Second Life Goes Open Source– First Thoughts, With Linden Lab’s Replies (Updated)


I was only hours from posting a sequel to the ongoing “Is Second Life Over-Hyped?” controversy, when Linden Lab spun the conversation in a totally different direction: starting today, the company behind the user-created world will open source the code that runs the SL viewer. Official announcement here, some out-the-gate enthusiasm from free culture advocate/Boing Boing blogger Cory Doctorow here; while you’re at it, grab your own copy of the code here.

But is this actually a good idea? Down the road, absolutely. But right now? There, I’m not so sure; actually, I worry this comes a touch too soon. I put my concerns to Linden CTO Cory Ondrejka, who managed to bounce back some brief replies during what must surely be a savagely busy day. His terse rejoinders to my rough-hewn first reactions after the break.

Before going further, by the way, I should first flash my full disclosure badge to indicate I have a vested interest in Second Life succeeding— so in this case, what follows is not just industry analysis, but personal and professional concern.

Revenge of CopyBot

As it happens, there already is an early project to develop an alternate version of the SL client, and it’s from a group of SL users known as libsecondlife. But the thing is, that very name can still elicit snarls from many in the Second Life community, because late last year, members of the team released a content replication program called CopyBot, and CopyBot provoked a brief but riotous protest that threatened to cripple the internal economy. Read my take on it here. An open source Second Life will surely enable a potential army of CopyBots, as developers (some with malice, some with good but unwise intentions) create alternate clients which, when interacting with the rest of the world, cause grief and panic through the community at large. (“We expect the improvements to features and stability to overcome the negatives,” Ondrejka tells me, in response to that.)

Dying Netscape’s Noble Death?

The official announcement begins by comparing Linden’s move to what Netscape did (motivated by “acceptance of the inevitable or simple desperation”), when they opened the source to their browser, which led to a fruition of Web development.

For the very young who are reading this, some history: Netscape, you see, was a company that once dominated the browser market. Why sonny, did you know its IPO in 1995 launched the dot com boom? Yes they did. Now, however, its share of the browser market is less than 1%. So while it’s true the source code was a marvelous gift to the web’s growth, in retrospect it seems like a heroic sacrifice— great for the Internet, but not so great for the company that did it.

“No,” Cory tells me, in reply to that. “The point is that we aren’t Netscape, we are doing this from a position of strength.”

By this, he presumably means that Second Life dominates the market for user-created, immersively 3D online worlds in which users retain their IP. Actually, they dominate because they’re the only ones in that specific market, but at least three companies/projects— Multiverse, Raph Koster’s recently announced Areae, and the non-profit Open Croquet— are gunning for entry. And by domination, we really mean Second Life currently has 200,000-230,000 active users; not insignificant, and by some projections, this number will reach 1.6 million by this April. Still, a user base in the hundreds of thousands is something a successful MMORPG can capture in a matter of months (or in World of Warcraft’s case, days). It took Linden Lab over three years to reach that figure. Going open source will spur SL’s growth in the short term, but will surely undermine a collective sense of a unified world, which may, in turn, cause growth to plateau, or even worse, fragment.

Which leads us to the next point:

Building Babel?

The great thing about open source projects is they empower brilliant, iconoclastic developers to innovate and improve. The not so great thing about them is forking, wherein those cantankerous developers go off on their own to create code variations which ultimately become untranslatable to each other. (This is at least partly why a Linux OS has been a failure at market.)

Is this where Second Life is heading? Not one SL, but a variety of Second Lives, some not open to each other? The end of the utopian dream of one metaverse under FSM, indivisible, with liberty and justice for all?

For Cory, theological questions like these will be suspended for the near term: “People won’t fork for a long time since they’ll need to stay in sync with the simulators,” he assures me. Which is also why it won’t affect the company’s revenue model: “It doesn’t impact revenues at all,” he says, “since you still need an account to log in.”

And so there it is: in the near future (i.e., at least the next few months), no immediate changes to the world, besides a renewed interest from coders eager to tinker with the way they interact with the world of Second Life.

Of course, when Ondrejka mentioned going open source only last month, it was couched in vague, “hopefully we will in 2007” terms. Either the company is accelerating their roadmap, or have been playing a canny game of underbet-bluff-show large all along. So now I wonder what other face cards they’re holding.

Because I can tell you this from personal experience: Cory Ondrejka is merciless at Texas Hold ‘Em.

Update, 1/9: On LL’s official blog, Ondrejka now has a status report on the open source initiative, expands on many of the topics mentioned above, and discusses other development issues.  Read it here.


Jon R.

When asked for comment, the Lindens replied:

“Fuck it. You’re all working on the content for us, you might as well do the client, too. Clearly, it’s not like SL players have anything better to do, and they can’t possibly do worse than us in any case.”

Gwyneth Llewelyn

Aaah but you see, what Linden Lab just did was “raise the ante” (now I understand why Cory is so good at poker ;) ). First, to compete with Linden Lab, you’d need better graphics. Hard to do in 2002/3, very easy today. Then you needed a better grid-based technology allowing far better scalability. Again, very hard to do in 2002/3, but easy today (see OpenCroquet as an alternative to centralised distribution — it works nicely on paper ;) ). Then, you need users — ok, this is where everybody has a huge advantage over Linden Lab. Also, better content-creating tools that are even easier to use than LL’s in-built ones — one could argue that something like SketchUp (not available when Linden Lab started developing Second Life) is easier for most people.

Then — a stable economy with a million of US$ of transactions between users every day. Oops. This starts to become difficult.

Next — attract dozens (nay, hundreds) of corporations willing to set up virtual presences on a contiguous metaverse (as opposed to running their own, tiny ones). Another big “oops”. Sure, there are dozens of tools around there allowing this sort of thing (Multiverse, with 6 months or so of being in beta, will “soon” have a few games out using the platform; it’s by far not the only one). But… how many corporations are really using it? How many companies the size of IBM have a thousand employees daily using a “metaverse platform” as part of their job functions? And, asking the question in a different way: how long will it take to attract them? (Even Microsoft is developing an island in Second Life these days; Apple is the only major “techie” company which is apparently absent without a good reason)

So, the last blow was making the client open source.

What this means is that any successful competitor will need to beat Linden Lab on all the above and do it during a timeframe while Second Life will continue its explosive growth. It seems less and less likely, even with huge and massive investments behind it — and a far sounder business model to support those long-term investments. After all, the post-dot-com VC firms are not so keen any more in having “very long-term” ROIs. A few are — Amazon, eBay, Google. Two of them are major investors in Linden Lab already :)

Does this mean that there is no room for “competition”? No. I could very much imagine Yahoo swallowing up IMVU, partner with Google, integrate SketchUp inside IMVU and put everything on Google Earth — and you’ll have a Metaverse in about a year or so (always in perpetual Beta :) ) sponsored by Google AdSense and Yahoo Adverts. Well, that could happen “overnight” and swallow up Second Life in a year or two if they released the client as open source, too.

Would they do it? I think not. A wise CEO at Yahoo and Google will simply wait a bit, assign a task force to play around with the now released source code for the client, and create a “Yahoo/Google SL-compatible client” instead. Imagine a client where the in-world search would actually work but to see the blindingly fast results of that search, you’d have to watch ads. LL never caught on that insane idea. Yahoo and Google might thing that targetting something for a 2.4 million account market (never mind how many are returning users…) will probably pay the costs of developing it :)


*This is at least partly why a Linux OS has been a failure at market.

What universe do you live in? Oh wait it’s the one filled with ping wangs..never mind..


Umm, I haven’t used second life for more than about 15 minutes, but as I undestand it, it’s the client, not the server that’s being open sourced. That means there will be multiple clients, but unless someone reverse-engineers the entire server, there will only be one second life universe.

What you probably will end up with is a lot of really annoying spambots who walk up to you and try to sell you viagra every 20 seconds.


Comments are closed.