The Wisdom of Pavig Lok

Learning from Pavig Lok

We’re now on the aftermath of the OpenSpace drama, as thousands of owners of OpenSpace sims, after weeping and moaning, are dropping their sims or consolidate them into “regular” sims. My friends report that almost a thousand sims were immediately discarded after the announcement (there is no market value for them), but I have no clue if that’s true: we’ll have to patiently wait for LL’s “official statistics” to see the results (most of these “sim counters” out there don’t take into account non-visible land, of course).

I think that dropping so many sims was a bit premature, though — after all, people will enjoy the same price until the end of the year, and OpenSpace owners could simply drop it at the end of December (timing it right so that they wouldn’t pay more monthly fees). But, alas, like the worldwide financial crisis, it’s the perception of a crisis that counts…

As usual, Linden Lab is counting on residents to calm down, bite the bullet, and focus on other things. They know that Second Life® residents have short memories: once a few thousands have left SL, nobody will remember any more what “OpenSpace simulators” are. As a proof, take a look at the vote count at the JIRA: after a huge exponential growth in votes in the first few days, it just got some extra hundred votes in the last week. People are forgetting.

In the mean time, I was looking for an answer for my own question on the last article on the subject — the question of why Linden Lab suddenly dropped the OpenSpace product and replaced it with something pretty useless. And by chance I have met Pavig Lok; we’re often together at the Thinker’s meetings (every Tuesday at 3 PM, in case you’re interested — it’s announced on the Linden Event list) and this was an opportunity to learn about his ideas.

The Three Types of Geeks

Let’s be clear here: Pavig Lok is one of the wisest persons I ever met in SL. Sadly, he hasn’t felt the urge to continue his well-thought blog for a while (“writer’s block”) but he’s always delightful to listen to. On this particular day he was quite inspired and very willing to share a bit of his vast wisdom to a tiny audience of two (LittleToe Bartlett being the other one!).

His first evaluation was discussing the changing population of Second Life. One thing that becomes clearer to me as I get older (in SL 🙂 ) is that “Second Life is not for the masses”. We’re sort of an elite of very special human beings, the ones that “get” SL, and we’re different because of it. LL’s purpose as established on their mission (“to improve the human condition through the use of virtual worlds”) will probably not work well — because SL is not a mass-market product. When you read people on Slashdot — many of which are expert computer programmers and system administrators — totally bashing SL, year after year, you start to think about the why. Why is a normal, intelligent, open-minded person so against SL? I asked this quite often when seeing my closest friends and former business associates to have such an aggressive anti-VW stance. But we worked together; we “understood” the Internet; we “got” the World-Wide Web. We shared the same ideals; we promoted open source software together; we were all for communities, social web sites, and reaching out for an audience where race, creed, nation, age, or gender were irrelevant; in many cases, we dreamed the same dreams and shared ideologies and beliefs. But I stood alone when embracing virtual worlds; they remained on the scornful side, laughing at me. Why?

I had no answer, but Pavig Lok has a very good one.

First and foremost, Second Life is, ultimately, GeekWorld — we might claim otherwise (I, for instance, am too old to be called a “geek” 🙂 ), but what we “special” human beings (a.k.a. “residents”) share is a certain amount of geekishness. It might not be too much. For some, it just means replacing Internet Explorer with Firefox on their desktops. Others use old CD cases to hold their jewelry but would never look at their faces on a mirror and admit they’re geeks too. Not all geeks are white male late-teenagers/young adults with glasses, living in basements and attics, getting sun tans from the glow of old CRTs. In fact, a certain geekishness is even a bit “fashionable” these days. You can fancy Prada or D&G and still be geek enough to text to your sweetheart. You might just watch soap operas with your friends, but when they go away, you secretely turn your Wii on — even if it’s just for the aerobics software.

But just the slightest touch of geekishness, even if surpressed, makes us log in to SL. As opposed to a “mainstream” product. And that’s why SL ultimately can only grow to fill a niche of “slightly geeky people” and not expand much further.

Having a product for a niche is not bad. After all, that’s what Apple (or Harley-Davidson) does. They still are profitable companies — and they make it quite clear that their products are “only for an elite”. An elite that buys the brands’ message and feels a certain belonging to the group of people that uses that brand, and behave accordingly (PC users refer to their computers as “their desktops” or “laptops”; Apple fans have Macs, MacBooks, iMacs, not “computers” or “machines”. It would just feel weird calling a Mac Pro a “desktop computer”. It’s a Mac Pro!). I refer you to Douglas Atkin’s The Culting of Brands, which was for me an eye-opener explaining how successful a certain brand can become if it appeals to a certain elite status of its users, which feel “different” from the mainstream, and how companies can so successful exploit niche markets without having a mainstream product. (Robin, if you’re reading this, pick that book!)

Now, the successful cases have, indeed, found their market. And this mostly means: addressing a product to a niche of people who are willing to pay (premium) for a service or product that makes them feel different from the rest of the masses.

Is Second Life such a product? Assuming that its niche market are geeks — computer geeks, designer geeks, architecture geeks, fashion geeks, social geeks, music geeks, you name it — the question is: are they willing to pay for this product, and, if so, how much?

Under Pavig’s model, geeks come in three flavours. I can’t remember the actual names (and my apologies for Pavig for forgetting so much so easily…), but the first type is the libertarian geek. These are the wild Californian types that embrace any kind of computer technology that will allow the world to become a better place (notice LL’s mission…). They obviously will flock to SL like moths are attracted to a flame — this is the place to be for libertarians, in a world they build with their own hands, with few laws, a lot of freedom (to create, to express yourself), and “a dream come true”. SL is the libertarian utopia, specially so when the client was released as open source, and there are open source servers available too. Nothing could be more perfect than SL — and that’s why these types are faithful until the bitter end and will never exchange SL for anything else, because there is nothing else like SL.

Unfortunately, these are also the type of geeks that are not willing to pay for such a service. They’ll eagerly become Basic accounts and use as many freebies as they can, and, of course, also create their own freebies to share. Sharing — not corporate economics — is one of their drives. SL is the ultimate sharing platform, so they live in it like fish in water. But they couldn’t care less about economics. If LL fails, that’s too bad, they’ll switch over to OpenSim-based grids and continue the sharing there.

The second type is the capitalistic geek. They certainly “get” SL, and totally understand the motives behind its existence, as well as being very good evangelists to promote the use of virtual worlds to their corporate bosses and clients. Unfortunately for LL, these are the types that will almost always just sell their bosses a different virtual world — one that is closed, proprietary, and far cheaper (and yes, it might mean Google Lively, or OpenSim). This basically means that LL won’t get a huge income from this group, since most will quickly leave and bring their business elsewhere.

And the third type is the clueless geek. They will come to SL because it’s the place to be. They’ll roam the world in search of that special thing that SL is supposed to be, but can’t find it. Nevertheless, they’ll still be good residents, always searching (in vain) what makes SL so special.

Unfortunately, while they search, they’re not going to spend much, until they’re sure they find out why SL is so special, which will be — never.

So this pretty much explains why SL can have 15.5 million registered users but only a handful — less than 1% — willing to pay for it. As a business model, it’s terrible. Searching for a “market” among a certain class of users (geeks) that is mostly unwilling to pay for it is not very scalable, unless, of course, LL figures out a way to extract from those 1% enough income to pay for the 15.5 million to have fun.

Which is exactly what LL does, of course.

And this is ultimately the reason why nobody else is using this business model. It’s too risky. For the “competition”, a less risky model is not making a simple “copy” of SL, but something entirely different.

Lucky us.

The Uniqueness of Second Life

So, well, what is so different about Second Life?

Pavig goes further to explain what’s behind the devious minds of the oldest Linden Lab employees — their founders and first architects. They found a few issues that are unique to Second Life and will not ever let go of them, even if that means their downfall.

The first one, of course, is user-created content, and I don’t need to repeat why it’s so special, or what the pitfalls of running a virtual world with user-created content are — these days, after Lessig made the case for user-created (and user-owned!) content, we all know about it, but we just don’t give so much importance to how it uniquely shapes Second Life and makes it totally different from anything else out there.

Granted, there are naturally many platforms allowing user-generated content (Multiverse comes to mind, but any toolset to create MMOGs/MMORPGs will allow that), but Second Life is quite different from those.

Pavig points out two things that are so often overlooked: persistence and contiguity.

Persistence is something so obvious that we totally forget about it. If I have permissions to drop a prim on a parcel, it stays there. For a long time. Measured in years. Now this is far less “obvious” than it seems: almost no other virtual world works like that. Croquet, for instance, does not have “persistence” built-in: content exists only as long as computers are on the same peer-to-peer network — unless, of course, that some of those computers are turned on permanently (effectively replicating LL’s client-server model and breaking true p2p cloud computing). Google Lively “rooms” naturally will keep content on a user’s own room for a long time, until, of course, the user cancels their account. Then their content is gone. This is obvious, after all — but Second Life doesn’t work that way. You can still get some cool 2002 objects — created by Philip Linden for instance — even if their owners have long departed SL or are not actively doing any new content, or owning land, or even logging in to SL. We can enjoy ancient buildings like the Governor’s Mansion, the Beanstalk, or the strange silver statue that is somewhere in the grid and which allegedly is the oldest item ever built. We can even have access to one of the very first slideshow presentations done in SL by Philip himself (some clever residents found their UUIDs and posted them on the forums), probably long after Philip forgot all about it. So, in a sense, content is independent of their creator’s continued (virtual) existence. That means that once a prim is dropped on a bit of land, it stays there forever (in practice, of course, the land beneath might change ownership; or the asset server might have a hiccup and delete the plywood cube; or someone might, by mistake, turn autoreturn on. But these are exceptional cases to the rule that “content is eternal”).

The other thing which was obvious in the past and that we tend to forget these days is land contiguity. Now, these days, there are many more private islands and mini-continents than mainland, but… at some point in time, that was not so. One of the key selling points of Second Life, very early on, was that it was a single world, growing endlessly, but all tied together — which was an alternative to the “sharded” approach of several MMOGs/MMORPGs (each set of users share a copy of all content, but they are not in the “same world”, even if they can, say, send IMs to players on other servers), or the “room” approach of more recent virtual worlds. IMVU, Lively,, Kaneva, Sony Home, etc. don’t have the notion of a “single world”, but rather users have either individual rooms and jump/teleport between them, or the “world” has several interconnected areas that you can join jumping through “portals” or similar tricks. Active Worlds also works like that, and so does MOOVE and

Second Life, however, is a single world where everybody is in the same, contiguous place. Granted, private islands and direct teleport have shattered the illusion of the contiguous world, but it’s still one grid, one world. The Open Grid Protocol will allow us (not earlier than 2010 though) to jump across grids, but… and here is the trick… if the developers can figure out a safe way to bring your inventory with you, this would ultimately mean that all these grids, put together, would be on a single, unified, distributed, but contiguous world. Already this means that the OpenSim grids interconnected to LL’s own grid have to arrange an area on a meta-grid so that individual sims will not overlap. The notion of many grids, one world is already starting to form in our minds.

Now Pavig claims that these three things together — user-created content, persistence of content, and contiguity of landscape — are so unique to Second Life that they cannot be “replicated”, or “copied”, or “imitated”. Forget inter-MMOG teleport — that will never happen to satisfaction. Ultimately what might happen is that some MMOG vendors, intrigued by the notion of “joining their worlds” into a metagrid, will adopt the Open Grid Protocol specifications to allow their “world” to be a spot in this metagrid. Effectively, though, this would mean that they’d be a part of Second Life as well. Or, if you wish, the “Second Life of The Future” will be a mix and match of several different technologies that just happen to allow contiguity on a single “virtual world”. But this is already happening, when we teleport between LL’s grid (which has their own server/sim technology) and OpenSim (which is an independently developed technology that has little resemblance to LL’s own servers, except that you can use the same SL client to connect to it). So this would mean that one day users might have their clients “speak” the SL and the Open Grid Protocols, and then you could jump between and SL — with a lot of limitations, of course, but it might work.

You can see my skepticism about “other worlds” joining up on the Open (Meta) Grid. The problem is simply that none of them were designed for contiguity, except, of course, to provide an illusory contiguity (ie. a single map, although when you jump to other areas in the “same” map you’re actually connecting to something completely different which just happens to look the same). Second Life, however, with all the faults of its “tiled approach” (namely, the lack of scalabilty — but we have 3Di coming to the rescue with some nifty tricks), is the only one that was designed from scratch (even if badly!) to support contiguity.

The Second Rise of The Mainland

So what does all this have to do with Linden Lab dropping OpenSpace sims out of their portfolio? The major reason is very likely what M Linden reported, and it’s purely a business decision: the cost of maintenance of those OpenSpace sims was cutting deeply into LL’s tight margins, and the reason being that although from a hardware perspective things are pretty much the same, from a networking perspective they’re not — more interconnected sims require more bandwidth, but the progression is not linear, but polynomial, as each new sim needs at least to be connected to their 4 neighbours (some claim that each sim is directly connected to 8 neighbours, although technically that might not be necessary). My maths are rusty, but this mostly means that twice the sims require 16 times the bandwidth to connect among themselves (16 is 2 to the power of 4). Please correct me if I’m wrong on the comments — but in any case, Gwyn’s maths can be wrong, but the result is the same: too many OpenSpace sims will consume the same bandwidth as regular sims (and all sorts of resources to keep the networking pipes flowing, e.g. memory buffers at all stages), but they will earn LL just a fourth of their income. So they were slowly coming to a huge financial problem, as the income for the OpenSpace sims was not enough to pay for the bandwidth costs. Why exactly they couldn’t foresee this in March this year, when they announced the “improved” OpenSims, is beyond my abilities to explain.

But OpenSpace sims are (or were!) just one of the problems. The other one is that too many private islands, outside of the mainland, would destroy the sense of contiguity. Even if technically it’s there, down below, at the conceptual stage, residents, in effect, would just teleport to each other’s private islands and totally ignore the spaces, distances, and sims in-between. This effectively means that for all purposes a private island is just “a large room”, but it makes comparing Google Lively with Second Life so much easier. If the vast majority of SL residents totally ignore the existence of the mainland — still the largest contiguous area in SL — the vast majority of SL residents is basically “forgetting” that contiguity is one of the key elements in SL. And once that’s gone, moving to a different virtual world — one that does not have contiguity — is much easier.

For philosophical reasons, Linden Lab cannot crush the contiguity paradigm and adopt the one that their competitors prefer — sharded worlds, or isolated rooms. If you start chopping at the key differentiating factors of SL, it becomes “just another virtual world”. In effect, plugging the analogue hole — that is, getting rid of CopyBot — is easily done by forbidding user-generated content, and that could be another serious blow to SL. Remove that as well, and people will possibly start demanding peer-to-peer grids next (because one can assume that at least you’d be allowed to generate your own content on your own computer), thus destroying persistence.

And this would simply mean that SL, as we know it, would simply disappear. It would be a poor man’s version of Sony Home, with ugly graphics and low FPS. While right now we endure the pain of low performance because of those three unique factors.

So here is the big dilemma:

Second Life, because of its unique aspects of user-generated content, persistence of content, and contiguity of the virtual space, attracts a niche market of geeks that find those three aspects immensely appealing, albeit just a tiny minority of them are willing to pay for all the others to enjoy a common virtual world.

But if you start chopping at those unique aspects, the geeks will leave — there will be more interesting alternatives — and LL loses their tiny paying customer base.

So… while persistence is probably guaranteed, and the lack of enthusiasm with which LL enforces content theft shows that they still believe in “User-Generated Content Über Alles”, losing too many customers to private islands is a Bad Idea, even if it’s (short-term) financially solid.

The question then is simple: how do you get people back to the mainland?

Consider the usual problems: the first one, it’s ugly. It’s pure anarchist urban chaos. The solution? Get some volunteers to make it look nicer, and launch new “planned communities”. We saw the big effort to create the nice new roads by the Public Works team. And then LL launched Bay Area and more recently Nautilus (a boon to land barons, who have finally managed to raise prices to stratospheric values on that continent), coming back into the content-creation business, and who knows, even community management. Honestly, Nautilus is not that good — it still looks like LL’s content in 2004, which was pretty good for those times, but residents have honed their skills and can do much better. Still, it’s a start. Is it successful?

Well, not totally. Because the second problem are, of course, griefers — or, at the very least, “lack of privacy”. Whereas on private islands you have tools at your disposal to deal efficiently with that. And finally, unlike what LL hoped to achieve, communities in SL do not really “grow organically” from scratch — some do, but they take years. No, the best communities are planned and are brought to life by a strong charismatic leader who attracts people, makes them talk to each other, manages the land, keeps griefers away, maintains urban chaos at a minimum, gives good technical support, and manages events to keep people interested and happy about living in that space.

To efficiently manage all that, LL has given private island owners a fantastic advantage: Estate Tools. And since you can now buy land anywhere on private islands “like on the mainland”, there are no differences between ownership on the mainland or on private islands.

Well. There are.

The most important one is that an island owner can always reclaim land — and thus, “ownership” is somewhat “shared” between the parcel owner (on a private island) and the Estate Owner. Ultimately, private islands are glorified rental systems which give the parcel owners more tools than on the mainland — but where the private island owner even has more tools than that. On the mainland, you can own a sim, but… once you sell parcels of it, they’re lost to your control.

“Ownership” on private islands has also lead to another change: there is no need to be Premium any more to be able to hold land. Since private islands grow so much more than mainland sims, the “need to be Premium” has become irrelevant — and Premium users continue to downgrade to Basic Accounts, since the extra benefit — customer support and a handful of L$ per month — is really not so important any more. LL knows this and T Linden has launched the debate around Premium users. Clearly they understand that the need to be Premium is not important if you don’t plan to ever live on the mainland again.

Sooo what do I think that Linden Lab will do?

Easy. The new generation of Class 6 Super-Servers could be an excellent pretext to revamp the whole pricing structure and revamp the mainland to make it more attractive. This is hardly “Old News”, people like Prokofy Neva have asked for this since at least 2005:

  1. Introduce Estate Tools on new sims on the mainland
  2. Raise the avatar limit to 100, just like on private islands
  3. Keep the US$195/month price on mainland sims
  4. Get rid of Premium accounts altogether

Number of developer hours to deploy this: zero.

So this would truly be a win-win situation. People angry about the high prices of the new “Homestead” sims would flock to the mainland, where for a little extra they’d get the full glory of 15,000 prims and 100 avatars and no restrictions — and Estate Tools. Mainland sims are auctioned, so, again, LL might start making some serious money there (private islands are sold second-hand, but not on the LL auctions, so the market for private islands is not so “visible”). Communities might fight to get back to the mainland, where prices could be slightly slower than on the US$295/month private islands — but, of course, they’d be limited to the available sims that they could buy wholesale, so large communities like Azure Islands, Dreamland, or even Caledon might still remain on private islands. However, most private islands out there are isolated sims in the middle of the ocean — and they could come back to the mainland with everything in place. If this is well planned by LL, they might even allow existing communities on private islands to “come back” to the mainland at no extra charge — or allow OpenSpace customers to “migrate” to the mainland without setup fees (just think on how many would happily do that!).

Sure, there is the problem of fitting the pieces of the puzzle together — after all, there is some terraforming that gives mainland sims some unique look, specially on the coastal areas. This would just be a question of adapting. Many, many communities on private islands just cover the whole sim — they could be simply dropped in the middle of the mainland, and remain there without a problem. Others are “coastal areas” which might require some adaptation — meaning patiently going on a case-by-case basis, putting some water sims to adapt, and they’d be “islands connected to the mainland” instead of “private islands isolated in the ocean”.

I think this could work. After all, most private islands are well-planned and look much nicer than the mainland — even the worst-looking community on private islands is not as ugly as the mainland. And, of course, the best-looking communities already on the mainland (Lusk, Ravenglass Rentals…) are at par with the private islands’ own rental facilities, but they just lack tools to enforce a better environment and keep griefers away from their community in an efficient way.

This would probably mean that from the 13,000 or so OpenSpace sims, perhaps half could be converted into mainland sims, with a little effort from LL. Many thousands of regular, isolated private islands would certainly move back into the mainland just for the lower tier. A few multiple-island communities would also do the same, providing that Linden Lab would “adapt” the landscape on the mainland to adapt to their requirements. And only very special projects — or the largest private island communities with hundreds of islands — would have no reason to “come back” to the mainland.

What is the biggest disadvantage of this model? Growth. On the ocean, you can easily grow to all sides, and even if you come in contact with other residents’ own expansion efforts, for a small fee you can move elsewhere and grow from there. On the mainland, space is restricted: you might be unable to grow beyond your original one-sim-community, if all land around you is bought and their owners have no intention to sell. Also, there is a problem with the existing, non-Estate-Tool-enabled sims — Governor Linden would still have to be the Estate Owner for them. However, things change — and many furious owners of plots on the mainland might simply move to neighbouring sims where a charming Estate Owner is providing a friendly environment for about the same price and go to live there instead. There would be a lot of reshuffling of parcels, which also means a lot of L$ changing hands (again) — specially if LL makes sure that the Class 6 sims with Estate Tools have better performance and allow more avatars for the same price.

I personally think this would work. The important thing to announce is that prices would not change — neither go up or down! — but that the newest technology would be available on the mainland only (at least for now) under the new model, and that people willing to drop their private islands and come back to the mainland would have special arrangements (either forfeiting the setup costs, or giving them first pick on the auctions, or something creative like that).

That way, 2009 might reveal a different landscape for Second Life — literally so — as the contiguity of Second Life is once more restored. As it always should have been.

Print Friendly, PDF & Email
%d bloggers like this: