Full steam ahead for new horizons!

Marooned in Tibet Every now and then, Linden Lab manages to surprise us in a positive way. After the big “splash” of Project Shining, promising all kinds of miracles — and yes, at least the HTTP Project Viewer is a small miracle in terms of texture downloading speed — we now heard the latest announcement: that Second Life® would be available on Valve’s Steam, announced on the same day that normal, specular and diffuse maps will be re-used from the code on the Exodus Viewer and implemented on the Second Life Viewer.

It’s been some busy days at Linden Lab. But interestingly enough, we’re seeing a quite different way of leadership, and this is truly encouraging (even if does not lead to immediate results!).

You might remember that we all speculated what Rod Humble, with his background in games, would do with Linden Lab and Second Life. While the immediate “gamification of Second Life” — a heresy to most veteran residents! — was a hypothesis, the rest was not very clear. We all know that Second Life is not really a “game development platform”. Not only the renderer is not up to modern engines, but programming fast action games is tremendously difficult (when not outright impossible); region space is outrageously expensive; and avatar limits prevent large-scale battles to be fought in the same region. Even if all these problems were theoretically solved, lag, the Eternal Beast of the Apocalypse, would be so overwhelming to make any “game” a ridiculous pretense (just imagine that SL’s engine would have been replaced by CryEngine 3 overnight, a thousand avatars could be on the same region, there were no memory limits on running scripts and no artificial delays on them, and whole regions would cost less U$9.95/month).

In spite of all that, what Rod and his team found out is that there are, indeed, people doing exactly those kinds of games in Second Life.

And so I guess he was intrigued. What makes those people enjoy designing quasi-real-time-FPS games in SL, even though the technology is completely against it? And we’re not talking about “oddities” — it’s not the lonely, bored gamer who twiddles and tinkers a bit in SL once in a while.

No, we’re talking about whole communities, often with thousands of members, who regularly engage in MMOGs designed and developed in Second Life. They have their own content creators and developers. There are even “standard gaming systems” — whole frameworks based on LSL scripts designed to create MMOG-like experiences in SL. They have their own websites, forums, groups, and even at least a viewer (Exodus). And all this gets often ignored by the majority of the SL resident population, who shrugs at the pathetic attempts to create something even vaguely looking like a 1990 MMOG, and considers them to be, well, oddities.

Rod isn’t doing the mistake of shrugging them away.

Instead, he found out an interesting thing about those communities. Like pretty much the remaining resident population, they have that “spark” that makes them return to SL over and over again, in spite of all its problems — or may these be called “challenges” instead?

What is that “spark”? I believe that  has figured it out on her article, “That Spark We Have“:

The love of learning is the spark that keeps people coming back to Second Life.

Tinkering with technology — or what used to be known as hacking, when that word didn’t have the pejorative meaning it has today — is part of the “love of learning”. People who love to learn are the ones that drool at a Lego set and think: “Oh, wow, what can I build with this?” These are the kind of people that in their youths would dismantle radios, driving their parents to madness, to try to assemble them again just to learn how they worked. These are the people who prefer staying at home reading a book about natural history than to go out with their friends getting a sun tan (or, if they go, they’ll bring their book with them). These are the people who learn recipes for cooking but throw them away and instead prefer to figure it out on their own. We can add up all stereotypes we wish — not all these people are “intellectuals” or have been “good students” (in the conventional sense) in their youth, or have careers connected with academics, innovation, or even technology. Still, no matter how diverse their backgrounds are, they have this “love of learning”, even if it’s in a certain very specific way.

If you love to learn — even if not in the usual, formal way in a classroom or so — then you can put up with whatever difficulties are in front of you, just to figure out what there is to learn. This is, to the best of my understanding, a relatively little explored area. We certainly know that different people react to the activity of learning in different ways. We know that a formal learning system is definitely not universally appealing — people blame the system, the books, the teachers, or even the family backgrounds of the students which work against their ability to capture just even the bare-bones of a field of study. As a result, Western countries, in a deluded result, have tried to “simplify” the learning requirements, make the learning environment simpler, and adapt the whole concept of “learning” to a variety of techniques to “stimulate” the students’ interest in acquiring knowledge.

I’m not an expert in the field of educational sciences, and so I will take furious comments from the whole community of academic researchers who have posited a lot of theories about “learning environments” when I merely point out what ought to be obvious: school systems became universal, but just a small fraction of students actually have the “love of learning” and succeed — a fraction not much unlike what we had generations ago. The difference seems huge now because, well, we have a universal school system — and so we can compare the huge difference between “good students” and the vast majority of “merely struggling ones”. Former societies were discriminatory and had far less students, most of which would never complete their studies anyway. But a few would — the Newtons and the Einsteins of the past, who remain peerless among generations of scholars, in spite of having endured the worst possible types of teaching environments, which were authoritative and despotic. They didn’t care. They had that “spark”, and someone who has it, will endure everything when the love of learning becomes its own reward.

(As a side note, it’s interesting to see how the Singaporean School System still continues to be widespread in Asia — with variants — and while it might look is too specialised, rigid, and elitist, as its critics label it, the truth is that student results are notoriously high.)

Contrast that with our current generation — from kids to adults, all seem to suffer from an inability to keep focused for more than a few minutes. Due to the way we’re constantly bombarded with external stimula, our minds barely fail to grasp one thing to start immediately on the next. We interrupt 10-minute conversations with friends to take a 5-minute phone call, but we interrupt that when we get a text message to which we immediately have to reply. People get bored with email and prefer Twitter or the Facebook/Google+ timeline to write to their friends. TL;DR became not merely an acronym: it embodies the lack of patience of the current generation with anything that is “too difficult” or “takes too long”.

Second Life residents, by contrast, are totally different. They’re not worried to spend uncountable hours clicking on things to learn how they work. Every time I find a resident with 2-3 weeks of age, I’m amazed on how much they already know — they have learned how the inventory works, they know about landmarks, they know about shops and buying boxes from vendors to unpack them, they figured out where to earn free money, they understand about friend lists and group IM chat. That means an astonishing amount of time spent figuring it all out, because it’s really so hard to do.

But all of us have “the right spirit”. We all went through this, because we’re not only unafraid of learning new things — we actually enjoy it. Even if we don’t admit it to ourselves.

Now, among the hard-core gamer community, the same also applies. The vast majority is just into it for quick excitement — few opportunities to think, lot of opportunities to react and feel strong emotions. But there is a tiny niche group who wants to figure out what’s beneath the game. These are the ones who begin by extracting 3D models out of their favourite games, or the music, or something else. Then they start tweaking the code itself. They develop “mods”. The recent fad is modding Skyrim, but modding goes back decades, at least from the ancient days of Civilization, and later with Doom, even later with Half-Life/Counter-Strike, and so on: there is a certain class of gamers enjoyed tweaking their favourite games much more than actually playing them.

And it’s precisely these kind of “gamers” (should we still call them that? Perhaps “modders” is a better name to describe their idea of entertainment) that Rod Humble is planning to attract to Second Life.

In a sense, this is addressing “a niche of a niche”. But Second Life is all about niches, and never a mainstream product. It’s true that Second Life, as a “game”, or even as a “game development platform”, has little interest — for a decade people have been complaining how SL’s features are so below the industry standards for gaming that they’re next to worthless even for casual gamers. But nevertheless SL appeals to a rather large community — counted in thousands of members — of people who do, indeed, come to SL to play MMORPGs. The apparent incongruity is simply explained that not all gamers are equal. Some, indeed, also have the spark of loving to learn — and what they learn is how to tinker with games to that they do what they wish. Put into plain English: SL is a modder’s paradise in terms of flexibility, even though it might trail behind in terms of high-end gaming features. But given a completely-closed environment where modders cannot play with, or an open one with less features but more ability to tinker with, and modders will flock to the latter. This is what happened in the past, and is happening right now with Skyrim. Sid Meier learned very quickly to release “modding tools” shortly after a new release of Civilization, because modders would immediate start to create new content for it — originally sharing it on BBSes, later on the Internet. Will Wright attempted to do the same with Spore. Most modder-friendly game developing companies release sooner or later some tools to get modders actively developing new content, new maps, sometimes even new rules, because that extends the longevity of their products — it’s a known fact.

But all those games have their limitations to the amount of modding that is possible to do. Games designed for single-play may be better suited for modding; multiplayer games are prone to cheating if too much modding is allowed, and so MMOG operators have to constantly check what their modding players are doing to make sure that the game remains fair. While crafting items is something that happens inside the game environment through tools designed to take the rules into account, modding allows — sometimes! — players to side-step the rules, and, as such, modding is frowned upon on most MMOGs.

Here is where SL has its appeal: everything is allowed (or, well, practically everything). SL has been designed from the ground up to allow its users to generate all content, so it’s a modder’s paradise: they can do whatever they wish, within what the technology allows them to do. With an extra bonus: you can navigate among different in-world MMOGs with the same character, and just adapt to whatever setting you wish to play. The same avatar can be used to play in a sci-fi universe or in a vampire MMOG. And when you’re tired of gaming, you can hang around other areas of SL and just have a chat and relax and do something completely different. That’s SL’s appeal. MMOGs certainly have their “off-world communities” — the sheer size of the World of Warcraft online community is staggering — but, well, the point is that they are off-world. You have to leave the fun of an interactive, immersive 3D environment to go back to boring 2D webpages to get in touch with your friends off-world. In SL, it all happens inside the same environment, and you can go around with your ultra-cool Vampire Overlord and all its weapons, spells, and assorted gadgets and dance in a nightclub — SL allows that very easily, and that’s unparalleled in any other environment.

Now the clever bit is the two-step strategy that Linden Lab is following.

First, of course, means applying a few finishing touches on the graphics engine. Mesh, and specially mesh applied to avatars, was a crucial part of that strategy, since modders are used to mesh and not to prim-based environments. Setting the stage for NPCs was also a requirement, and while there is a lot still left to do on the Pathfinding code, it’s a start. Finally recognising that WindLight settings out to be part of the environment was a sorely missed (and now much welcome!) feature. The use of normal, specular and diffusion maps will mean simpler meshes (for those who are used to all those techniques) and not only increase the realism, but, more importantly, reduce viewer lag by allowing meshes to become less polygon-intensive — a technique that game developers are well acquainted with, and modders are at least semi-professional or very advanced amateurs at developing content for games. They know all about those tricks; every game out there uses them.

Complementing the technology innovations — which are important in terms of visual appeal — Linden Lab also started earnestly to engage the War Against Lag. Project Shining is the first big step taken this year, in terms of an actual roadmap — how many years ago have we had the last roadmap? I think it was around 2009… — but there is quite a lot being done “under the hood”. Everybody is aware that the very peculiar way of hosting a virtual world with visual contiguity and user-generated content implies several technical challenges: with every other environment out there (yes, even Cloud Party!) relying on closely isolated areas and download-first-generate-next content (or download-everything-first, which is the way of most MMOGs), SL is completely different, and it’s very hard to achieve similar results when the viewer has absolutely no clue about what content will be changed in real time. I won’t discuss the many possible strategies that can be used for this scenario; I will just mention that most of them are theoretical and have only been demonstrated in a laboratory environment. OpenCroquet, for instance, demonstrated how a network of users could use a peer-to-peer network to keep content synchronised among all users, without relying on central servers — but that “demonstration” was possible on a Gigabit LAN. Start adding people with their ADSL links, and specially, start adding many people (where “many” is “much more than a dozen”), and the carefully controlled lab environment starts to rip at the seams. I’m not saying that SL has the best method to deal with the requirements of our virtual world. It’s just that it’s a decade-old technique which has been shown to work, even if it has a lot of problems.

So Linden Lab is mostly enhancing and improving things, and not radically redesigning the underlying architecture. There is still quite a lot to be optimistic about: after trying the HTTP Project Viewer, it’s with extreme reluctance that I fall back to the main viewer again, because it loads things so much faster — and this is just a very early demonstration of what will become widespread among other parts of the viewer in the months to come. With bits of Havok on the viewer side, and Pathfinding-enabled regions, in a not-very-distant future, the viewer will not need to ask the simulator about crucial physics simulation events — like the “simple” calculation of where the avatar’s feet should be on top of solid ground. How far the physics engine on the viewer will be able to offload some work from the simulator still remains to be seen; but I think that the goal is to have a future generation of simulators that handle nothing else but tracking avatar position. All the rest would be on the viewer and on external servers loading textures, objects, and inventory, as well as handling chat and all the remaining ancillary data — the more these things are done via HTTP, the more they can be cached at several levels, as well as pushed out into the cloud. I’m already looking forward to an updated article from Tateru Nino or Vex Streeter to deal with the extra HTTP request types that the viewers are making, and being able to cache them all locally 🙂 Extra functions on LSL like the ability to instantly teleport people elsewhere or temporary attachments, even though they can be used for completely different (like shops!). And then we’ll get the sim-side avatar baking I’m sure we will see in the next few months which should finally eliminate half-baked, half-unrezzed avatars with weird texture cycling on them, as well, of course, making impossible to CopyBot the individual items being worn (at least those on the texture level).

As Nica Pennell so well put it:

Oh, man. For years I considered SL’s technological evolution hopelessly stagnant, and now all of a sudden they’re checking off some of the top items on my wishlist. Could user-defined avatar skeletons be next? Improved terrain mesh texturing? Who knows! I’m actually feeling optimistic! — Nica Pennell, commenting on New World Notes

The second step is something quite unusual for Linden Lab: marketing!

We all saw the media going mad with Cloud Party when they announced that, thanks to their registration integrated with Facebook, it was now a VW available to 800 million people. That meme was repeated so often that people actually started to believe it. But Cloud Party is not going to be a mainstream product for 800 million people — probably not even for a million. Nevertheless, this lead bloggers and opinion makers to ask Linden Lab to be more careful about the competition and think about what that means — in terms of media splash, announcing to the world-at-large that “800 million Facebook users are now able to use Cloud Party by merely clicking a button” had a large impact.

Linden Lab’s reply is much more cleverer. Finally understanding how Second Life only appeals to a niche market — the ones with the love to learn — they announced integration with Steam (and not Facebook!). This sounded strange to most of the resident population, for which Second Life is clearly not a “game”, and distributing Second Life via a “gaming platform” was odd.

But actually, it’s not. Let’s consider some essential points:

  • First and foremost, Steam is a community of gamers. 40 million of them. Unlike the 800 million Facebook users, most Steam gamers are willing to pay to get entertained. This is quite different from the 800 million Facebook users who use it because it’s free. See my comments on why that business model is flawed. So Cloud Party is targeting free users but Linden Lab prefers paying users.
  • Also note that while gaming on Facebook is declining, gaming on specialised platforms is not. Big Fish Games, Steam, Garage Games continue to expand and grow. Why? Because instead of positioning themselves as “mainstream” game distribution platforms, they prefer to deal with gamers as their market. Paying gamers, that is. The market is far smaller in numbers — but so what? What matters is what a company earns at the end of the day. Remember: supporting free users is a drain on resources. Having paying customers is what makes a company be both long-lived and profitable. This should be obvious, but even past the dot-com era, the mainstream media has a difficulty understanding the difference…
  • Steam users are used to downloads. Sure, downloading from Steam is simple and straightforward. So is downloading from Apple’s App Store, Google Play, Windows Store, and similar sources. While embedding everything on the viewer is great for some kinds of applications, others really are far better if they’re installed natively on the operating system — and all those mechanisms to get applications as easily on one’s computer as on a web browser are paving the way for the future. Goodbye complicated downloads and installation packages; hello, point, click, and run. So to answer to Cloud Party’s “click here to join our VW”, Second Life will be also “click here and join”. And 40 million Steam users know exactly how it works.
  • Steam is not just about games. They are also opening up their new iteration of application download which will feature social tools. This is exactly where Second Life may fit best.
  • Steam is also about community. But instead of the mainstream usage of Facebook — which, uh, is… dating? Well, just remember what Facebook was developed for! — Steam’s community is about groups of friends who love games. So everything on Steam is designed to get users find their friends, add them, see what they’re playing, offer them gifts, discuss on common forums/groups, see what achievements they have unlocked, share snapshots and videos, and, well, compete and cooperate… the kind of thing you expect from a community of gamers. What does Facebook offer gamers? The ability to pester your non-gaming friends with stupid requests for things you don’t want to waste your time with. So clearly Steam’s community is more valuable to Linden Lab than Facebook’s 800 million users — most of them non-gamers — are to Cloud Party. This is rather more important than it seems.
  • Instead of offering Facebook registration, Linden Lab will have to offer Steam registration. This might actually be interesting to see. Will they abandon my.secondlife.com and instead port all tools to be flawlessly integrated with Steam? Will users registered on Steam be immediately available under Second Life? If they can pull it off — the ability to send IMs out of SL to Steam users, for example — then this would effectively mean that Second Life would go from 24 million or so registered users (but most of them inactive) to 60 millions users (most of them on Steam but who never logged in to SL). From purely a marketing perspective, they can say “Linden Lab  grows three times the number of users able to immediately connect to Second Life” just by joining Steam. It’ll be quite interesting to see how this will be announced, once the integration is finished.
  • I have already seen comments claiming that it’s preposterous to believe that “40 million Steam users would immediately join Second Life” — indeed, it’s as preposterous as claiming that Cloud Party is “open” to 800 million potential users. In theory, anyone in the world could claim that their product or service is now available to 7 billion human beings — which is theoretically true, but not quite right. So, at best, what Linden Lab is aiming is at the small social community on Steam and the slightly larger (but highly interesting!) community of modders who have never given a thought about Second Life (probably due to prejudice after having a bad experience in the past, and equating Second Life with a hopelessly primitive graphics engine [pun intended!] not worth tinkering with). The important thing for LL is that all these people are potential customers. Sure, a lot will log in for free; but it’s the ones willing to pay for entertainment that are LL’s biggest catch. I think it’s possible to get a million new paying customers in a year or so, if LL handles the marketing well. No, not all of them will buy islands hehe — most will just expend a few Steam Credits now and then, and these will probably become “legal tender” at the LindeX
  • It paves the way for Linden Lab to add their “other products” to Steam as well. We have no idea what those “other products” are supposed to be, when they’re going to be finished, how much resources LL is really investing into them, etc. Looking at the amount of coding that is being done on Second Life, it seems actually strange how they still have time left to develop “other products”. The whole Web services team — which is not that small — is very likely busy with Steam integration, and that will mean little time left for anything else. Unless LL sneaked in some new developers, or is co-opting independent teams under their brand (which is what game development houses do when they reach a certain size), I don’t see where that extra labour is coming from… but we have to assume that Rod was being honest and that they are really planning something. Whatever it might be, it’s sure to pop up on Steam sometime.
  • And finally, of course, this will get Linden Lab’s announcements on Steam to be picked up by the gaming media. Even if they hate Second Life, they will not be able to ignore it, as they have been doing so since 2007. The worst-case scenario is how they will negatively review SL’s engine and features compared to modern games and platforms, but, as the saying goes, it’s far better to talk negatively about a product than not to talk at all! There will always be people willing to try SL out just to see how “bad” it is, and some of those will be… modders. Who will become instantly hooked. On the other hand, LL can rely on a huge base of fans who will quickly “set things right” on an exceptionally bad review 😉

When all the above is put together… this is big. Sure, it’s not “mainstream big”. It’s big inside a niche of a niche. But it’s far better than years of silence. Linden Lab’s press releases are frankly boring, and they didn’t send out a single press release in 2011. This might begin to change as soon as Steam integration is announced.

Way to go, LL!

Oh, and while LL is in the mood, I’d add Second Life to both the App Store and the Windows Store. It cannot hurt. Right now, a free text search for “Second Life” on the App Store returns an application called… BibleReader. Eek! lol

Now one can only wonder if Linden Lab will drop their own rendering engine and adopt Valve’s Source Engine instead 🙂

As to the picture above: this is an example of what Second Life can look like today on 6-year-old battered hardware. I cannot get much better results on modern “games”. The difference is that on the above scene I was getting 3-5 FPS, while I’d probably get 20 or so on a modern rendering engine. Or perhaps not.

Print Friendly, PDF & Email