Jan 15, 2011

Requiem For Blue Mars

Written by: Will Burns [Aeonix Aeon]

Twitter: @DarianKnight

::: TLDR Warning :::


Connie Sec Blue Mars Crossing Bridges


Ridiculously Short Summary:


Blue Mars is Broke. The CEO is gone and most of the team. They’re using the money they have left and the board of directors is betting the remainder of the farm on mobile/web technology with OTOY, because they can’t afford to develop a desktop native client any longer. Chances are the cloud rendering option will eventually bleed them dry and what is left of Avatar Reality will no longer be a reality. Read on for full details and synopsis.


Full Article:


In what can be seen as an unexpected move on Friday, Avatar Reality (makers of Blue Mars) announced their intentions to restructure the company and focus mainly on mobile platforms. This will, inevitably, come at the expense of the PC client many currently use, and over time the PC client will be phased out entirely in favor of the iOS (Apple) mobile client application and possible web client.


Of course, this would have come as a surprise had it not been for the simple fact that this was not unexpected.


Rather than speculating on the news of Blue Mars, I did what anyone in my position would do; I simply had a chat with somebody who actually works (worked) at Avatar Reality and asked what happened. Reasonably speaking, I had asked not because I didn’t actually know what had happened, but I wanted to hear it from him directly.


Here is what actually happened:


Avatar Reality has run out of money.


The investment capital has more or less run dry and without further capital, the powers that be (the board of directors) have decided to put what capital is left in the coffers toward Mobile and social media integration development, betting what’s left of the farm on OTOY cloud rendering. In August 2010, their plans for utilizing OTOY cloud rendering for Blue Mars did not come to fruition, and as a result the delay in implementing that solution caused the company to more or less fold. In honesty, when the CEO of a company resigns and takes a large portion of the team in a “restructuring” effort with them while the remaining team is focused on social media and web applications it’s a clear sign that… something… is…


Why does this sound familiar? Oh.. right… Second Life. *sighs*


That’s not to say that OTOY is at fault for this, because if the success of your company relies completely on another company to deliver a solution that your own company cannot provide (when clearly it is your company’s job to provide that solution) then the fault lies with your company for having a broken system to begin with.


While Avatar Reality still exists as a shadow of its former self, at this juncture it is now a matter of when and not if the company will completely deplete their remaining investment capital and run ashore with their ship. However, maybe the plan here is roughly the same as with Second Life?


You know… Facebook and twitter integration, a cloud rendered streaming web version, all in the hopes that some kind corporate giant will buy the company and reward the board of directors and team leaders for making bad decisions and blowing millions of dollars in investment capital on poorly thought out virtual environment architectures that could have been avoided by simply doing basic research. As an aside:


*Waves to Mr. Humble, the new CEO of Linden Lab*


Part of a prior conversation with Avatar Reality and myself roughly six months ago outlined this issue directly and stated that the servers alone will cost more money than can effectively scale at cost, citing bandwidth requirements alone in serving premade worlds in bulk download plus rendering and streaming on adequate PCs capable of at least offloading some of that rendering cost. While it is true that OTOY can effectively render and stream complex scenes over a broadband connection, the factor which was ignored was at what cost for that computation and bandwidth.


As it turns out, and rightly so, with or without OTOY in use, the cost of rendering and streaming Crytek level graphics costs far more than the company could sustain, and by all means more than any content creator was willing to pay to have hosted.


Which brings us back to the Mobile/Web decision.


With a PC client, there was at least the vague option that the client was running on a computer remotely capable of offsetting the rendering costs locally. After all, the Crytek Engine is nothing short of hardware intensive (to say the least). However, choices were made internally to stream the rendering server side at what can be quickly assumed to be more money than they were capable of making to offset the bills.


With a mobile platform in the works, that cost reduction no longer exists. Avatar Reality, in some twisted logic has decided not to choose a solution which would allow them to offset the costs but instead would inherently dictate that the full brunt of the costs be on their own heads. Can we say that the iPad, iPhone and related mobile smartphones are actually capable of rendering Crytek level graphics in real time?


Of course not.


Most PCs short of monster gaming rigs have a hard time using the Crytek Engine, and to assume that a mobile platform would be more capable than a high powered personal computer (whether an Apple or PC) is simply ridiculous.


So the mobile option now involves not less centralized computing to lower costs but more centralized computing, bandwidth and rendering to vastly underpowered devices. It seems plausible that having one foot in the grave at Avatar Reality isn’t enough to invoke common sense within their board of directors.


As noted with SecondLife Skylight, cloud rendering and streaming does in fact work, however the bandwidth used to stream that information (not to mention the costs involved with cloud rendering) are something to be alarmed about. In tests done by various users of Skylight, the standard bandwidth used over the course of a 60 minute session was 1 GB or more.


I’m not aware of a mobile provider that can or will support 1 GB of data transfer per hour on a data plan, and of those who may allow such a thing, the “unlimited” plan would surely be brought to its knees fairly quickly, causing massive overage charges to the user in question. Then there is the question of who exactly is footing the bill for the cloud rendering and streaming of that data to begin with (aside from the massive data charges for the mobile user).


So let us assume that a web based system (or mobile platform) based on cloud architecture is really aimed at WIFI tethered Mobile devices or websites where the average user may as well be using a personal computer (again, Apple or PC). But in this case, the question then becomes not if this is possible, but rather why anyone would not simply want to use a native desktop application for this purpose and gain more benefits from doing so.


In the case of Skylight, I can understand the plan to somehow create the ability to make Second Life into an easy to use application for Facebook, of course stripped down considerably in ability. Regardless, we’re still looking at much more bandwidth in use for a fraction of the functionality and possibly additional costs server side for rendering and streaming that information.


So who will foot this bill?


Surely it won’t be the content creators of Avatar Reality. The cost is simply too high and the returns too small to consider it an adequate ROI.


Maybe Avatar Reality will translate those costs to subscriptions for the end-user? Again, the cost is too high to be worth it to the end-user and the returns are still too low to justify it.


In the end, Avatar Reality will be footing the cost of this platform, with marginal (at best) returns from content developers and users. In no way will those returns remotely close the gap for the costs.


Looking Back at the Future

The really interesting part about all of this is that in order to see the future of Avatar Reality, and subsequently Blue Mars (or any virtual environment today), we need not look into the future but instead look to the past.


When I originally used Blue Mars in the early beta, I had this nagging feeling that it seemed familiar to me… like I had seen it before. After a few days of thinking about it, I finally realized why Blue Mars seemed familiar, and I was correct in thinking I had seen something like it before; as a matter of fact I had seen something exactly like it before.


Worlds Inc + Crytek = Blue Mars

For those of you in the industry seasoned enough to remember, Worlds Inc was one of the grandaddys of Virtual Worlds platforms in the early 1990s. The WorldsPlayer client was downloaded and each individual World had to be downloaded separately and in advance prior to having access to it. Those worlds were also pre-made and “published” as scenes. The client came with a starter world (welcome area) and had a “map” of other worlds available to visit (after varying download sizes and installations).


At this point, there really is no differentiation between whether I’m describing Worlds Inc or Blue Mars anymore. The difference, though, happens to be in whether or not Blue Mars managed to learn absolutely anything from history.


In this case, the answer is no.


Worlds Inc was in a position where they had to think creatively in order to minimize bandwidth because the average user was lucky to have a 33.6 Modem let alone a 56k Modem. Blue Mars, deciding that they wanted to look good, decided to license the Crytek Engine and then have designers create 50 to 500MB scenes (not to mention a 300 + MB client download) that needed to be preloaded in full and installed prior to entry. Instead of trying to find ways to minimize bandwidth usage and graphics rendering requirements to something a “normal” computer could actually handle, Avatar Reality saw fit to create a system in which the only possible way to utilize it is to have a veritable supercomputer cluster pre-render the entire program for each user and stream the results over broadband.


Is it my imagination or has the virtual environment industry become mentally handicapped?


Instead of learning from past mistakes, and doing everything in their power to avoid them, the entire virtual environment industry is stuck in some weird stupidity loop, literally doing the exact opposite of what they should be doing as demonstrated by every possible prior attempt in the industry. Even if we look at web clients like web.alive/Avaya, I can point to it being a rehash of VRML and X3D in the 1990s, and we already know what happened with that.


None of this is new or innovative.


In 1990, which is now 21 years ago, the solution to this madness was given. Every year after that, a new virtual environment springs up, doing the exact opposite of what was said, and somehow is left in surprise when millions of dollars in funding dries up and they are left with a company (and virtual environment) that has failed.


Remember the early 1990s? VRML was all the rage, and there were companies springing up all over to offer some sort of Virtual Reality software or service. Blaxxun Interactive, ActiveWorlds, Worlds Inc, you name it… Virtual Reality was new and exciting! It was the future of communication!


And then VR died.


Of that wave in the 1990s, I’d say that ActiveWorlds is still around but barely alive. Worlds Inc is apparently responsible for Utherverse and Red Light Center which is essentially a XXX pay per play virtual environment (and still terrible quality by today’s standards). I haven’t the slightest clue what became of Blaxxun Interactive, but truth be told, we don’t hear much about that company any more (assuming they even exist).


What did we learn from this wave? Apparently nothing.


In the early 2000s there was yet another wave of virtual environments and services, There.com, SecondLife, Blue Mars, Kaneva, web.alive/Avaya, and at least half a dozen more (Open Croquet, Cobalt, Wonderland, Project Darkstar, Proterra, Proton Media…) and the weird thing about this is: The people using, and those responsible for, these recent systems have amnesia.


I remember reading a statement by Rivers Run Red claiming they were the first business in the Metaverse, and obviously I laughed, but then I got very worried. A company like RRR who is supposedly representing corporate interests in the virtual world space, should be smart enough to never have made that claim, and yet they did. Much like listening to Philip Rosedale claim that Second Life is the Metaverse and that Metaverse really means “Metaphysical Universe”.


In 1990, the solution was given by two people to all of this madness. Chip Morningstar and F. Randall Farmer, authors of Lessons Learned From Lucasfilm’s Habitat. Strangely enough I had asked Mr Farmer about Linden Lab and he informed me that he was actually called in as a consultant in the early days, and not surprisingly, ignored.


The answer they stated: Centralization is the bottleneck and does not scale cost effectively. A decentralized approach is needed in order to effectively build a system capable of scaling to millions or hundreds of millions of simultaneous users.


So what has every system since 1990 done?


They’ve all become increasingly more centralized, and each company has eventually met their demise as a result, or have become footnotes in the history books, forever ignored and forgotten by the next generation who apparently enjoy making the exact same mistakes.


Avatar Reality had ample warning, however, that this was going to happen. I know this because I personally warned them. Just like I warned that Second Life would have similar issues. And, strangely enough, just as I warned Activeworlds that their system would need a complete overhaul to remain both competitive and relevant.


In each case I was told I was wrong, and while I watch ActiveWorlds implement my proposal list four years later, Avatar Reality hits a brick wall and runs out of money, There.com ceases to exist, and Second Life “restructures” and hemorrhages CEOs… well…


They say that hindsight is 20/20, but you’d think that there have been enough mistakes in the past to convince people today not to repeat them.


For somebody that is constantly told he’s wrong, I seem to be getting an awful lot right. The only difference is, it’s not costing me tens or hundreds of millions of dollars in the process.


  1. Anonymous12:05 PM

    Outstanding post. People ignore past lessons at their own peril. I wonder why that's so often the case with businesses designing virtual worlds.

    I remember reading "Lessons from Lucasfilm's Habitat" back in the early 90's. Still a brilliant read today. I didn't see a link posted to it in your blog, so here's one: http://www.fudco.com/chip/lessons.html

    What are your thoughts about Hypergrid connectivity and Opensim? I think it's a compelling way to avoid the bottleneck/centralization issue in the same way the current model of the Web has solved that issue. Constellations of interconnected worlds, and the ability to move freely between then...

  2. @Becunningandfulloftricks

    Thanks for the comment! I didn't post a link to Lessons Learned because I've reposted it in whole and in part on this blog a couple of times. Thank you, though, for offering a link to it for my readers :)

    As for HyperGrid, I think it has a great idea but unfortunately is based on the extensive bottleneck baggage of Second Life architecture, so I feel that while they have mostly the right idea, in the end it seems like they are likely to succumb to the same issues over time.

    If anything, I'm intrigued by fully decentralized approaches such as Solipsis (http://www.solipsis.org/) however, I believe a real solution may be in a median approach between the two extremes. Hybrid Decentralized approach.

  3. A recent comment came to my email, apparently unable to read the captcha to post, so I'll put it here with my reply:

    From William Glascoe:

    I'm a Web3D Consortium member so I take umbrage at remarks suggesting VRML/X3D (H-Anim) is defunct. Anyway, I'd like to read your critique of Octaga and Bitmanagement operations, which are built on VRML and X3D.

    My reply:

    It wasn't a suggestion that VRML or X3D are entirely defunct, it was more an acknowledgment that the initial hype and wave concerning those two technologies have (at best) taken a back burner and lost favor with modern approaches. Clearly these two technologies (and standards) aren't as mainstream as they once were, where I remember the entire world was buzzing about everything being in VRML or X3D as the future, now I rarely (if ever) hear a word about them. So for all intents and purposes, they are defunct in terms of serious consideration for a new platform which may utilize them on a mainstream scale.

    I'm sure there are companies and efforts still in progress which base their efforts on VRML or X3D, but in 2011 they have the same weight in the industry as say, Blaxxun or ActiveWorlds would have to enact progress. Maybe this will change going forward and there will be a breakthrough in those two standards which warrant a second look on a wide scale, but until then, at best these technologies are "latent".

    For my readers:

    If you're interested in seeing where VRML and X3D stand in quality and use today, feel free to check out the links provided by Mr. Glascoe of the Web3D Consortium.


  4. Amazing :)

    You are bang on about the rush to mobile being a waste a time. Data plans are so restrictive in terms of usage and so incredibly expensive... Joe Average is NOT going to be accessing his VW on his iPad or iPhone.

    Not to mention culturally... as it exists right now, people are private with their interactions in virtual worlds. Who the hell wants to be on a subway or in a line up at the Deli (wink) and be trying to navigate their avatar around and catch up with friends?

    It's not how people USE it. And they ought to know that by now.

    Those that know that VW's are making huge mistakes that will lead to their demise... have the painful obligation of trying to shout it out... having absolutely no one listen (or be accused as a nay sayer) ... and the sad sad role of watching all predictions come to fruition.

    I liked Blue Mars. Found it too hard on all my computers graphics wise (and I have some high end stuff)... but it lacked so much. I always thought it would scale back a bit and become more mainstream accessible... I had hoped. Because it was a very visually superior experience to be in there.

    Very sad to watch it tank...

    PS: You were right. But I know it sucks to be right in this case :(

    PPS: I feel like I should send a card to Hamlet Au :( Poor guy.


  5. "As for HyperGrid, I think it has a great idea but unfortunately is based on the extensive bottleneck baggage of Second Life architecture,so I feel that while they have mostly the right idea, in the end it seems like they are likely to succumb to the same issues over time."
    So is Second Life doomed?

  6. I read the habitat paper (via @pathfinder) earlier, and while I think their understanding of "cyberspace" was a lot more game-focussed (more like a MMORPG than what we now think of when we hear "virtual world"), their ideas and experiences are still valid - twenty years later!
    Just as they predicted, neither the huge technological change to "gigabaud fibre optic connections" nor "rendering three dimensions at high resolutions" have invalidated any of their theories and axioms.

    I bow to these pioneers of cyberspace and hope that one day some company will finally heed their warnings and guidance.

  7. @Joe Rigby

    The broader sense the question is not "Is Second Life Doomed?" but whether the entire industry is afflicted with amnesia and doomed to repeat mistakes of the past. At this juncture, I can confidently say unless changes are enacted to the fundamental architectures involved overall, then yes, Second Life (as well as every other virtual environment) is doomed to repeat history.

  8. "so I feel that while they (Open Sim) have mostly the right idea, in the end it seems like they are likely to succumb to the same issues over time."
    So lets say Open Sim is 70% the way there with the decentralized approach does the other 30% involve inworld collaborative UGC development to some extent.

  9. Very good pot, Aeonix, and definitely required reading on this subject! However, may I be the discordant voice and not fully agree with you on each and every aspect? :)

    From a purely technical perspective you're right in your assessments, even though "centralisation" is not that bad as it sounds, and it depends a lot on what you actually mean with "centralisation". Facebook is "centralised" — just heavilly distributed. Google is pretty much the same: "central" data sync'ed around hundreds of thousands of edge servers. Second Life distributes assets via Amazon's cloud — that was perhaps one of the most "silent" accomplishments that might have gone unnoticed to you. Aye, I agree that SL's model is not "fully decentralised" (in the sense of Facebook or Google or Hotmail or any other similar model...) but a hybrid approach: redundant central asset servers which distribute cached content over the cloud, and partitioning of the grid in independent regions, where each region simulator just handles a fraction of the overall load. No, it's not a perfect solution. Yes, I'm aware there are serious bottlenecks. Yes, I'm aware it's not a "classical" decentralised or distributed model. But it's not as bad as you paint.

    From a business perspective, I feel you're misrepresenting Avatar Reality's position. They're not "broke" in the usual term of the word: they're a VC-funded start-up. Start-ups start with an idea (in the techie world, this is usually just a techie idea, not a business idea) and enough capital to survive 2-3 years. 8 out of 10 will utterly fail, and that is expected by business angels and VC company owners. 1 will break-even. 1 will be a huge success and let them recoup the investment on all 10 companies. That's how things work; that's why 80% of all VC-funded companies "fail" — they never had such a good idea that translated into a successful business model.

    LL was also "broke" in 2003 when their funding was exhausted, and like all start-ups, unless they came up with a successful business model, they would have folded like 80% of all start-ups. Luckily for us, they actually struck gold — they kick-started a virtual economy where people can rent land and buy items from other residents for L$. This was not planned from the start (LL predicted that all their income would come from monthly subscriptions like all commercial MMORPGs), but rather a very lucky decision, which by sheer coincidence actually translated into a successful business model. They never needed to apply for funding again.

  10. Avatar Reality, while in theory emulating LL's business model, were not so lucky. And here I think you hit the nail with the technological hurdles: they adamantly refused to create a "lightweight" viewer from the very beginning, because they "thought" that their target market were hard-core gamers and game designers which would love to come to a "serious" game platform (as opposed to LL's renderer). This shows that they, like most start-up companies who have since then failed (and some industry giants like Google with Lively), never did any actual market research. By 2009, when Avatar Reality launched their open beta, they could easily have seen that SL's resident population is anything but "hardcore gamers", and even though SL's viewer is far too demanding for "regular" users, and way too complicated to use, it barely allows a million regular users to be happy to log in every day or so. Anything far more demanding in terms of hardware would simply fail. I think they were encouraged by a lot of professional designers with state-of-the-art high-end desktops who joined BM and were very encouraging on the forums — they felt they were designing for a company that knew what kind of requirements a professional designer needs. But how many of those have been attracted to BM? 100? 1000? Probably not much more. That's not enough to cover the running costs.

    I have to humbly admit that I had no idea about the costs involved in a streaming solution; your article was an eye-opener in that regard. So instead of doing a lightweight viewer, trying to get a larger user base, and thus offset the running costs by generating more income, they opted instead for an expensive solution which would push the break-even point to an even more distant date in time (if at all). No wonder that their funders were unhappy about the decision.

    I also have to confess that I didn't understand that Avatar Reality would be streaming content to iPhones! I thought they would be doing a native iOS application — the description of what they intend to do is something like IMVU for the iPhone. But I always thought it would be a native application, roughly inspired in Blue Mars and using its brand, but pretty much detached from the BM experience. Hmm. Well, pushing 1 GB/hour over a mobile phone network is definitely a business for the operators; perhaps they are expecting to establish some kind of business relationship with them. For operators with metered bandwidth, intensive-bandwidth applications are quite welcome! I can very well imagine Avatar Reality getting a fraction of those fees to support their running costs in the streaming solution; it's just a question of making the right deal with the right kind of operators. For the ones with a flat fee, of course you're right, they will probably ban BM on their networks...

    But thanks for the historical perspective. I totally agree that most start-up CEOs never do any homework outside their technological know-how (and their funders even do less). It's not just looking back to history; it's also about market research. The "build and they will come" approach is hardly an established plan that leads to success, as so many residents in SL (and companies which tried to establish virtual presences there) have found out in the hard way. Sometimes, of course, you're lucky. LL was lucky by hitting by chance on a successful business model based on tier and a free (in the sense of not controlled) market economy of user-generated content. Avatar Reality was not so lucky.

  11. @ Gwyneth

    Definitely a refreshing counterpoint! Although I cannot say that LL "got lucky" with their approach. Massive cutbacks in staff and the musical chairs approach to CEOs would indicate otherwise. If anything, I'd say that LL is simply holding out longer, but the outcome is likely the same.

    When I say "decentralized" I mean just that. The cloud is not decentralized and relies on a single point of access to work, and as we've seen before that doesn't work reliably (over the summer of 2010 I recall many times when 500 Internal Server Error was common).

    1GB per hour in based on observations concerning Skylight and Cloud rendering, and I'm merely speculating that it would be similar with OTOY cloud rendering for BM. In terms of streaming a lot of bandwidth and the mobile providers, I believe it would be the exception and not the rule that such options would exist, which to me seems like they are needlessly cutting their adoption rate further than they already have.

    As for "broke" you're right, it's not entirely true. BM is broke in the sense that additional funding has dried up and they are using whatever they have left to put toward mobile applications. But seeing as their business model is essentially the same, I would say that even in mobile it will have the same consequences if not magnified.

    In the case of LL, again, I'm not so certain they were lucky as you put it. I think the top level business model is a good one, however it relies on bad underlying architecture to sustain, and that is building a castle on sand.

    Thank you for your input :) It's much appreciated (and highly interesting) to read.

  12. @ Joe Rigby

    I think 70% right is pushing it. Decentralized (as I've said) means just that. It means reducing or eliminating the need for many people to stand in line at a single source for redundant information. OpenSim I do not think does this, however well they manage to decentralize the overall servers, we're still standing in line at each of them for redundant information.

    This is, I believe, that extra 60%, so if anything, the numbers for openSim would be reversed: 40% right, 60% wrong. Which is still better than SecondLife which I'd say are 80% wrong and 20% right.

  13. Anonymous7:56 PM

    "Only in the end, do they realize the "ignorance" of the Dark Side.;)"

    The lack of wider business models that can succeed and be sustainable has been the way of web3d tech/vc companies for almost 20 years.

    LL did not "invent" virtual economies of 3d assets. They go back to the 90s as well.

    Each failures "answer" is always to "build" a new widget, since thats what software/hardware engineers do to spend others money....

    making products, selling products, and managing services are not these companies leaders/owners strengths.

    Its always the same bankers, with younger -newer- grad students every year.- Each generation more "wired" into the internet and it's "relevance engines" for information. History is constantly being deleted from value. And beta or obsolete is the only mantra of the machine makers and machine made (humans).

    5 years from now this cycle will be repeated again.

    The medium of RT3D will continue to find usage in entertainment and simulations communications... but mediums alone are not a buisness... but things that you do business within.

    anyhow. hawaii must have been nice...and so it goes.

  14. Could you comment on Open Croquet / Open Cobalt / Teleplace, which uses a decentralized (P2P) approach? Do they do it right?

    Also, if server-side rendering is such an awful idea, then how does OnLive stay in business? What's their secret sauce?

  15. @Troy

    As I've said, it's one extreme or the other. Centralized or P2P methods. Each has their own strengths and weaknesses, however when taken together, they compliment and fill in where the other may be deficient.

    As for OnLive, it works because it's a gaming service that is rendering limited instances. Plus the subscription fee is compelling to the user and covers costs and a profit margin. I'd rather not compare a gaming service to a virtual environment or vice versa.

    As for open Croquet, Cobalt, etc. The very thing that gives it strength (Peer to Peer Networking) also is its biggest Achilles heel. The same is true for centralized approaches.

    The secret, when it comes to Virtual Environments, is knowing that they solve each others problems.

  16. The original intent of Open Croquet was in fact a peer-to-peer approach, which is how we usually demonstrated the system. When we built Teleplace, the enterprise market required a more centralized approach (for many reasons, including security, document management, credential management, etc...). My original intent was that portals would be generic links to any other virtual world - basically a 3D web. I still believe this is the proper approach.

    I also agree that using video to provide 3D interaction is not a good idea. Interesting story is that when I wrote The Colony (1987), I decided to go with 3D primarily as a compression medium. You just could not fit a big graphics intensive game onto a 440K disk. You can see it here . I believe this is still the right way to approach the problem, and the key to making this a success is embracing procedural models of the world.

  17. Just want to comment quickly that Bitmanagement.de is, in fact, the successor to Blaxxun.com. Blaxxun finally folded their website when the licences for their community software expired. Only a Korean university and a private individual bought Blaxxun community 7 server software. The Cybertown world still runs on community 3 or 4 software. I go back home to Cybertown occasionally just to see if it's still on life support.

    Personally, I would think X3D might be a viable alternative, if it could be incorporated into HTML5 without the need of a code plug-in, or stand alone viewers.