-related, which could

  • hmm (Score:2, Insightful)

    by BuR4N ( 512430 )
    In normal cases when you see news like this I would be tempted to say that this is something that will never materalize, but Japan have a trackrecord of going their own way with for example mobilephone networks. WIll be interesting to watch if they getting anywhere with this.
    • Re: (Score:3, Informative)

      by dsginter ( 104154 )
      Did anyone else picture an ethernet cable jammed into a can of Folgers [folgers.com] when they read th title?

      We've secretly replaced Yoshi's 100Mbit internet connection with Folgers Crystals. Let's see if he notices!
    • First - wire their own country for it.

      Second - provide gateways and translations from the old Internet to their new version.

      Third - provide the specs in an Open standard so anyone else can also implement it.

      Fourth - provide the specs for tunneling their new Internet through the old one until the new Internets are connected to each other.

      At the very worst they end up with their improved version for their own people. (If it really is improved.)
      • Fifth - Profit!

      • First - wire their own country for it.

        Not if they stick with the Internet Protocol (IP). The data will move through the existing network. But it might still not be compatible with any external network. How about if they implement the 'Son of SMTP' to make it automatically secure and SPAM resistant? That would be a worthwhile improvement but it would need some form of gateway/interface to transfer email between Japan and a network that didn't implement it.

    • However, with cellphone networks, it's easy to have your own little internal network for your country, and then switch to something else to communicate with the rest of the world. However, the Internet is more global, and if everybody isn't using the same technologies (http, tcp/ip, smtp, etc) to communicate, then things get hard to manage. You could have some sort of translation utility to communicate between Japan and the outside world, but I don't think it could work very well for things that require en
  • I guess the next wave of the internet will be based around much freakier porn than today's internet.
  • by BibelBiber ( 557179 ) on Monday August 20, 2007 @08:39AM (#20291893)
    The thing i snobody wants to pay for it. Compare this to the AOL and CompuServe networks that were available for a long time. Competing with the free internet. They don't exist anymore. Just because anybody who owns it can put restrictions on you. It's not gonna work.
    • by Lumpy ( 12016 ) on Monday August 20, 2007 @08:51AM (#20291955) Homepage
      You are completely ignoring what everyone is thinking.

      how much DRM are they gonna shovel onto this thing? The current Internet setup is near perfect because of it's flaws. It is why it took off like a bat out of hell. "fix it" like these researchers and corperations want it and it will be Cable TV. Bland and icky.

      They want to shove so much DRM into the internet as well as have all your packets signed by your information, etc...

      I have a suggestion for the researchers, give up now, it will be a failure. good god look at how long ipV6 has been around and it is still being ignored. I think I read my 100th article about how we are running out of IP addresses that was worded identically to the one I read in 1999.

      • That at least is compatible with existing stuff. That is to say you can implement IPv6 on parts of the Internet (like your own internal network) and still talk ot parts that don't use it. I can't conceive a new network that is secure against hacks, viruses, and so on but is still compatible with the old Internet, which has lots of those sorts of things. I mean sure, you can design a network where there's all sorts of controls on it and nodes have to authenticate themselves and so on. You can set it up such
      • ignoring? This is what I was saying by writing "restrictions" (the "R" in DRM).
  • I bet this is going to be the same as the old one, except that all the addresses will comply to the following syntax: pika.youraddresshere.chu
  • I remember the last time the Japanese announced that they were going to change the whole face of computing, with this [wikipedia.org] project. After a few years, it was going to be the only hardware/operating system/networking combination that anyone would ever use. I wonder how they're getting on?
    • by Goaway ( 82658 )

      As of 2003, the TRON system (or more specifically the ITRON derivative) is one of the world's most used operating systems, being present in millions of electronic devices.
      • [citation needed]
        • [citation needed]

          Sure thing: http://www.linuxinsider.com/story/31855.html [linuxinsider.com]

          "The ITRON specification is a standard real-time OS kernel that can be tailored to any embedded system. ITRON already has been ported to a wide range of microprocessor architectures and has quickly become Japan's de facto standard for embedded systems. Today, the specification is used in an estimated 3 billion microprocessors."

          So the Wikipedia article is wrong, as of 2003 TRON was used in billions of devices, not millions.

          • Ah, so it is only ubiquitous in Japan. And that is hardly impressive considering that those little handheld poker and blackjack games are considered embedded devices.
            • by Goaway ( 82658 )
              Being used in three billion devices in a single country is less impressive to you than being used in three billion devices world-wide?

              PS: How many of your electronic devices are made in Japan, again?
              • Technically, close to none. ;)

                But, when we consider all of the embedded electronics that get tossed around today (look in a Toys R Us for example), this number quickly becomes irrelevant.

                Now, what percentage of embedded devices use this system.
    • Re: (Score:3, Interesting)

      TRON has been a ridiculous success being one of, if not the most popular embedded operating systems in the world, meaning that it probably has more devices running it than the number of PCs running Windows/MacOS/Linux/etc. combined. Sure, I think it would be difficult to argue that it has changed "the whole face of computing", but really, is that anything to scoff at? I mean, how many technologies are there that have?
    • Wow. I assumed you were going to link to this. [wikipedia.org] TRON is spectacularly bad as an example of something that failed.
  • Once upon a time, France had complete domination of network information communication thingies [wikipedia.org].

    France probably laughed too, a big gutteral Gaulic laugh: "Silly Americains, think you can replace the Minitel? I fart in your general direction!"
    • by igb ( 28052 )
      You're taking the wrong lesson from Minitel. The real lesson is that pan-national will always triumph over local. See how far OSI got, even in the countries that mandated it (such as, of course, Japan).

      ian

    • Minitel was a marvel at that time. Its biggest flauw : a monopoly.

      France Telecom was in charge of it. They kept as much as 70% (or more) of any revenue you could possibly make out of it.
  • by jollyreaper ( 513215 ) on Monday August 20, 2007 @08:47AM (#20291935)
    Whatever they replace it with has got to be a) self-aware b) housed in a really cool-looking robotic body c) flail phallic, cybernetic tentacles on command and d) be preoccupied with conquering neighboring nations and cowering schoolgirls. I predict it will be called EcchiNet. Nuclear war and terminator endoskeletons to come later.
  • by Yvan256 ( 722131 ) on Monday August 20, 2007 @08:57AM (#20291997) Homepage Journal
    The american internet is made of a series of tubes, right? Well, then we can guess that the Japanese version will be a series of tentacles.
    • by AccUser ( 191555 )
      we can guess that the Japanese version will be a series of tentacles

      Don't you mean noodles?
  • Costs (Score:3, Funny)

    by peterpi ( 585134 ) on Monday August 20, 2007 @09:04AM (#20292037)
    Great. We can ask them how much it's going to cost [slashdot.org]
  • by dwater ( 72834 )
    Anyone seen serial experiments lain? I think the Japanese are the last people we want inventing any internet....bazaar, to say the least, but strangely gripping, in a cultish kind of way.

    Good music though.
  • Damn, that response time was faster than a freaked out web browser company resolving a security hole!

    Just days after we heard Internet TV would crash it [slashdot.org] they're working on a fix. And they're working on an Internet, not just a security hole. :-o
  • I don't even trust this old and busted internet. The new internet would be nothing but pandering to government 'security' concerns and big business DRM demands.

    Count me out.
  • 1. Will it be backwards compatible with the existing internet?
    ------------- (If "no" for #1 above, it must be a Microsoft product!)

    2. This kind of claim sounds like a marketing campaign, is this a marketing effort?
    ------------- (If "yes" for #2 above, it must be a Microsoft product!)

  • by Danathar ( 267989 ) on Monday August 20, 2007 @09:16AM (#20292119) Journal
    http://www.nsf.gov/cise/cns/geni/ [nsf.gov]

    "With support from the Directorate for Computer and Information Science and Engineering (CISE), researchers are working together to design a bold new research platform called GENI, the Global Environment for Network Innovations. As envisioned, GENI will allow researchers throughout the country to build and experiment with completely new and different designs and capabilities that will inform the creation of a 21st Century Internet."
  • Not likely to work (Score:5, Insightful)

    by Sycraft-fu ( 314770 ) on Monday August 20, 2007 @09:19AM (#20292149)
    The problem that all these people who want to replace things like e-mail or the Internet run in to is the whole thing that makes these technologies great is interoperability. The great thing about the Internet is that you hook in to it anywhere and barring your ISP or government having blocks up, you can talk to everything. You can switch ISPs, areas of the world, devices, etc it all doesn't matter. It's not like we didn't have networks before the Internet, what we didn't have was a network that everyone and everything could work on.

    So if you are going to replace it, you have to do it with something that works with the Internet. I am not going to sign on to a new network, no matter how good you say your technology is, if I can't access what's already out there. Of course a big part of what people want to do when creating a new standard is to cut off the problems that the old standard had, and thus it becomes incompatible and thus isn't workable.

    I mean the problem with a new e-mail system isn't designing one that's resistant to spam. That's easy. The problem is designing one that is resistant to spam but not incompatible with existing, unsecure, e-mail. You aren't going to get people to switch otherwise. It doesn't do me any good to have a spam proof technology if all the people who need to contact me don't also use that.

    Same deal with the Internet at large. I don't care how cool your new network is, if it doesn't provide me with access to everything on the Internet, and give everyone on the Internet access to servers I run, then it really isn't very useful to me.

    Really, the Internet, for all its flaws, is here to stay for a long time I think. It's not that we couldn't do better, it's that we aren't willing to redo everything from the ground up and switch over. Same shit with plenty of other things. With modern technology, a HVDC power grid might be a better system than what we have. However that's not what we have, and we aren't going to replace what we do have entirely, so we keep adding to the existing system. The Internet is much harder given that you are talking about a network that spans the whole world (and that you actually can convert AC to DC and back).

    It's a nice thought that "Hey, let's just tear down all this crap and rebuild it right, based on the better knowledge we have now," but it usually isn't at all practical in reality.
    • by vidarh ( 309115 )
      I agree with what you say, but for 99% of applications out there, being "compatible with the internet" means being able to address nodes in [insert new system here] that needs to be publicly accessible with ipv4 addresses and talk to them via tcp and udp. Now, the applications running on that new system does not need to know tcp or udp - they just need to know a stream oriented and packet oriented protocol that can be reasonably easily proxied to look like tcp or udp to the world at large. Same goes for a l
      • The problem is that you eliminate any of your nice security bonuses by doing that. You can design a nice, secure network that requires all sorts of authentication and is setup so nodes are uniquely identified and verified and so on. However that isn't going to accomplish anything if you still want to communicate with the Internet since it DOESN'T play by those rules.

        As I said, it is similar to e-mail verification deals. Sure, I can design a system that verifies senders and thus keeps spammers out, however n
  • Do you think IP6 will be implemented by then?
  • Can they make it Godzilla-proof?

  • Good Luck with that.

    -Hackus
  • I hope they would think about the children and peer connections will be not possible. All we need is a solid corporate content provider to user lines.

    Seriously though :( ... why not just switch to IPv6 finally?
  • Originally MicroSoft was going to grow MSN as its "private internet" and conquer the Net with proprietary protocols, much like it did in operating systems, office software, and browsers(afterwards). MicroSoft proposed "extensions" to TC/IP or replacing it altogther and losts of people were upset. The Bill had his famous "Damascus" episode where he turned MicroSoft 180-degrees into embracing the then InterNet.
  • better control of what people sees of course !

    theres always this main motive behind 'new internet' crap.
  • http://www.theonion.com/content/news/earthquake_se ts_japan_back_to_2147 [theonion.com]

    It's just that an earthquake set them back over a few hundred years.

Math is like love -- a simple idea but it can get complicated. -- R. Drabek

Working...