Virtual Reality Getting its Own Network?

loganrapp writes "We've all watched the Matrix, and regardless of how we felt about them, the concept of plugging into a virtual reality appeals greatly to us. It appears that a nonprofit group called the International Association of Virtual Reality Technologies plans to build a network purely for virtual reality. Its name? Neuronet, and the first generation is planned for 2007, with "consumer applications" planned for 2009. There is some fear, however, that the whole thing is a scam."

huh?

[Swimport]

Wouldn't using an existing network like say the internet be much cheaper?

Re:

[garcia]

Well, that's why we have Internet2... It's so that the data transmission speeds are faster and better research is possible. So while it seems pointless overall, they have a decent idea.

Why do *they* need to do this? I haven't a clue. I just don't see how VR applications are important enough to the rest of the world to have a separate network just for it.

Sounds bogus to me.

[sowth]

The statement "...new standards must be created, and network hardware must support those standards..." sounds like they either don't know what they are talking about or it is compeletely bogus. Exactly what modifications to ethernet hardware would be needed for this service? In fact, what changes to IP would be needed? There are already realtime flags/protocols and multicasting built into it. They would just need to buy the right routing equipment, wouldn't they?

I can see why they can't use normal ISPs, since most of them don't support multicast at the home user's end and their latency and bandwidth allowed is usually bad for games/vr. But why would they need to engineer whole new protocols? I just don't see it.

Re:Sounds bogus to me.

[PhiRatE]

Actually, they're right. Not saying they're any more realistic about what they're doing, but the Internet2 research that has been going on has run into plenty of problems with the basic IP protocols, from packet overhead to issues with TCP backoff algorithms. These things just weren't designed with gigabit+ speeds in mind, even though they work there, their efficiency is poor and in some cases really quite pathological.

Of particular concern above and beyond the basic failings when confronted with very high speeds, the balance between bandwidth ad latency starts to warp significantly as the bandwidth of the link increases. The speed of light, and thus the latency from A to B will never change, although response times improve due to better switching, but the amount that can be sent in any given moment constantly increases as we improve our ability to transmit more data per second. As a consequence, the idea of, for example, the standard TCP handshake SYN *wait* ACK will never improve despite greater speeds, it'll always be constrained by the need to wait for light to get to the end and back. At the moment, this is efficient, in the near future you would be better off sending a whole HTTP request off in a single packet and if the other end doesn't want to talk it'll send back a RST instead, reducing the connection times significantly.

There's a considerable number of related issues to do with high bandwidth that need serious investigation, from security implications (brute force of TCP session numbers occasionally rears its ugly head again until someone manages to squeeze better security into the protocol) to better protocols to support routing mechanisms (ipv6 being a good case in point, ipv4 is computationally expensive to route in comparison, causing your megafast link to choke because the hardware can't handle it).

Re:

[networkBoy]

Re: TCI/IP over gigabit: "their efficiency is poor and in some cases really quite pathological."

Damn I've never hear a better discription of those issues and I was at the commercial R&D birth of 802.11 gig PHYs (7 years ago...)
We knew that there were packet issues and such, but just getting decent protocol analyzers and such (SMB2000 IX1600) was a PITA for about the first year and a half. Broadcom's fuck-up of the spec on their first gen parts was no help either (but hey, anything to get to market firs

So don't use TCP

[Mateorabi]

TCP only matters at the endpoints, which presumably they control. If TCP is causing problems, use something else or invent something better to shove in the IP payload. Now if there is a problem at the IP layer of the protocol stack then, yes, your're SOL if you want to use the internets.

Re:

[joto]

Exactly what modifications to ethernet hardware would be needed for this service?

Increase packet size. Sure, some cards and routers have it configurable, but probably not in an appropriate range. A megabyte or ten would make more sense than a kilobyte or two. Also, gigabit ethernet is not really fast.

In fact, what changes to IP would be needed?

See above. Internet protocols require a MTU of 1500 (because of ethernet and hysterical raisins).

But why would they need to engineer whole new protocols? I

Re:

[Jon Kay]

I used to look at this 3d environment space, and I instantly agreed with the assessment of fraud.

There's no need for a separate network. Do notice that plenty of MMOs operate at decent quality today, over today's network. They mostly use bittorrent-like stuff to deal with performance issues.

Re:

[mjmalone]

Well, that's why we have Internet2... It's so that the data transmission speeds are faster and better research is possible.

Internet2 is a non-profit consortium that develops new network technology, it's not a network. It's true that the Abilene network is run by Internet2 (with help from others), but the difference between Abilene and "the Internet" is subtle. Abilene doesn't peer with the public Internet, but most of what's available on Abilene is also available on "the Internet." Further, Abilene doesn't stray very far from the standard protocols used on the Internet, although the adoption rate of new technologies is typic

That would ruin their business model.

[wasted]

Wouldn't using an existing network like say the internet be much cheaper?

If they did that, they couldn't sell the .vr and .cin domain names for their proposed network, and it looks like the sale of these names is to be their near-term source of revenue.

Re:

[desenz]

Well, we wouldn't want the tubes to get clogged with VR.

Re:

[FoXDie]

Mod parent up! It is so true! It's not like it's a big truck, sheesh.

Re:

[sgt_doom]

Wow! What a great scam! It's about time someone used this one (at least in this century, would have been used already back in the old 20th). Gee, of course we're all familiar with that great tech genius, Christopher Scully (guess things were getting kind of slow at the old Wentworth Institute of Technology, huh?).

Can't wait to jack into this 'net...or, maybe not....

Looks completely bogus

[Animats]

Yes, looks bogus.

Domain names are so Web 1.0, anyway. In virtual reality, you have virtual real estate, like the "islands" of Second Life.

Re:

[nbannerman]

Yep, agreed on that one.

If I'm going to be using a VR-type setup, nothing is going to ruin the experience as well as typing a domain name!

Give a give a give a give a Garmin

[tepples]

If I'm going to be using a VR-type setup, nothing is going to ruin the experience as well as typing a domain name!

Think of it as like typing a street name into your GPS receiver.

Re:

[Yetihehe]

I would love to simply type address in some realworld(TM) console and be automagically transported to destination in less than second.

Re:

[H3g3m0n]

With Second Life becoming Opensource soon it would seem a much better solution would be to work out a system of servers based on that.

There would need to be client side storing of items, player avatar, scripts and textures which would either need to be uploaded to the server on client entry or transfered player to player to save server bandwidth.

I don't think the distributed system that second life uses where servers run a fix block of land and you can walk off the edge of one and onto an other server would

Re:

[StuffedFrogYK]

Well, you can't count on them to have enough imagination to get out of a "domain" mindset. Anyway, maybe having domain names for sale instead of virtual real estate is simpler for the average domain name buyer to deal with. I don't know, maybe "virtual real-estate" is too esoteric-soundng.

Not a chance

[slimjim8094]

Bah, I say. This is about as likely to happen in 2007 as DNF being released.

They talk about "not enough bandwidth" to transmit the necessary information.

But wait, there's more! There are so many unanswered questions. How do you connect (not by DSL or cable!)? What's the interface? Does it run on a computer, or a separate appliance?

On another note, this should not make references to the Matrix. It's nothing more than a Second Life, with lower entry requirements (for the providers, of course)

"Not enough bandwidth"

[jd]

This is a common complaint amongst Internet 2 users and is even to be heard from facilities on the European-wide gigabit network. And these are far from being as network-intensive as MMORGs would be if you had comparable screen resolution and frame rates.

So we're looking at 10 gigabits/second minimum, for the kind of really heavy-duty traffic we're talking about, for any reasonable number of servers on that network. There's plenty of dark fiber around and I believe that the record for 200+ mile distances over fiber is in the order of around 4 or 5 petabytes/second. The backbone isn't going to be a technological problem, then. It would be damn expensive to light up enough to cover even as small a region as the United States, but it isn't impossible.

But that's the backbone. How do you get that traffic into people's homes? We're barely at the point of getting people to pay for single gigabit connections, never mind ten gigabit ethernet drops. The NICs are not exactly cheap either. And it's not just any old PC that can sustain a data stream through the PCI bus at those kinds of rates. You're looking at a fairly expensive piece of machinery, and one that is to be used not just solely for games (gaming machines are always expensive) but solely for games on that network. The more you use it for anything else, the less return you get for your investment.

Do I think this is a hoax? Yes. Because it's impossible? No, it could be done. But either it won't be done well enough to be worth having a new network for it, OR it will be too expensive for gamers.

On the other hand, a high-performance VR network for the scientific community, an order or two in magnitude faster than anything currently out there, could be done tomorrow and you're damn right that DARPA, CERN and the other Really Big League users could afford to pay the connection charges. Compared to the cost of the LHA in Switzerland, a ten gig drop per office in these labs that went to a secure petabyte trans-atlantic backbone would look like chump change.

Re:

[MikeFM]

I'd be more interested if someone was actually going to try rolling out a nation-wide fiber to the home initiative that could get us all gigabit or better speeds at a reasonable price ( $100/mo). I don't give a damn about some stupid VR project but I would like to have decent Internet access.

The biggest shortcomings in VR are in the user-interface. If you can't provide a decent experience with all the information sitting on the local machine then you can't expect to do it over the network.

Re:

[jd]

Well, yeah. It would be relatively simple to provide a fiber-to-the-home network that was based on a hierarchical arrangement of meshes. We do this with roads, we do this with electricity, we mostly do this with water (although some areas use wells). There are cable-laying robots that run lines through pre-existing pipes (of which there are plenty) and if you're really cheap, you can run waterproofed cables alongside the railway lines.

I've posted before that we need fairly uniform coverage, not just metro

Re:

[MikeFM]

I totally agree with your statement on cost versus value. Sure it'll cost a lot to build fiber and gigabit+ speed Internet to every home and business in the US but the economic benefits would be huge and the cost of not doing this will hurt badly as we'll slip further behind in education and our companies will not be equipped to keep up with companies in countries such as Japan and Korea. Do we want to lose our technological lead? Having universal gigabit Internet access is required if we want to stay a wor

not even neuro

[yosofun]

when i first read the blurb, i'd assumed that they would have neuro-interfaces, i.e., direct link from brain to the VR net. however, it seems like the only new innovation is the idea of a faster network, designed for such content.

given the fact that technology already exists for this, this might not be a scam. however, making such high-speed networks both affordable and ubiquitously available to the general public might not happen for a while...

Re:

[ortholattice]

But wait, there's more! There are so many unanswered questions. How do you connect (not by DSL or cable!)? What's the interface? Does it run on a computer, or a separate appliance?

Apparently, it will conveniently interface directly to your brain, so there may be no need for a separate appliance. They issued a press release that unfortunately seems to be down on their site right now, but the following is from a Google cache [64.233.161.104] of it:

Vancouver, Canada - The International Association of Brain Interf

Re:

[nuzak]

> Bah, I say. This is about as likely to happen in 2007 as DNF being released.

But of course -- haven't you heard, this is DNF's new game engine!

Cyberpunk much?

[deadhammer]

Neuronet eh? I think a group of researchers got a little too much into the Neal Stephenson and William Gibson novels. Wonder if it's going to be open to the public, I've got some Shadowrun hacking fantasies I've been dying to act out.

Re:

[Anonymous Coward]

If they want to appeal to the Matrix masses they could just get it over with and call it "Neo-net."

Re:

[dangitman]

Dude!

They could call this Vapornet also

[viking80]

From the article:
quoting IAVRT co-founder Chistopher Scully: ...IAVRT is overseeing the registration of Neuronet domain names, the group said. Trademark holders can get an early start from February 5 to June 1; the general public is set to get access after June 4...

and ...Funds raised from the sale of network domain names will offset the considerable costs associated with the creation of the network...

If this is not the definition of vapornet, I do not know what is.
I wonder who are that easy to fool and will pay the registration fee.

Some paid SCO a license fee for Linux, so they might have a customer base here.

Re:

[nr]

Yes, in order to create a separate network they would have to either purchase black fiber and buy and install and operate their own equipment or lease dedicated MPLS/Lambada channels from existing global network operators (Sprint, etc). All financed by sales of domain names? smells very fishy to me *laugh*.

Well, I wish them good luck with their task anyhow. ;)

Re:

[nr]

Correction, it should be "GMPLS/Lambda [techtarget.com]".

I hate how big business is given more rights...

[sowth]

I hate how big business is given more rights over everyone else. If you own a trademark, you get first crack? What if you want to register your name? Say your name is Bill Crack, and some company decides on a whim they want billcrack.vr--why should they have first chance at registering your name?

I could see if they had a terms of service saying no abusing other's tradmarks or acquiring a trademarked name and auctioning it off to the highest bidder, but giving companies first crack just because they happen

Hold on...

[slimjim8094]

They're trying to create an entire new network? Not even the telco's did that (used existing telephone infrastructure). And, they plan to finance it by selling domain names in this network. But, no company will buy one (they must be expensive...) until they see a need. There's no need until the network exists...

Anybody with any idea how they could possibly create an entire new network spanning much of the US (forget the world), with essentially no prospect of money until it's finished?

Re:

[cnettel]

.mil

Re:

[NitsujTPU]

Billions of dollars and no need to turn a profit later?

Re:

[sowth]

Well, I think they will probably start small. Say in high density cities and move out from there. I think that is how cable broadband started...

I'm not sure if this is real or will work, but I do think someone trying this would probably need to create their own networks. Current ISPs just don't have that kind of latency and bandwidth. Nor could they supply the constant stream of packets needed to run a fully realtime high bandwith VR simulation. Just think of all the traffic shaping going on for small tim

Re:

[fyoder]

But, no company will buy one (they must be expensive...) until they see a need.

Not necessarily. You'll want to register your company's domain before your competition or some evil squatter gets it. And look at how all the good .com domain names are taken, whereas here you can get in on the ground floor and get whatever domain you like while the getting is good.

I'm skeptical. If I was to do this as a con this is exactly how I'd play it. Appeal to people's fear and desire to get in on the next big thi

Re:

[meliux]

sell domain names ... no company .. until they see a need ... no need until network exists This is exactly the catch-22 surrounding the IPv6-rollout to end-users. A company/ISP generally won't invest unless there's customer demand. Customer demand doesn't exist until a company publicises a service generally for the masses. The product usually has to result in immediate or direct profits for a company to invest without a prospective customer-base already demanding it. IPv6 in itself does not generate profit

Visit my Neuroblog!

[Quaoar]

I'd tell you how to get there, but your primitive HTML tags are incapable of displaying the 3rd dimension required to visualize the hyperlink.

Re:

[muridae]

Primitive HTML? Look, I stopped using VRML back in 99, now you want me to use it again just to see a bloody hyperlink?

I have one of these

[ipooptoomuch]

(in a deep Austrian accent) My CPU is a neural net processor, a learning computer.

Infinium Phantom 2.0

[XHIIHIIHX]

Deja Vu of when I wasted good money parking domain names for the infinium phantom. [phantom.net]

Re:

[Finn61]

Geez, those Phantom geezers seem to be slightly behind the eightball. They say:
"The Phantom Game Service is anticipated to be the first end-to-end, on-demand game service..."

Yet this has already been around for some time:
http://www.gamenow.com.au/ [gamenow.com.au]

Speaking of VR

[localman]

Is it possible anywhere to get a full field of vision helmet or goggles? From what I've read on VR, the effect of immersion doesn't work until your field of vision is very near completely covered. All the alleged VR gear I see these days just make it look like you're watching a little TV screen. The term VR has been watered down to the point where it covers Wolfenstein 3D if you sit close enough to the screen. I'm surprised there's no truly immersive gaming googles for WoW or something yet...

Cheers.

Re:

[MrSteveSD]

Is it possible anywhere to get a full field of vision helmet or goggles?

Not as far as I am aware, but this company (http://www.leepvr.com/index.php) realises the importance of full FOV. You can't buy one yet. They'll need to get some big player involved like Sony.

From what I've read on VR, the effect of immersion doesn't work until your field of vision is very near completely covered.

I think that's largely correct. It's probably even more important than stereoscopic vision, which breaks down quite quickly with distance. I've been waiting for home VR since I played Dactyl Nightmare in the early nineties. I really think it's long overdue. Providing a VR headset and VR-Suitable games would

Re:

[copec]

Since the demand for HMD's for consumers has been dead for like 10 years there isn't really anything affordable. There are however a lot top end HMD's that are way cool. This company makes a high resolution HMD (4200x2400) with a 180 degree field of view.

http://www.sensics.com/ [sensics.com]

Is it too late...

[testednegative]

... to submit an entry for the Vaporware awards ? I think we have a sure shot here.

Even better, virtual domain names!

[noidentity]

Hurry up people, my virtual domain names* are going fast! Buy them before it's too late.

* Domain names only valid on my local network. Reliability not guaranteed.

Everybody wants to rule the (virtual) world

[Wesley Felter]

I know, I'll set up a new Internet (with blackjack and hookers, natch) where I'm a monopoly so everyone has to pay me. Why should ICANN and IANA and telcos have all the fun?

At least it will serve as a lesson to anyone clueless enough to get scammed by these people.

Re:

[crowbarsarefornerdyg]

On second thought, forget the hookers and the blackjack. Awww, forget the whole thing!

Go ahead, laugh

[PopeRatzo]

I'm sending in my application right away. You guys are gonna be sorry when I'm in my virtual network and you're all still on that old-fashioned, flat-world, so-called "Internet". I'm getting in early, as soon as my broker sells off my shares in Phantom Entertainment. I'm starting to think those Phantom game consoles aren't going to be coming out in time for Christmas, anyway.

Meh...

[Chabil Ha']

I'm preemptively adding it to my list of Vaporware 2007-2009

Suspicous

[edwardpickman]

The fact it's being financed by selling internet based swamp land is a concern.

Not in the wildest dot-com days

[istartedi]

This would have been hard enough to pitch in '99. They're going to build a whole network for a niche application that isn't even consuming a single-digit percentage of the existing internet? That's nuts.

OK, maybe VR is consuming a significant percentage of the net if you define it as "network gaming" or something. If you do that though, you immediately provide an argument against the need for another network, since these applications are successful with the current net. You might be able to argue that you could provide more bandwidth-intensive applications with the dedicated network, but a logical first step is to write the software and run it over the existing network first, and then run demos on a LAN showing how a dedicated network would help. If your LAN demos blow people away, then maybe you have something... but if that were possible, you'd already be hearing network gamers say things like "this rocks on the corporate LAN, but is worthless on my cable modem". I haven't heard anything like that.

Then, as that linked blog pointed out, you'd want to be able to communicate with the Internet at large. So. Then you'd need a Neuronet to Internet gateway of some kind. Even if this conveyed an advantage, just think of the cost--bringing in another ISP just for one app that most people don't even care about???

This just makes no sense to anybody who knows anything. Maybe they'll fleece some really stupid VCs though.

Internet-like growth needs decentralization

[Morgaine]

Even if we assume for a minute that this isn't a scam ...

They've got their underlying model entirely wrong if they're expecting massive growth and success of their VR network by analogy with the Internet.

The Internet bloomed in popularity because it was decentralized and uncontrolled, growing branches at all points and sprouting leaf nodes everywhere. The explosive growth of content "at the edges" happened because of a total lack of coordination and restriction, ie. because people could do their own thing without asking, and almost without cost. And its millions of contributors were driven by fun and interest, not by earning money from their sites.

In contrast, these Neuronet folks seem to be starting with a centralised and tightly restricted registration scheme, plus costly membership that is clearly creating an elite and a money-driven pyramid right from the start.

Well that won't work, if they expect growth modelled on the growth of the Internet.

And it also won't work because of the lack of community-based VR systems to run on such a VR network. The few existing ones that could qualify (Second Life, all online MMOGs and game worlds, clan-based FPSs, etc etc) are almost all proprietary or centralized or both, and hence don't meet the two key requirements for explosive growth.

Frameworks for making non-proprietary and decentralized VR systems do exist, in fact there are many of them (in the guise of open-source 3D game engines), but that's merely a potential rather than a reality for today.

Re:

[shomon2]

Yes, and and all that is needed for a decentralised network that could support VR type stuff is pretty much what Second life or any MMORPG have got, but split up so that each computer attached to it uses their own hardware and storage space to serve their bit of the virtual reality that we create together.

It's so easy to start an open source project to do something like this (I guess requires time, love and hacking skill). Could be as easy as hacking some code from existing things (such as libsecondlife eve

Re:

[cruachan]

Actually my understanding is that Second Life is to some extent decenteralized as the load is spread between areas (sims) on different machines. The part of the system that isn't decentralized, and so has been causing them major headaches of late, is their asset server which is a single cluster.

So Asset Server aside (a pretty big aside) it would in theory be possible for Second Life to grow beyond Linden Labs by additional sim servers being attached to the network. In a sense they partially already do thi

I'm not sure I understand

[thePowerOfGrayskull]

So this is a separate network, independent of the Internet -- indeed, not even connected. Kind of like a LAN. Perhaps they're planning to run fiber to the door for anyone who wants to use their VR network?

Matrix incoming...

[xufos]

There is some fear, however, that the whole thing is a scam.

Be fear if the whole thing is coming true!

Neuronet

[DaAlien]

If you die in the Neuronet do you die out here. The body cannot live with out the mind

Re:

[Bieeanda]

I wouldn't say that; between zombie computers and reams of first posts, there's plenty of evidence for mindlessness on the net.

Wrong end of the argument

[dissy]

Why don't they start with, say, the VR interfacing hardware, and then, I don't know, maybe the computers and software needed to generate the environment.

Get two or more of these nonexistant devices in the same room talking to eachother over nice fast cheap ethernet first. THEN worry about the external network to hook said nonexistant devices together over a distance.

Oh, wait, my bad. Not as easy to get peoples money that way. Nevermind.

Serial Experiments Lain

[darkain]

I would have called it "Wired" or "Seventh Gen Protocol" if it where me...

Alas, the woes of the bleeding edge

[trancertong]

It's almost unfortunate that this would never work out, be it a hoax or not. As someone mentioned earlier, until there are users on the network, no company will buy domains for it, but until people buy domains for it, there will be no network. The exaggerated scope of it seems to make me think it's a hoax though. If they would just make small VR arcades, or VR cafes or what have you, and then network those in a small area, and then gradually grow larger and larger, then maybe that would be viable. But to p

A seperate network?!

[Quai]

A single telephone line was enough for VR.5 [imdb.com]...

worst scam EeveRr

[hazygin]

soo basicly they are trying to make a MMORPG with no cash. great...

there is realy no need for linked 3d maps, since the it would be much less efficent then the current system in tranfering information. if it is to be implimented the map coding/graphics must be procedurel form to achive the enough data compression for the transfer to be more efficent then the current system. allowing the same graphics to be adapted to any system with extremely high compression. but no real advantg here other then it looks pr

web3d.net

[mdickey]

I certainly agree with others that this "smells" like a scam. But if this is a scam, then what are other projects aiming to unify the "3D Web" with a proprietary platform centrally-owned by a single entity? Since Second Life is the subject of so much negative attention lately, let's use them as an example. Are they really so different than IAVRT?

Second Life claims to provide a platform that can grow to handle the demands of the metaverse, yet their servers can handle only a few active users at a time [com.com].

It could be worse...

[Seahawker101]

they could have called it Skynet.