1. Joe Rigby said on August 4th

   How about the future of VW’s based on the tried and true robust gaming engines. Immersive environments using high polygon static meshes and photo-realistic textures. Well just a thought!

   http://www.calebbooker.com/blog/2008/07/03/second-life-vs-unreal-2004

   http://mellanium13.blogspot.com

   http://dusanwriter.com/?p=481

   http://electricarchaeologist.wordpress.com/2008/04/22/autocad-into-unreal2/

2. dandellion Kimban said on August 5th

   Availability of the world through the browser, and, especially, lack of the need for additional installation (which exclude Lively) can do a lot for the popularity of a specific world. But, imho, there are more important tings than that. Intuitiveness of the UI, performance, content and features will have greater impact. Of course, there is overall experience that will count as well.

   But, at this moment, it is hard to expect that two models will fight so that only one survives. It is more likely that they will both develop, each in its own direction. One to the easier access, other to the more features and stronger experience. (With this in mind I am happy to hear Philip talking about porting SL to browsers.)

   Also, I don’t think there will be a definitive winner in the world competition. Sure, some will die, butt there will not be only the one at the end. The winner will be the one that will succeed to set the standards and make Intergrid.

   Microsoft’s interest in 3D, actually their attempt to catch up with Google on that field is not promising much. Hardly tt Microsoft will step out of their platform, and leaving Mac and Linux users witout a tool for a technology that should cover all of the Internet users is not a good strategy at this moment. Also, at te moment wen standards and opening of technologies comes into play (and that is already happening) MS will be in a problem as many times before.

3. Digado said on August 5th

   @Joe - Thanks for commenting and providing these links! The Unreal2 Engine as described in your post and the ones you’ve linked is certainly an option to provide additional value in client-based virtual worlds. Blue Mars is going to be a virtual world based on the Crytek2 engine - supposedly one o the most graphically advanced engines out there.

   But smaller (and cheaper) engines are also an option - the OS OGRE engine and the engines as provided by virtual world developer platforms Blink, Unity and Quest3D. But they all face a problem in getting people to accept the application of virtual worlds, and move them away from being ‘just a game’ in the perception of the masses.

   @ dandellion - I agree there might not be ‘one winner’. A lot of the media (TechCrunch) has played the launch of the browser-based virtual worlds as a race against Second Lifes ’success’. Like Cuil was supposedly competing with Google, suddenly each virtual world is competing against Second Life.

   However, there is a certain truth in the competition - the visible brands have moved away from Second Life, user numbers seem to have frozen and even the large metaverse developers like MoU and ESC have moved their business elsewhere. Second Lifes growth is at a standstill while virtual worlds as a whole is a growing industry. In this context, Second Life is indeed being threatened by other, perhaps more accessible worlds in the near future, its not hard to see where the business model of SL fails without ‘fresh blood’.

   Furthermore, I wonder if ‘more’ immersion (experience) will really outweigh accessibility for a significant niche. People are immersed in books, its really not the graphics that do the trick, its the story (or context), and embedded worlds might just prove to have the edge in providing this context through the 2D web (and provide customizable UI’s, and don’t have to deal with the ‘noise’ of other, unrelated content).

4. Pavig Lok said on August 5th

   Browser based worlds won’t replace client based worlds for two reasons:

   It’s easier to compile a browser into a rich 3d world than it is to compile a rich 3d world into a browser. Most of the compelling use cases for virtual worlds rely on immersion and attention - the more one feels one has a presence in the space the more powerful the medium becomes.

   If you’ve ever attended a well run conference in a virtual world like SL you’ll have seen some of the positives of this approach in terms of focus. A bunch of people all looking at the same speaker and slide keeps the experience fluid and folk on task. Breaking that out into multiple windows destroys this benefit. The tools could certainly use work, but as rich clients and worlds increase in funtionality we’ll see more of this. Education similarly requires strong focus and presence, which is broken by splitting learning tasks into multiple media rather than keeping them as part of a shared experience.

   Secondly: the browser is an extremely poor user interface for virtual presence. Lively for example runs in the browser, and suffers from the attention splitting that happens when you put anything on a web page. Most folk have multiple tabs or windows open in browsers, and when you’re looking at one thing you’re not paying attention to another. Interaction in these worlds tends towards folk spending most of their time effectively AFK as they shift focus and cease to pay attention to wherever they’ve left their “body”. The browser is becoming much more complex, but it’s roots are as a document display system and there’s only so far you can push it before it makes little sense.

   So for the strongest use cases monolithic clients will win out in the long run. Lean 3d worlds which run in the browser will appeal mainly to the tweens market due to their limited applications - which are probably better served by leaner chat systems such as IM. Technically we are closer to seeing solid flash content on a browser inside a virtual world client (croquet demo anyone?) than we are seeing extremely rich virtual world interaction inside a browser plugin. To achieve the complexity required the plugin simply ceases to be a plugin and becomes an application of its own.

5. Digado said on August 5th

   Think its the first time I don’t agree with you Pavig :P

   As posted above, I think its rather odd to place ‘immersion’ exclusively in the hands of sovereign applications such as client based VW’s - Immersion is about the narrative, the white rabbit to show you the affordances that trigger the imagination (and therefor the experience). I am convinced browsers will, and have offered similar experiences to the immersion in Second Life. If it was really just a technical issue, its not that hard to create a full screen flash world from inside the browser (doesn’t even require a pop up, see how its done in the suite xx3 demo somewhere on this website). Flash player 10 will address the graphical card, and physics engines and multi user applications have already been build in full 3D. Given the engine based client applications had at least the last decade to develop, and Papervision started in 2007 I’m not all that convinced of your theory on flash in VW’s.

   To me, your arguments sound really technical (therefor not wrong per definition at all - we are in the stages of technical development), but int he end its about people. Will they adopt ‘game worlds’ without seeing the benefits first? Is ‘the level of immersion’ really the deciding factor in the success of virtual worlds if indeed, SL offers better immersion just because it has glow and rendered shadows (well not yet, but it could have) - or will those just get in the way of creating meaningful applications rather than pretty ones. I’m just not as sure those questions will be answered in favor of client based virtual worlds in the near future.

   I don’t really like anecdotal ‘evidence’ - but there is one i keep thinking of so I’m going to add it anyways. I have the latest Outlook installed on my computer - has a lot of neat functions and a decent interface. It synchronizes with all my email accounts and I like how it works in general (especially the ’select folder > mark all as read). However, I never use it. Why? Because either at start up it launches when i don’t need it, continues to download my email for a very long time, even the emails i dont want to read, then the application keeps running on my computer all the time even when i don’t need it (can’t prove it but I think it really slows down Vista for some reason). The program I do use is gmail - i visit it when i need it, it doesn’t run when i don’t need it except for one subtle reminder. No loading times, and I can access it anywhere, I can use the data in other parts of the web (iGoogle for instance) and it runs on my iPhone. You don’t think similar choices will be made when it comes to virtual worlds, and the applications thereof?

6. Pavig Lok said on August 6th

   I understand your objections regarding the rapid development of flash based applications but there are many problems to solve before fully immerseive worlds are available within the browser. For twenty years we have had 3d game worlds and avatars available to us, but even now a high quality experience of that sort, which also functions as a dynamic virtual space, requires more than just the ability to display 3d. There is the issue of streaming content.

   So if something like imvu or lively can give us an in browser experience, which is 3d and prettier, but still about the complexity of lucasarts habitat (1986) in terms of avatar interaction, we have a long way to go. To increase the available options in such a world we need to include more tools - so anything useful for a particular application will by deffinition be either a botique solution, or make strong concessions to practical limitations of the technology.

   The simplest way of expressing this argument goes thus - gmail is the example you used, though I’m going to use google maps as an example as the use case is easier to describe. Say i want to discuss a map or tour itinerary, etc in a virtual world… to do so i choose to stick a google map on the ground and walk about it with some folk to discuss. Google maps runs on a browser currently - making a google maps plugin for a virtual world without browser hooks would require a botique coded solution at best, and would have reduced functionality. So i choose to use the full web based google maps rather than hire coders for months or buy an off the shelf solution that only solves half the problem.

   So now we have a browser in the virtual world, as an example of an imported tool fit for purpose.

   In a browser plugin based world it would be rediculous to compile a full featured browser into the plugin that runs in the browser. Where do we stop? Does this virtual browser run flash? If so then we have a browser with a flash plugin that runs a browser with a flash plugin and so on ad infinitum - there will be an arbitrary limit as to how far you can go. And in doing this you have made the enclosing browser redundant.

   A simpler solution to these rich applications is to run a monolithic client - it can easily encapsulate browsers, office applications and other such things without leading to problems of infinite regress. As virtual presence is a representation of self - and it makes sense to only be present at one place at a time in real-time - a single virtual presence via a rich client makes more sense for this application.

   Unlike the web which is temporally flexable and can be dropped and picked up at a later time without impacting the experience, virtual worlds are a realtime medium. Just as it is extremely difficult to engage in multiple voice conversations at once - it is difficult and counterproductive to be virtually present in multiple places. The browser however allows one to have multiple documents open and switch between them at will. So virtual worlds are not a comfortable fit for the browser, though within limited applications they may be ok.

   What I’m trying to get across here is the hard limits on complexity that you get with a plugin. As the applications within virtual worlds develop in complexity there will be a point at which the browser plugin becomes so complex that it becomes an application of it’s own, and the plugin itself will simply be window dressing wrapped around the client software to make it “look” like a plugin. This is in fact the case already with google lively and imvu even with their limited functionality. They aren’t plugins, they’re clients that look like plugins. Secondlife itself has already been demonstrated running in a browser using the same tricks.

   So for limited applications, sure plugins are the way to go. For virtual presence though, the complexity of providing a rich feature set will push us into larger client software. It is simply hugely inefficient to build in a rich enough feature set and wrap it in a browser when simply putting a browser into a seperate virtual world client is so much simpler.

   Working completely within the browser has many benefits but also necessitates many concessions. Sometimes those concessions will make sense, but for strong or complex use cases they will be unacceptable. A rich client is the only thing that makes sense. Google docs for example is often useful, but we haven’t replaced openoffice or ms office, and we have a long way to go before we do. We may be critical of the limitations of a monolithic application like openoffice, but it is fit for purpose and focused in a way that the browser based alternative is not.

7. Digado said on August 6th

   Okay, so if I understand you correctly (and I want to make sure I do) you are simply saying browsers will not be able to offer the same kind of experience as client-based worlds can provide in the foreseeable future? Because that would be where I simply disagree but that’s not something either of us knows, just what we think.

   The only point I wanted to add, is you talk about browser based virtual worlds making concessions, but client based virtual worlds are making a lot of concessions as well - they are closed, require significant hardware, not likely to be visited lightly, need huge adoption curves, are sovereign applications that don’t integrate well into everyday life and are inflexible. All of these points are, what I believe toe be, the major deciders. The rest is fine for differentiation and smaller niches, but is not going to win anyone over from a people perspective in the situation we are in now.

   PS. I have long, long since replaced my MS office (and certainly OpenOffice, absolute junk) with Google Docs (or the fancy adobe buzzword), but lets not get into anecdotal evidence anymore :p

8. Tele3dworld said on August 7th

   The future is shurely in the projection based 3D technology like CAVEUT; the nerdy PC screen as an interface will gradually be replaced by tele-immersive spaces. PC’s and peripherals will disappear into the background enabling a greater experience and learning just like the real world.

9. Rob said on August 7th

   I don’t think that downloading a plug-in to deal with a browser-based 3D world of lesser quality will make users not want to download a client for something like Second Life which is nicer, faster, and contains more than something such as Lively.

   Of course this may change if/when browser-based plug-ins deliver equal or better quality and features than a fully downloaded client 3D world.

10. Pavig Lok said on August 7th

    Browser based plugins will _never_ produce a better quality feature set than a standalone client because, as they increase in features they cease to be plugins. Even so - if they act like browser plugins, the browser becomes an artificial limitation on possible functionality. You will always make concessions to embeddability.

    That’s not to say that plugins may not become compelling - just that they are at a severe disadvantage in outrunning standalone clients when it comes to feature integration. The browser layer is another level of abstraction to bug hunt within… and we all know that there are a lot of browsers to contend with. Mozilla, IE, Safari etc. So in terms of adding features we are not just dealing with multiple operating systems but multiple browsers, and combinations thereof. This significantly complicates the provision of a compelling in browser experience across multiple platforms. Standalone clients are simply easier (and so cheaper) to develop and maintain.

    This might not seem like such a show-stopper. A platform could get a significant market share by simply targeting the most popular configuration of browser and OS. The problem for virtual worlds though is that the killer application will probably be something to do with collaboration - social networking on it’s own doesn’t make a lot of money, business does.

    In a collaborative environment the trend is towards cross platform tools at the moment, and business (including edu) and government purchasing and investment follows the general trends of it installation. Large universities and corporations typically have mixed technologies as a legacy of differing layers of purchasing or departmental policy. Windows free boxes are high on the purchase lists this year, especially in europe, and more macs are being integrated into departments too. A few boxes on the wrong platform will kill a purchase decision for any collaboration tool, which puts single platform solutions at a disadvantage - IT departments don’t trust them.

    So that’s the logic - plugins are more expensive to develop cross-platform to exploit the markets which actually have a dollar to spend. Applications which have a hope of leveraging virtual worlds in a way critical to business require full feature sets and cross platform compatability. Plugins deffinitely provide easy access to the kind of reduced functionality that is required by the tweens market. This will be a growth area, but is currently saturated with a lot of projects competing for fairly shallow pockets. There’s room in the market for both, but plugin worlds don’t present a threat to client worlds for the forseeable future, as we have a long way to go before those markets are established, let alone merged.

11. Digado said on August 7th

    I really don’t see these ‘concessions on functionality’ - take a look at the games at ‘Instant action‘ - a plug-in game center offering high quality 3D, mutliplayer options and fast gameplay. Of course this is a pretty extensive plug what if such a thing because as ubiquitous as the Flash or Java plug-in?

    I don’t think you can say plug in worlds are more expensive at all based on different OS’s. Java and Flash are both cross-platform while SL and AW struggle to keep up with the Mac (AW is completely incompatible with the Mac and SL ‘no longer supports it’). It all depends on (1) the goals of your virtual world, (2) the usage and (3) community dependency to determine what platform would be cheaper to develop.

    Also, you represent the plugin like it has to scale with functionality which is just not true. Weather its java, flash or the Lively plugin, as you well know its just the browser understanding a certain type of ‘language’ that can add new functionality, and provide it in an accessible manner. Once you have flash installed, you don’t need a new, or add-on plugin to provide physics or different types of data exchange/functions - I really don’t see your point in them ceasing to be plug ins with increased functionality. They are just ubiquitous gateways to applications that bring down the obstacles that would otherwise turn a large percentage of users away, in exchange for the concessions as listed in your post above.

12. Opensource Obscure said on August 7th

    The title of this article seems a bit provocatory and misleading to me: even if other platforms are now providing high graphical quality environments, noone of them leverages on unrestricted user-generated content like Second Life does.

    I think this is a really important and peculiar feature of Second Life - maybe even more important than its working microeconomy system or its native programming language. And it’s difficult to think that platform A can be replaced by platforms B and C if these ones don’t offer a fundamental feature of A.

13. Pavig Lok said on August 7th

    AHA! There’s the rub - within something like flash, which will grow to be sure, there are limitations on functionality. One of those limitations is that anything complex needs to be coded within the plugin or compiled in. Creating a word processor in flash for example is extremely baroque compared to compiling one directly for a target application/os…. so to create a virtual world with shared text editing in flash for example (to provide a silly example) one must twist the available resources in flash into a shape they were not designed to be in . Anyone will tell you it _can_ be done, but the solution is hopelessly baroque and heavy on the client. It will also take a lot of downloading for that code, making the flash based solution laggy and uncomfortable to work in.

    Solving a single problem this way may be workable, but it means an expensive to develp botique solution, because feature loading in such an environment produces diminishing returns. This is the basic laws of coding - just because you can, doesn’t mean you should.

    So flash, sure, can do some things that are like rich virtual worlds, but not many things at once, and at a severe penalty in terms of performance, bandwidth and functionality.

    When the balance is right and the application is light, flash wins…. When the job is complex, a dedicated client is orders of magnitude more efficient and fit for purpose.

    You don’t have to believe me on this - ask any coder. All computing approaches are by deffinition “turing complete” so you can do anything using any system available - the only limitations are memory, bandwidth and processing cycles. Flash can certainly create rich virtual worlds, but it will do so ineficiently, in a bandwidth unfriendly way, and with a clunkiness that puts it well behind a purpose coded client. Just because you can do something doesn’t mean you should.

    Google docs is a good example of an application which gets an 80% solution by dropping functionality - it exists quite happily in a browser. Then again people don’t use 80% of what word offers so there’s not much to complain about. Then again it hasn’t taken over the market yet.

    20 years ago browser world level functionality was demonstrated in the client domain but that hasn’t changed the world either. Only in the last 5 years or so have more flexable virtual worlds emerged which can do anything compelling. So browser based worlds may well provide a competitive advantage, but they have 15 years of evolution to catch up on to provide the same functionality as SL or Croquet. This may be the time when restricted use worlds start to garner public interest, but assuming that they’ll catch up in terms of functionality with client based worlds is a folly. Technically they’re just not fit for purpose.

    An easy way of understanding this goes thus - google docs is fairly simple, apart from the saving online bit. You could find and download an application which does pretty much everything it does which would fit the same memory footprint of an image on a web page. To use google docs you suck down this data many times within a session - making it a convenient, but resource hungry solution. Imagine now that we take the same approach to producing a rich virtual world - providing the working code to run the world, and unoptimized fat processes to keep it running, will take much of the bandwidth and processing power we could better use for responsive, realistic in world interaction. It’s just wasteful.

    Clients though can be finely tuned. Probably the one unsung radical idea that secondlife bought to the table was streaming 3d - Philip Linden brought that technology from his work in streaming media at Real Networks. This is why secondlife can load a quarter of a kilometer of world within seconds whereas something like lively may take a minute to load a room of 50m diameter. Those kind of optimizations need to be part of the underlying structure or the implementation will be klunky and wasteful. Flash may well be able to display the data - it just can’t load it in the most efficient way, nor can it optimize display with the same level of tuning that the SL client provides - unless it is designed to do it.

    If they design flash to have all this functionality built in, then it’s no longer a plugin - it has the complexity of a client so is effectively a client. The browser then becomes supurfluous - just extra baggage for the plugin to carry around. It’s not about adoption, or hardware resources or anything like that. The plugin route is a stopgap and virtual worlds will eventually outgrow it. If you don’t like outlook you’ll deffinitely dislike being in a virtual world of the complexity of secondlife (evolved a bit) running inside a browser window. It’s un-needed fat.

    It may seem that i’m banging on about this subject but I feel that it needs to be said. The vogue at the moment is for browser plugin based worlds. Browser plugins have taken over casual gaming to be sure, and they’re certainly fit for purpose in that. They’re also not doing anything particularly fancy that wasn’t done in the early ninties, game or graphics wise. Virtual worlds have been around for about as long - and mostly unsuccessful until recently. They’re complex applications and folk are only just starting to work out how to get them to do interesting things. Plugins can not be expected to catch up.

    Plugins are by deffinition either generic or purpose specific. If they’re generic then any solution will be sub-optimal. If they’re purpose specific then any solution will be of equal complexity to a client solution. Either way they’re at a disadvantage - as it’s not technically possible to add functionality (such as browser embedding) without a hit in optimization. If the race is hearts and minds, plugins are fine. This talk I keep hearing though that plugins will overtake clients in functionality is mis-informed about the engineering challenge. You just can’t argue that something embedded in a browser will ever outpace a dedicated client on technical grounds.

    The market will gravitate of course to the most convenient solution for various use cases. A client is a barrier to entry of course. When reduced functionality is needed a plugin will probably be the preferred solution due to lower “cost of entry”. Whenever the application techically pushes the boundaries of what is possible within a plugin, the cost of entry of a dedicated client will outweigh the inconvenience.

14. Digado said on August 9th

    @OpenSource: You are right of course to say the title is a bit provocative, but that’s what its intended to do. Not just to draw that extra couple of readers, but by provoking a (counter) reaction you could be more inclined to think about it, and express your thoughts and create valuable input - so it servers a purpose.

    @Pavig: No it doesn’t feel like you are ‘banging on at all’ - I posted this for exactly the reason of this discussion and I’m very thankful you expressed your thoughts :) The thing is - I agree on your argumentation - yes flash has limitations and has to make concessions, but my point is so do client based worlds. Now who makes the bigger concession at the moment. Is it:

    1. ‘Limitless possibilities in terms of functionality’ (i.e. combining applications, tools, and graphics more fluently)

    or

    2. ‘Reach and integration into new markets and the routines of people’ (i.e. having access to/being able to play anywhere, anytime, to a potentially huge network, and play nice with the rest of the web)

    Between those two, your argumentation has made it clear functionality has a long way to go in browser-based worlds, but to make a definitive conclusions, you’d also have to look at the potential for clients to reach into untouched markets. Will the tweens growing up adopt client based worlds or will they look for something as convenient as browser-based worlds and settle for a ‘quick fix’. SL’s ‘millions’, even with There.com, HiPiHi and Kaneva and other client based worlds are but a fraction of Habbo, Neopetz, Cyworld etc. Perhaps the question is if these browser-based tween worlds are more likely to cross over to untapped adult markets, or the more adult clientbased worlds are likely to tap into the (growing up) tween markets.

15. Pavig Lok said on August 9th

    Interesting questions…. oh and a timely announcement - the SL client as a browser plugin http://nwn.blogs.com/nwn/2008/08/introducing-xen.html (not the first such announcement but certainly the first such announcement today :P )

    I don’t think the 1 (limitless possibilities) or 2 (tightly focussed integration) are either/or propositions as virtual worlds develop. I really think worlds need a backbone which plugins don’t provide - but also acknowledge that the lightweight plugin approach has plenty of benefits as well.

    As worlds evolve I believe we’ll see integration between the two. I have a pet prediction about how this may happen.

    Second Life is the canonical (at the moment) large multipurpose complex world - and lively is the classic browser based rooms world. Lively can’t presently expand to secondlife scope, nor can secondlife reduce down to something as simple as lively. However - land can be divided up in secondlife and “parcels” can be alloted different types of functionality - and this may be the key to a crossover space between these different ends of the spectrum.

    In SL one can turn off scripts within a parcel - at the sim level one can even turn off physics and other such complex weighty stuff. This switching off of functionality is often done to allow greater numbers of avatars into an audience for a function. Imagine that we could designate a parcel of land in SL as a “room”. Rooms would have reduced functionality (simplified physics/scripting etc).

    For people within the secondlife universe it would be a part of contiguous space and could be entered like any other location, but would have reduced functionality. For a web based client though the room would be a lively like unconnected area - entry via a web plugin would only allow you to see whatever was in that parcel or room (and anything outside that could be burned into a skydome to provide an aparent “outside”. Folk on a lightweight (web or iphone?) client could hapily coexist with people using a full client in such a space. The “room” could be given a web url and accessed via a plugin when required.

    This kind of approach, where there’s a solid complex multifunction world, with gateways into lower rez spaces would allow for much greater integration. The lightweight world client can act as a gateway into a more featured environment, and of course the complex world also gets an easy public access crossover space. Like a foyer.

    This kind of approach makes more sense to me than a large number of worlds competing for the same space but remaining unconnected, or making sacrifices to fit into a market space which will later leave them painted into whatever niche their public decides they fit. Creating a highly flexible virtual world with lots of tools and uses, then stripping out the fat when required in order to serve up snack size slices would solve a lot of our problems.

16. Client versus Browser Part II | Digado said on August 9th

    [...] will the rich, client-based become the dominant model for virtual worlds? Not even a week after a lengthy discussion on my previous post on this topic, some major clues arrived to the scene as to how this ‘war [...]

17. Digado said on August 9th

    I summarized part of the discussion, and ongoing developments in a new post:

    http://digado.nl/client-versus-browser-part-ii.html

    And included your vision and argumentation Pavig, I think that is what we agree on for the most part (99%) :)

18. naima aya said on August 16th

    Hi, Just receive the newsletter and came to check your review. Thinking that I was writing an article about it too (in my native language), not in the same perspective because I can’t compare the web browser 3D environment with Second Life it self. So I am on focus the Second Life future with IBM “partnership” recreating a new way to look the 3D virtual worlds, It’s possible in some years we can see almost the same type of 3D worlds connect each other and I do believe that they are going to keep their characteristics. But I do like the new way to see a web page based content in a 3D environment and it gives a magical touch to a static page.
    Lively does not excite me, because since I’m an second life resident and and content creator I would preffer in 100% to keep my second life. Lively as I saw you must live with their content, (lol pretty weird), but sure others that never saw Second Life or don’t understand it, would love lively as a 3D dynamic chat room :)

    Nice to read you,

19. Dale Innis said on August 18th

    I may be overlooking the important points here, but it seems to me that there’s no such thing as a “browser-based virtual world” (or a “no-browser-based” one for that matter). Whether or not the default, or most popular, or officially supported, client for a given virtual world runs inside the browser window or not seems like a relatively small and not that important fact about the VW. So any blanket statement about how “browser based VWs” will (or won’t) dominate the market seems like a category mistake.

    Lively is “browser-based”, but it requires a special download, so it doesn’t have that “runs without a special download” magic about it. Second Life is usually used with a large client that runs outside the browser window, but if for some reason you want to have it use the browser window there is (as was pointed out above) a plugin that enables that. Clients based on Flash 10+ will eventually be able to do true 3D, so an SL client that ran inside the browser window without a special plugin (at least if you’re alreaady running the right version of Flash) will be possible. Does that mean that SL will instantly jump from being “non-browser-based” (and therefore doomed), to being “browser-based” (and therefore not doomed)?

    I think this whole “browser based” issue is really a red herring. End users are perfectly capable of pressing a “download” button. Zillions of non-technical people use WoW, for instance, which requires a heavyweight local client. Unless we have some real evidence that being “browser based” or not is really a big differentiator in adoption, I would suggest against obsessing over this particular feature of a VW. Others (like user-created content, for instance) seem much more likely to matter…

20. Digado said on August 18th

    Thank you for commenting Dale,

    The thing you might be overlooking is integration into existing processes, which is really the only way VW’s will ever reach out beyond a gaming niche. Dominant (sovereign) client based applications don’t generally do very well unless its object oriented, in other words, knowing what you will do inside the application beforehand instead of casual use applications which is what we are looking at now in the case of Second Life, or other social virtual environments which are not game-driven.

    So this type of required accessibility is really more than just a better interface or better software, to me its changing away from downloads and installations as a large barrier for the casual user (WoW players are NOT casual users of VW’s or your every day internet user) and provide different portals into a virtual space - the browser being the most obvious, and most convenient one as its already ubiquitous and fully capable of running 3D with the least amount of effort for the end user.

    Of course going into the browser is not the end all solution, but its the start of opening up to a wider audience you wont reach with better graphics, or more features as long as they need to download and install simple because they think it might not be worth the effort. But if you could provide an incentive before the download of a ‘full feature’ world for instance, the process of persuasion becomes a lot easier.

21. Dale Innis said on August 18th

    Hm, there’s two possible meanings of “integration into existing processes”. In terms of integration into the way that humans share pointers to things, the important thing is being lightweight and easy to consume, not so much (I don’t think) displaying yourself in the browser window. That is, if the first time clicked on a URL pointing into virtual world X, they ended up on a page where they had to click a “download” button, and every time after that the client just came up quickly and put them in, that wouldn’t be any particularly large barrier to entry I don’t think (it’s basically what happens with Lively for instance as I recall, and everyone counts that as “browser based” anyway). What’s most important here is that the thing be easily consumable, not that it run inside the browser.

    If we’re talking about integration into back-end, programmatic processes, then that’s pretty much entirely different, again, than the question of whether the VW runs in a browser window. As far as I know there is no way to get data in or out of Lively or Habbo Hotel or any other browser-based VW that I know of, whereas Second Life at least has limited HTTP support (outgoing, soon to be incoming) and RPC (incoming, soon to be obsoleted) and email (both ways). So if for instance you wanted to integrate part of a build in a VW with some datasource (or control knob) in RL, right now you’d have to use one of the “non-browser-based” VWs as far as I know.

    I think we’d do better teasing out these issues that actually matter (”how easy is it to start using this VW?”, “can I get my data in and out of this VW?”) from the issue that’s simpler to talk about but doesn’t really matter (”does this VW’s most widespread client sit inside a web browser window?”). Something like that anyway. :)

22. Digado said on August 18th

    Well that all depends on how you look at it - you say one of the issues is ”how easy is it to start using this VW?” - Now when you look at the data we have from current VW’s the ones running inside the browser outscore thouse outside the browser 10 to 1 - if not more; so perhaps one of the solutions is placing VW’s inside a browser, a ubiqitious tool which is already a window into many types of data i’d want to use inside a VW, except it hasen’t been visualized yet. That COULD be a solution to one of the issues of virtual worlds right now - which is what i wrote about. Then you can go back to an abstract level again and raise the same question (”how easy is it to start using this VW?”) and see if there are other ways- one of which came up in the comments and in a second post i wrote, like modular worlds. Different portals into the same data and visualization.

    Are we there yet once we have it solved? Of course not, and to be honest i think these two ‘important’ questions are not really all that important at all as long as we keep looking at 3D rooms with powerpoint as the ‘killer application’ of the metaverse :) Meaning, application, goals, those are the important questions, this stuff is secondary but is bound to play a part in the adoption of VW’s none the less. Which is also where it goes back to my original point, once virtual worlds provide more clear goals, more people will be prepared to download, and the obstacle becomes less of a problem in the process of persuasion.

23. Dale Innis said on August 18th

    Okay, that’s fine; the current browser-based virtual world clients are in fact generally easier to start using than the non-browser-based ones. Using only what’s generally available in widely-deployed browsers is one way to help make a VW client easier to consume. Those are all true. I’d just like us to focus on client consumability, which is I think a much more important and valid category than the “web-based VW” category (which I think is not even really coherent).

24. Digado said on August 18th

    I still don’t have entirely clear what you mean by ‘client consumability’ - I assume it just means Virtual World accessibility (how accessible are virtual worlds?). But then you place it in the context of ‘there are no browser-based worlds’; where you seem to be more after some sort of technicality. Habbo, Webkinz, Club Pinguin, Webflock, 3DXplorer all seem perfect examples of virtual spaces within the browser, a ubiquitous tool for navigating the web…

    – Edited out 3D

25. Dale Innis said on August 19th

    Yeah, I probably mean the same thing by “consumability” that you mean by “accessibility”; the latter just has special connotations in my field (see for instance “http://www.w3.org/WAI/intro/accessibility.php”).

    My point about there being no browser-based worlds is that it’s only the *client* side of the world that’s browser-based, and since in general a world may have many different client programs (some browser-based and some not), being browser-based isn’t generally a property of a world as a whole, but rather it’s just a property of some of the ways to access that world at a given time. Once there’s a commonly-used browser plugin that allows SL to run within the browser, will SL be browser-based for that reason? Or will it still not be browser-based if most people still use an outside-the-browser client?

    To put it another way, it makes a certain amount of sense to talk about red-headed people, because that’s to a significant extent a property of the person. But it doesn’t really make sense to talk about a red-hatted person, since people can wear different hats at different times. To my mind, “browser-based VW” is more like “red-hatted person”…

    I don’t know if that makes it any clearer… :)

26. Pavig Lok said on August 19th

    I think one of the problems in this discussion is where we draw the line on virtual worlds. If we say took all the tween targeted worlds, many of which run in the browser, and re-labeled them “multiplayer social games” things would look very different.

    The tween market worlds can be very tightly functioned and focused. If we start including say Barbie World in our figures against worlds with a true business use case - then our figures will be totally skewed towards worlds with no business impact (outside their creators.) The market can only support a small number of players of this type as the model relies on huge numbers of tiny pockets, and churn. Webkinz or Barbie worlds are about a product…. whereas WoW or Second Life IS the product.

    Virtual worlds need a degree of complexity to allow comfortable interaction and hours of use. Hours of play is a different matter. For more than cursory or purpose specific adoption one must create some barriers to entry in terms of complexity, for without that complexity the world is uninteresting enough that folk won’t actually lay down roots there.

    If we are to develop virtual worlds which are as critical to our daily lives as the web and email has become, we must find ways of making them livable. Freedom of movement and action are fundamental to this. Newbies might find it hard, but the results are worth the investment. You wouldn’t want to actually live in Second Life as it currently stands; but say a virtual world became a large part of your daily routine due to business or education - Second Life would be a far more comfortable world to spend time in than webkinz or habbo or imvu. Even WoW would be more comfortable than some… and that’s saying something!

    According to the criteria of user options for action and creativity, World of Warcfaft qualifies more as a virtual world than many of the supposed VW’s that are often discussed. If we are going to disqualify WoW as a “game”, we should also disqualify many of these other “browser based worlds” which are “10x as successful as sovereign applications”. Moving the furniture does not make a virtual world rather than a game - you could do that in The Sims.

    In short… I think we’re mixing domains in this discussion.

27. Dale Innis said on August 19th

    Speak for yourself, Pavig; I would *definitely* like to live in Second Life as it now stands. :)

    I’m not completely convinced that a world complex enough to hold long-term interest is therefore necessarily harder for a beginner to use. Certainly it’s *harder* to make a world that’s both easy for a beginner and complex enough to hold the old-timer’s interest, but I don’t think that’s a fundamental limitiation. It’s just harder.

    And I’m not really sure I understand your paragraph about the Barbie Worlds of the world. Are you saying that the market can only support a small number of brand-focused worlds? I would tend to think the opposite: there could be zillions, if they’re all similar enough that the learning curve to start using a new one is pretty flat. People won’t tend to live in them for extended periods unless they’re really diehard Barbie otaku or whatever, but I can see that general niche surviving pretty far into the future with lots and lots of (perhaps ephemeral) players…

28. Pavig Lok said on August 19th

    Dale… my day job is working in Second Life, and it still has a long way to go before it’s a comfortable work environment. Even so it’s miles ahead of most of the competition in terms of being a workable virtual workplace.

    I think worlds can be scaled in complexity from a single base - so complex barriers of entry to a complex world could be shaved off the newcomer expereince to allow folk easier entry. The Mac user interface is a brilliant example of this - the mac hides most of the tricky stuff that oldbies know behind the fog of aqua which presents the things you probably want to do right in front of you. VW’s have a lot to learn from clever design like that.

    Barbie world is an actual virtual world - during the peak growth period it outpaced SL growth 10:1. My running joke with virtual world pundits who talked account growth as an indicator of the significance of a virtual world was to use reductio ad absurdum and say; by your argument in two years time we’ll be holding business meetings in Barbie world. :)

    It’s an extreme example, but i think an illuminating one. It is easy to see something with wild growth, or hype cycle attention, as having more significance in the market down the track than it might have. Barbie world is a prime example, as it is a high growth world but of course has eventual limits on what it will become - Barbie limits. Just as it would be absurd to ignore that it’s just about Barbie and start talking about it as a general purpose virtual world, we should also be careful of the inbuilt limitations of the other worlds we choose to compare.

    Some worlds will be built for simple entry, massive growth, and light engagement. Others will be built in such a way that they become a more substantial element of our digital lives, and fulfil many purposes.

    Lively is a fine example of a world which has only been active for a month but already faces problems. For widespread adoption by the business community for conferencing - something it is uniquely placed to do well (with gtalk integration already compiled in waiting for the on switch) it has run into it’s own roadblock. The Lively user interface was never designed to cater for large groups or fluid communication. For one of the main features of their strategic plan, it was built broken.

    One month after launch they are going back to the drawing board with their entire communications UI to give it the capacity and elegance to deal with real events. Rest assured Google will find a reasonable user interface design for this… but it won’t look like lively, and to have the complexity it requires to deal with the difficulties presented by large numbers of participants, it won’t be as easy to use.

    Other virtual worlds with more entrenched chat metaphors, building metaphors, etc will also need redesign as they grow out of their narrow use case roots. As they do so they’ll also find criticism from the E Z 2 UZE camp… if cleverly done that criticism will be minimal.

    But back to the Barbie thing; many of these virtual worlds are so focused on such a narrow use case that they will never jump out of their niche - it just makes no sense for them to. A narrowly focused world is only diluted by increasing the complexity of interactions possible in it - one hits diminishing returns. If such worlds are effectively hobbled, tied to their own product too tightly, is it fair to compare them to general purpose worlds which support multiple products and services? Personally I think Barbie world and Second Life are very different products, and can’t even be compared - they belong to different categories of application.

29. Dale Innis said on August 19th

    That’s fine, Pavig, I was just reporting my own subjective fact: I would like to live in SL, even if you wouldn’t. :)

    There are certainly contexts in which, as you say, it doesn’t make sense to compare the more generally targetted worlds like SL and Lively to strictly branded worlds like Barbie. On the other hand there will inevitably be some things that they can learn from each other, so I don’t think we should treat them as entirely different beasts, either. I do like your idea of using the Barbie growth-curve as a reductio, though…

30. Digado said on August 19th

    @Pavig: “In short… I think we’re mixing domains in this discussion.”

    Though its hard to pinpoint virtual worlds as a whole, i usually use it as an equivalent of virtual environment, - a computer simulated multi-user platform which offers some kind of immersion through displacement of presence. I never dismiss WoW as a virtual world but as its goal oriented (as opposed to open-end VW) every function evolves around a fictional, pre-programmed goal, very different from a virtual, social one. Habbo, and other tween wolds evolve around social interaction and have no ‘goal’ - no game to win, which is a different model for virtual worlds - the kind of model which is still in front of the ‘chasm’ instead of having crossed it like MMORPG’s have done.

    Furthermore - the barbie example not jumping out of its niche deserves better. Look at what disney does, creating a chain of worlds to introduce, persuade and capture a growing (aging) audience. Now Barbie doesn’t scale as well as Disney does, but the niche that went to barbie-world, and outgrows barbie, will look elsewhere for a similar experience (SL Gossip Girl if marketed right for instance). Especially in the adoption of virtual worlds these niche worlds have a much better shot at helping VW’s from the early adopters to mainstream than generalist worlds like SL and There.com have. Niche worlds are able to provide clear function and value for their users because they know who they are and what they like, instead of SL’s eternal mistake if trying to be everything to everyone.

31. Pavig Lok said on August 19th

    This is where i must disagree. I think there are very strict limits on what a virtual world is… namely that its a “world”. This implies that it can be trusted to act like a real place in enough ways for folk to find immersion in it. By immersion I’m not talking about what the lifers and gamers mean by it, but a sense of personal agency.

    One of big issues I see with many of the “worlds” around now is that you have space… you can have a flat… you can arrange the furniture… but beyond that you are totally restricted by what the world creators give you in the library. The only thing that differentiates this kind of world from a freeform game like the Sims is that you don’t have so many stats, though of course there’s always the metagame of “popularity” etc. These worlds can not be used for much that wasn ‘t designed into them, so avatars have very little true agency within them.

    As we are realizing in the web 2.0 area folk may use many applications, but they only inhabit the ones they can mash up as they please, where the users are free to create the use cases rather than relying on the application creator to design these options in (thus focusing the application so tightly that it’s uses are restricted).

    For me a virtual world is one in which the tools are provided for folk to invent uses for it. An avatar chat or virtual social space is one where chat is the predominant activity and content, use cases etc flow down from (or are mediated by) the world designers…. and a “virtual social game” is something like webkinz, where pre-made activities are the dominant theme. All of them have spaces and avatars, but they have very different structures and design goals.

    About a year ago Entropia was often discussed as a virtual world… it had an economy, user content, was (or could be) freefofrm. If you made a bullet list of it’s properties and compared it to second life they looked kinda similar. Entropia however had structures in place which firmly held it within the game domain. These days it’s seen less as a virtual world, and more as a mmorpg with gold farming built in.

    For me to be convinced that many of these avatar chats (which have been around for 15 years or so) have the qualities we associate with the word “world” I’d need to see a pretty coherent argument. Lively for instance is not a world at the moment so much as a building - it has rooms and places, but they’re poorly connected. All the space metaphors are broken. It looks kinda like a place but doesn’t act like one in any way you could trust your instincts. It doesn’t leverage the intuitive knowledge we have about the real world to make us comfortable in it.

    It is my belief that the general public will become comfortable and trust interactions within virtual worlds when they feel that, like the real world, they are in a solid and reactive space that obeys some basic rules with some fidelity. A simple example of this is (unlike SL) in lively someone can animate my avatar without my compliance. This breaks one of the most fundamental intuitions we have about our real selves - the sovereignty we have over our own bodies and what we choose to do with them. A virtual world should act in such a way that basic knowlege of how the world works shouldn’t be broken in such a cavalier fashion, no matter how amusing it might be.

    So that’s lively acting unlike a world and more like a game - in such a way that it breaks “immersion” and turns our avatars into detatched puppets. Conversely in second life, where avatar attention is tied to the mouse, you can read peoples body language enough to know what they’re interested in, or if theyre distracted, or a confused newbie. As subtle as it is that’s something LL got right - it’s the virtual world acting enough like a real one to provide intuitive situational awareness. If a world acts like a world we inhabit it as such without thinking about all the visual cues we may be processing.

    Anyways I’d better not rant so much :) But that’s the distinctions I make. I think avchats are a good entry into the virtual world area for newcomers, but I don’t think they’re the killer application they’re made out to be. Well not yet anyway. And I don;t think they qualify as virtual worlds particularly, though they share many qualities with them.

32. Digado said on August 19th

    Please do ‘rant’ so much - I really appreciate your input and you make an interesting argument though I feel very one-sided :) Its from the builder/expressionist point of view you argue ‘the world’ has to use the metaphor to its full extend, instead of just a virtually simulated environment.

    Furthermore, the open content creation and ‘free form what the users make of it’ world sounds more like VRML, or even Flash rather than an actual platform/application. The tools in which a 3 dimensional space can be build. And before you disagree think of the qualities on the ‘bullet list’ you mentioned. You are very selective in highlighting what makes a virtual world to you, and what doesn’t ‘deserve’ that qualification.

    Especially the part as seeing UGC as part of the primary VW toolset rather than something part of the world itself is not for everyone, in fact i think its for a very limited audience only - the 3D photoshoppers. I think of it as PHP development for the web, or posting comments/blog entries in pre-made formats (applications). I don’t want to build wordpress, or even a blog per se, as an end user I just want to write down my thoughts and bookmark some points of interests publicly - and this is where the masses are.

    Lastly, there is just as much to ‘avchats’ as there is to 3D photoshopping. They are both about connecting, expressing, socializing and communicating in a visual, immersive environment in the end. They just dont necessarily appeal to the same audience. Just like instant messaging and blogging doesn’t necessarily appeal to the same audience - different applications for the medium but I can’t really make a distinction in which is a better use of the internet - neither can i between avchats and other applications using virtual, rendered environments.

    So thats why i think they are more than an entry point, do fit the VW descriptor in the widest sense, and stand a better chance at providing applications (and therefor integration and eventually mainstream or even ubiquity) VS the UGC generalist worlds. The facts kind of speak for themselves at this point and I dont think that is about to change soon - the trend being more towards walled gardens, secure and private worlds rather then a rush into Second Life or similar approaches.

33. Dale Innis said on August 20th

    Haha, I was just having this same discussion inside our corporate firewall!

    I think that user-generated content is very important for a successful virtual world that wants to be a general-purpose virtual world rather than just a MMORPG like WoW or whatever. And it’s important for two reasons:

    First and I admit somewhat idealistically, I think that everyone wants to create, and will create if we make it easy enough and foster a culture that encourages creation, and that everyone will be richer and happier as a result. It doesn’t have to be the creation of the shapes of virtual objects (”3D phtotoshopping”); it can be that, but it can also be textures, or music, or design specifications for an object that someone else will actually build, or room layouts in a building, or text and writing or all kinds, or sound effects, or the mechanism of a quest game, or a set of jokes, or a new way to organize a committee, or a script to power a funny hat, or… But in any case empowering people to be creative in these ways in the VWs means providing User Generated Content.

    Second, and much more practically, user generated content is important because without it the owners of the platform are a bottleneck for every single stupid thing that anyone needs built. Someone wants to open a store to sell either real or virtual goods, and they want the store to have a distinctive look rather than being just Generic Store #3, they have to file a petition with the platform owners and hope someone gets to it. There’s a craze in RL Korea for skirts with huge flower-shaped ribbons, by the time the VW owners notice (unless they’re in Korea themselves) and provide the corresponding virtual content, the craze will have been over for a week. A corporation wants a conference room structured around the basic principles of their new Seven-Sigma Continuous Improvement Business Innovation for Stakeholder Success Philosophy, they aren’t going to want to queue up behind the people who are badgering the platform owners for custom houses done the Dark Elvish style.

    UGC frees the economy. We know that the way an economy produces the right goods and services is by way of a price system and a free market (modulo market failures, externalities, rights violations, and so on). We know that central planning of the means of production doesn’t work in the real world. Why should we expect it to work in the virtual worlds? Why would Linden Lab be any better at predicting what ought to be designed and created for its residents than the Supreme Soviet turned out to be for theirs? UGC puts the decision-making power out in the user community, where imho it belongs.

    And note that this isn’t just about the people who want to create stuff. Even if my idealistic idea is wrong, and some people really are born to be passive consumers who only want to buy, never create, UGC is still the right way to make sure that the stuff that they want to buy is available. A vibrant economy will do a pretty good job of making that happen; a bunch of people sitting in a room at Linden Lab trying to decide what objects to add to the world next will do a very bad job of it.

    End rant. For now. :)

34. There.com beats sluggish Linden Lab | Digado said on October 30th

    [...] fact There.com seems to be the first to pick up on the earlier discussed ‘modular virtual world’ probably comes as no surprise to the runner-up in the Second [...]

35. zhorzha said on February 12th

    what will happen is the web will get deserted once it goes onto tv sets. sl should not go on the web if it wants to serfive it should make a deal with the tv set makers to have their tv set inside sl

Join the conversaton!