It’s that fun time of the year again, when one looks up what predictions for last year actually came to be, and what the new year might us bring. Let’s take a look at what I have predicted last year and see how good those predictions were!
I had cheated a lot last year, since some of the “predictions” were really just a matter of time until they were released, and not really “predictions” in the sense commonly used — predicting what the future might bring. In other words, I deliberately mixed up “planning” the future with “predicting” the future, and that’s why in general I didn’t fail on every prediction!
- “Second Life 2.0?, LL’s newest viewer, will be released. Indeed, that one was rather easy to predict. What would be harder to predict was its level of acceptance. We’re at 2.4 now, and although I have noticed a recent increase in the number of users (suddenly, everybody around me seems to be able to “see” Display Names, so that shows a lot of people are using 2.4), I have no idea on how widespread the acceptance is. Back in 2009, I assumed that 2.0 was a “mandatory” upgrade and would be used by everyone, which did not happen.
- XStreetSL will be rebranded. Cheating again, I knew that LL would abandon that brand, which is awkward to spell in casual conversation and made little sense except for die-hard SL residents who have been following the history of web-based shopping since October 2004.
- Interop will become a reality. Totally failed this one. Not only LL broke up with IBM and Intel before the IETF launched the first draft of VWRAP, and thus the announcement of that draft was pretty much ignored, but LL dropped all pretense of continuing to develop any kind of interop with other grids or other VW vendors.
- VastPark and other similar virtual world platforms will join the OpenSimulator Foundation. Hmm. No 🙂 I guess I was too optimistic about that. 2010 showed the rise of Unity 3D as one popular framework to develop new virtual worlds, and instead of consolidation, we see more and more fragmentation. We reached the 1 billion users of VWs in October 2010, but most of those users are not consolidating on a single platform (See KZero’s charts here and here).
- Linden Lab will experiment with new avatars but not release them yet. I could cheat on this one and say that the new meshes will allow new avatars, with shapes and clothes like we have never seen before. Linden Lab is still not allowing new skeletons or new avatar meshes, but through the new way meshes will be able to get attached to avatars, this might amount to the same thing. So, well, I’d say that the prediction is half-right; and no, LL has not officially released this in 2010, even though we can try it out on the Preview Grid.
- The “Account Dashboard” will become more and more Facebook-y. Sort of! One year was not enough to see much advance in this area, but clearly it’s part of LL’s plans.
- The European Grid opens. Nope. I didn’t predict LL’s “downsizing” which shrunk the company by a third of its former size, and one of the things to go was grid hosting on an European co-location facility.
- Validation/registration will slowly become more widespread than anonymity. I don’t know if that’s really true in SL. Sure, people can now use their real names as Display Names, but how do you know that these are, in fact, their real names? In fact, Second Life seems to be the last refuge for people with security and anonymity concerns. Since businesses and academics were kicked out of the grid, two of the strongest drivers towards “real identity” were removed. Now the last drive is just for people looking for dates with real people (as opposed to merely having fun with anonymous persons). Since that accounts for a relatively small audience (let’s say, 20% of all residents), the push towards more validation or more registration needs seem to have abated somewhat.
- Raph Koster will give up his company and start developing virtual worlds in OpenSim. Mmmh, no 🙂 If at all, he’s worrying about how to integrate virtual worlds in Facebook.
- Philip Rosedale will finally assume that his new company is really just a front to look busy, but in reality he’s back in SL doing what he likes. Amazingly, in 2010 Philip came back as Interim CEO for Linden Lab and went away before the year ended 🙂 But he managed to shake things up. So, mmh, I guess I was lucky with that one.
Work at grid interoperability has not stopped at the standardisation level, and the next milestone is due in 2012. In the mean time, OpenSim grid operators have isolated themselves more and more, although Hypergrid teleporting v. 1.5 was released and showed great promises of seamless intergrid teleporting with all inventory and respecting creator tags and item properties and attributes.
Interop, these days, is just pretty much web-based shopping across grids (including Second Life). This is interesting in itself, because the biggest hurdle for universal acceptance of OpenSim grids seemed to be content transfer and safe commerce. Web-based shopping allows that using a completely different model which is not based on “technology”, but merely by relying on content creators to port their content to other grids while making a profit that way.
All in all, I managed to be correct on perhaps half of the items, which is rather bad — I’m afraid I’ve got no prediction powers!
Let’s see what I predict for 2011:
- Dramatic land price changes as LL moves to new revenue model. The reasoning behind this is not very easy to follow, and it requires to postulate that Linden Lab understands a bit about economy (who knows, Humble might be good at it). As a true child of the dot-com era, LL “assumed” that SL would have “exponential growth”, and what we saw in 2006/7 really looked like that. Under an exponential growth model, the biggest issue to deal with is infrastructure. The rest is easy, since you’ll get more and more money, and more and more users every month, so that even if you lose a large chunk of users from one month to the next, the subsequent month(s) will bring as many users as all months added up before that. Exponential growth systems allow companies to be sloppy — they can anger their customers as much as they can, because all they need to do is to wait a few more months until they have the same amount of brand new users as before.Now unless LL is seriously deluded, they will have noticed by now that we are at a close-to-stagnation phase, i.e. the market niche is filled and will very likely not grow further. Even if it grows, it won’t be at the same exponential rate as before, no matter what LL does (see below!). Under this economic model, the only way that LL makes a return above the population growth rate is to twiddle with the elasticity of the prices. Put in plain English, reduce the prices, and you’ll get more users buying land; do it right, and you’ll get a bigger return. How does this work? I have lost count recently, but let’s assume that there are 30,000 residents willing to pay US$295/month for a full sim right now. Now imagine that sims start costing merely US$30/month (same number of prims, same characteristics): if LL gets more than 300,000 residents buying sims, then the strategy has worked! Well, you’d have to take into account the increased cost of supporting ten times as many sims, of course. But technology can help: get more sims per server, thus keeping the cost per sim lower!
The problem is how to reduce the cost per sim to as little as, say, US$10 or 20 per month. Right now this seems impossible. However, a possibility actually exists: just use OpenSim 😉 This would be a dramatic change for LL — they could become their own OpenSim grid operator, selling ultracheap regions in OS, but fully interconnected with their asset servers. So instead of paying US$30 for an OpenSim region from an independent grid operator, people would pay the same amount for the same kind of service — but have it fully integrated with LL’s asset servers.
A “normal” company would never do this, of course, because that would spoil their brand — effectively competing with itself using a different product. Just imagine that Microsoft would release Microsoft Linux, using exactly the same UI as Windows 7, but under a free licensing model (or an insanely cheap one), to prevent their users to switch over to Red Hat Linux, Novell’s Suse, or, well, Ubuntu. This would never work, it would destroy all remaining credibility of Microsoft in supporting their own product for a higher price.
Then again, LL is hardly a “normal” company… so they might figure out a different way to offer supercheap sims and actually make more profits that way.
- Linden Lab will acquire Tipodean Technologies. While still tinkering around with other ways to deliver SL via a Web browser, LL will not pass the opportunity to get a Unity 3D viewer able to display SL content. In the so-called “serious” VW market (business & academic), Unity 3D has established themselves as a serious and more appealing alternative to Second Life, so the ability to say that a customer can have both — Unity 3D viewer, SL environment — is too strong an argument to ignore.
- Second Life 3.0 Viewer gets released. The new viewer will have as a major feature a complete detachment of the UI from the rendering engine, and include third-party plugins. By default, residents will be able to pick a 1.X look or a 2.X look, thus ending, once and for all, the long drama surrounding the UI.
- Meshes to revamp the economy. Like any other technology introduced in SL (e.g. flexies, sculpties…), content creators will create fantastic and amazing avatar attachments using meshes, and this will boost the sales of major fashion designers, as fashion-conscious residents will discard their old hair and clothes and buy new meshed ones. As a nice side-effect, overall client-side lag will be much reduced.
- Second Life will incorporate Google +1 technology. Maybe LL doesn’t need to worry so much about being “more like Facebook”. Instead, we’ll be able to use Google’s upcoming social sharing technology from inside SL. Watch your timelines as messages start to appear saying: “JohnDoe Resident shared with you a new landmark”. Properties for objects will start to feature the number of “likes” they got from fellow residents, and all this will allow clever programmers to elaborate lists of “most liked” event/fashion designer/location and publish them online, thanks to Google’s upcoming APIs.
- The 500 kids in SL will be completely forgotten. By the end of 2011, few people will remember that the Teen Grid has been shut down and merged into the Adult Grid. Teens are having fun on myriads of other virtual worlds and won’t bother to be in user-generated content virtual worlds like Second Life. Except, of course, for a few outstanding, brilliant minds, but they will be so few that nobody will remember they’re around.
- New APIs will make programmers happier. OpenSim leads the way in terms of interoperability at the programming level, since it’s so modular; developers can get a lot of statistics and functionality just by calling APIs, and not write a single laggy line of LSL. Linden Lab has already started experimenting with similar things (e.g. SimConsole). With the move of presence/chat towards Jabber/XMPP, things like chatting between SL users and Google Talk might be as easy as adding them on your list; but a lot more might be revealed, like a NPC interface for Second Life which doesn’t require a ‘bot to be logged in. I was particularly impressed with the easy way programmers can now use their favourite development editor or IDE to automatically edit scripts in 2.4, showing that some cool magic is happening beneath the viewer code; I’m expecting the ability to sort through your inventory using your operating system’s directory features to become a reality with, say, SL Viewer 3.0. And what about using L$ in payment gateways? Mmh 🙂
- Humble will try to attract universities back to Second Life. After the recent purge, LL will give academics new incentives to come back to SL, giving them incentives like never before — say, the first sim for free for 36 months if they commit to buy a second sim after 12 months, or increased discounts with each subsequent sim bought, to allow for large virtual campuses to be created in SL cheaply. Reception will be lukewarm as researchers figure out that they can import meshes on OpenSim and disregard the free content in SL anyway — the biggest incentive they had to stay in SL.
- Linden Lab will outsource Search and advert/classifieds functionality. After the recurring fiascos in trying to “fix” search and classifieds, they will finally give up and outsource it to a third-party for a recurring fee. In economic terms, it will make much more sense that way. Since the search/classifieds in the 2.X series is just basically a web page, this can be easily done, so long as LL gives access to external parties to an API that allows 3D content indexing. That way, LL gets a regular, recurrent stream of income, leaves all the risk to the third party, and pushes that “problem” outside their sphere of influence, letting others to take the blame when things don’t work as expected. After all, LL is not Google, and is not in the search business.
- Validated Premium residents will be able to save and restore content in their inventories and/or sims where they own all content. All TPVs allow saving/restoring content, so it seems rather pointless that LL’s viewer is the only one not permitting this. So LL will finally revert their policy and include that functionality on all their viewers. Estate Owners will also get a new tool to allow them to save a whole sim to disk — provided they own all content. Drama will spread on the SLogosphere as usual, but the point is that everybody already has these tools if they wish to use them, just not integrated in the “official” viewer.
That’s it for now 🙂 A very happy 2011 to you all!