Linden Lab Approaches A Crossroads
In the space of two days, Linden Lab, maker of the virtual world of Second Life, has made three separate announcements that indicate trouble may be on the horizon for the company — if it hasn’t already arrived. Two of them have raised renewed alarms about the platform’s scalability. Though CEO Philip Rosedale blithely assured the world last summer that Second Life could “scale to inifinity,” he seemed to be the only one who believed such an absurd claim. The skeptics are now being proven correct. And instead of concentrating all of its firepower on solving the current problem, the Lab is also casting about for ideas on how to beef up its 2D Web interface — despite the fact that any number of developers large and small are already working on the same thing. Is trouble ahead? It sure smells like it. Read on for our conclusions as to why. Though this post may be overly alarmist, to a certain extent, we’re looking at serious issues here. Take it with a grain of salt, but not too big a grain.
Thursday saw the announcement that Linden Lab was seeking a leader of global technical operations. “Linden Lab is going through enormous growth and rapidly getting into operations challenges that most companies have never seen,” Rosedale wrote on the company blog. Those challenges, as any regular user of the service knows, center mostly around the fact that Second Life has been hard pressed to handle the influx of new users in recent months. (So much for infinite scaling.) To address the issue, the Lab will try to hire someone “at the VP or Director level” to “provide high-level strategy and technical leadership” and expand “our elite operations staff with the best possible people.”
This is a tall order. So as a stopgap measure, Linden Lab announced today that it is planning to limit access to the service, if needed, on weekends (when it sees the most use), blocking logins by users who have never transacted any business with the company but who rely on free accounts. Last June, we noted that Linden Lab was creating a de facto tiered service by labelling users’ accounts as “unverified” if they had never bought virtual real estate or purchased currency through the LindeX currency exchange. Now that de facto has become fact; it’s the unverified users that will be locked out in the case of a weekend overload.
Between those two announcements came a plea for assistance from users as to what they’d like to see in a 2D Web interface to be known as my.secondlife.com. The effort is described as a “major redesign” in coming months that will give users access to certain functions via a 2D Web site. This is great stuff, and well needed. But taken together, all three announcements present some problems.
First off, while the locking out of unverified users on weekends may not come to pass, it certainly looks like Linden Lab feels it will be needed. It can’t be a good sign that this solution, described as a “contingency plan,” was cooked up after the problem arose that it’s designed to solve. Nor can it be reassuring to investors that the Lab’s best solution, at present, is to lock out potential paying customers. Rosedale describes SL’s retention rate as in the area of 10-15 percent. If potential paying customers can’t even get on to try out the service when most of them want to, expect that number to decline further.
The move could also have the effect of alienating metaverse services companies like Millions of Us, Rivers Run Red, and the Electric Sheep Company (sponsors of this blog), among others — and alienating Linden Lab’s best source of new users in the process. These companies and their clients rely on the fact that Second Life is free to all. More than a dozen of the most prominent corporations in America — including all three major television networks, several Hollywood film studios, AOL, auto majors in the U.S. and Europe and many more — have asked their customers to look in on the virtual world experiences that these 3D services firms have created for them. Except that now they won’t be able to do it at the time that’s most convenient for the most users. And unfortunately for the Lab, these companies have been responsible for directing a large stream of users and paying customers toward the platform in recent months. They will be hard pressed to invite their users in if they have to tell them they might not actually be able to show up. Again, if you can’t get in to try it on, you’re not going to buy it.
In addition, corporations and news outlets like Cisco, IBM, Reuters and others are now beginning to use Second Life in the course of doing business. This is a crucial time for the Lab to show them that they’re providing a viable platform for such pursuits. Competitors are just now getting ready to nip at SL’s heels. If major third-party players are turned away by technological problems, they may not even be interested in such competitors. More would be the shame.
At the end of the announcement of the contingency measure, the Lab also laid out its technological roadmap for the next few months, which will include additional caching systems, reducing write load on the central database by partitioning or removing data, addressing critical bottlenecks, and deploying more internal ‘web services.’ “Beyond that we are building ambitious plans for re-architecting the Grid, so that in the future we can realize the full potential of the Second Life Grid to support millions of concurrent users.”
If all this is the case, why on earth is the Lab launching a simultaneous push to increase its services to users on the Web? Not only does it divert resources away from the work on scalability (though of course not every engineer at LL is even qualified to work on that), but it comes at a time when the very foundations of one end of that interface are being rethought. If there are “ambitious plans for re-architecting the Grid” afoot, would it not be better to wait until those plans are settled before writing a system that is to interface with them?
Another important ripple from this 2D effort, though, will be felt by the many developers currently working on the very same services. Companies and developers large and small are currently taking advantage of the fact that the SL client code is now open source to build the same kinds of services LL mentions in its announcement: Web-based management tools for things like friends lists, account management, and various other things. Does LL expect to do an end-run around them and provide a native service that will box them out? It would go against many of the precepts on which the world has been built, but it wouldn’t be the first time this has happened.
Readers of the Second Life Herald will recall that the LindeX currency exchange was opened in August of 2005, in direct competition with a popular user-run currency exchange, which, though it wasn’t forced to, chose to close up shop rather than compete with LL’s 100 percent penetration. Unfortunately, this occurred after LL had had several meetings with the users who ran the exchange, and had allegedly extracted from them many details about its workings.
It now looks like something similar is in the works. 3pointD hears that several of the parties developing Web-based services for Second Life have met with Linden Lab to discuss what they would need in terms of access to database information that would smooth the implementation of such things, and Rosedale has said in the past that the Lab would provide hooks into that information that would be accessible to all. Now LL is forging ahead with its own implementation. This approaches something that stinks. LL’s argument in the case of the currency exchange was that such an important function couldn’t be entrusted to a third party, and there’s some merit to that argument, though not particularly much. The same cannot be said here. The Lab has shown no sign of being particularly handy with Web-based services; why duplicate the efforts already under way by people who may well be better qualified for the task?
Linden Lab’s apparent goal has been to become the gold standard of 3D virtual world platforms. They’ve been successful in that so far in large part because there has not been a competitor as free and open as Second Life. What the Lab looks like doing now, though, is putting up more barriers, rather than pushing further into the openness that has been their biggest comparative advantage. The problems they face in scalability and the task in redesign are not small. (The very concept of a wholesale re-architecting of the platform should rattle nerves.) The good will they’ve won from metaverse services companies, from those firms’ clients and from the press, while it may appear to be stellar, is in reality quite fragile. A backlash has been straining at the leash; the only thing that’s kept it from breaking free is that SL has continued to outpace it. But this week’s developments have the Lab dancing in several different directions at once, and very few of them can be thought of as moving forward.
This is a state of affairs that can only last so long. Linden Lab will have to decide, and soon, which direction it wants to move. Rosedale’s unfortunate flat management style, in which even key teams have in the past been largely self-directed, and feedback is provided by a love machine, cannot stand up to the exigencies of cutting-edge technology. If you want something done, you have to tell your staff exactly what it is.
First, though, you have to decide just what it is you want done, and it’s here that the Lab probably has the most work to do. It’s no use striking off toward three different futures at once; you’ll never reach any of them. Does all this spell the end for Second Life? Probably not, or at least, not yet. But it does put Linden Lab in a more precarious position, and it would take only a little bit of good press directed toward the right competitor for Second Life to be knocked from its perch. That competitor has yet to raise its head, but several are gestating. This year will be a crucial one not just for the future of Linden Lab, but probably for the future of the metaverse itself.
If Second Life falls apart, will the 3D Internet go with it? Probably not. But we will have lost one of the most exciting platforms to come along in quite some time. It will be interesting to see how LL handles its current challenges, and who comes along to further challenge them in the meantime.
[UPDATE: Read a follow-up post featuring a few comments from Linden Lab chief technology officer Cory Ondrejka.]
(Woah. . . . At over 1,800 words, I believe that takes the prize: Longest. Post. Evar. And I vaguely remember a time when all I did in Second Life was a little virtual reporting, some hanging out and some ham-fisted building. Ah, those were the days. . . .)
Feeds
It might be a little alarmist, but ultimately it really is a pretty scary turn of events. The future of the 3d internet clearly isn’t LL, unless they make a lot of foundational changes (open source server, 3rd party hosted islands). They need to let go of the “not hosted here” concept and realize that they have the opportunity to influence (not control) the evolution of the internet.
I think if you listen to what Mitch Kapor says in the Davos interview with Adam Reuters some of the longer term thinking comes into play around the not hosted here.It does not alay all the fears raised but sounds very much more like the way I would like to see it all going.
I am sure though this long post will get lots of comments, so I will leave mine to this brief one :-)
Well…if SL does collapse, the upside will be one of the biggest parties every thrown, courtesy of Valleywag. Those guys look for reasons to hate Second Life.
Instead of focusing on the 2D web UI aspect of the announcement, you might consider that this means that (if they’re smart) they’ll create web APIs which can be used by both the Second Life web app and third party apps.
This could be another sign that LL intends to move non-3D functionality out of the SL client and onto the web, which would go a long way to merging it with the rest of the world.
For true scalability, I keep my eye on the OpenCroquet guys, they seem to have smart people thinking about the problem. And if how long it is taking them is any indication, then SecondLife has their work cut out for them. Because the common approach I see to scalability nowadays requires dollars, pure and simple, and the dollars either come from deep pockets internally (Google, Yahoo, etc), or you make the customers pay (World of Warcraft).
While I’ve been keeping an eye on OpenCroquet, it never seems to go anywhere, and has one of the least usable interfaces I’ve ever witnessed. Linden Lab is going through some growing pains - this happens. It happened in the mid-90s to companies like eBay, Yahoo, Amazon and CDnow who all suffered significant downtimes with a much smaller scaling problem that Second Life presents. Projects like OpenSim are already leading the way to hosting your own simulators; they’ve made quite a bit of progress. More here: http://opensecondlife.org/wiki/OpenSim With the open sourcing and external development, it’s not hard to imagine that (the SL protocol + open simulator + something like Amazon’s S3 storage as an asset system) might equal the 3-D web we’re all waiting for. I see Linden Lab’s future as something of a cross between ICANN (who gets to go where on this gridded map), running of the micropayment economy, and something similar to MySQL AB’s model of service.
I don’t see this announcement as “the sky falling” - I’d much rather have a grid that’s usable at 95% capacity than one that’s unusable for anyone at 120% capacity. A throttle makes sense until these other problems can be fully addressed.
I don’t think the sky is falling either, Flip. I really hope it isn’t LL that becomes the ICANN of the 3D Internet, though. ICANN is plagued by enough problems without a profit motive thrown in.
I agree that a throttle makes some sense. There’s just not a good way to do it. I dislike the idea of locking out unverifieds, but I also dislike the idea of locking out people who have paid for access to land.
On the other hand, you generally get locked out of most Web-based services thar are struggling with load on a first come, first served basis, whether you have a premium account or not. tbh, I think that’s the preferable model. It avoids making value judgments about customers and prospective customers, which is something LL needs learn to do.
[…] Mark Wallace raises a number of valid concerns in a record breaking 1,800+ word post which is recommended reading. I am less concern than Mark about Philip, Cory and the Linden Lab team getting the next steps right or the future of virtual worlds. With input from OpenSL and a number of other communities work around’s will be found. […]
Walker, as you can see, FlipperPA, who represents ESC at least in some capacity, is not one of these people willing to publicly bitch about the Lab and make these concerns more explicit. Indeed it is FlipperPA who has led the charge in public statements about the need to throttle the free accounts. He can only be happy at the free accounts being swept away, as he doesn’t need them on the SLB which he sold to ESC — SLB needs to have the Linden system of payment and dollars verifying people first, before they can verify an avatar, and a free account with no payment means on file is useless to SLB and other shopping sites. Free accounts in fact, in the eyes of the foremen of Second Life like Flipper, are merely drags on the CPU, and competition to their own business, since they go to the games, casinos, whorehouses, and other venues that are outside the FIC domain of business for the most part.
I want to take you to task on this statement, as well:
“And unfortunately for the Lab, these companies have been responsible for directing a large stream of users and paying customers toward the platform in recent months.”
I don’t see anything that really confirms that claim. Where do you find that
After seeing this really ridiculous claim in PC magazine that IBM had “thrown 2000 engineers at Second Life,” I went to one of the IBM guys and parsed it bit by bit and came down to the fact that it was more like…200 people who had logged on and joined an internal IBM group and had made an avatar and such. Not even “engineers involved in thrashing the problems of SL” but just IBM staff, which could be anybody, and just be playing around in SL just to have a presence.
I don’t think from my own interpretation of SL, which is based on my customers, polls, and explorations inworld, that these big companies are providing any stream of customers for SL. Baloney. They bring in their staffs — if that. They are managed by the sherpa companies, and they don’t even log back on ever, much of the time. They come for the big events and then disappear. They may occasionally come back, but not in any “streams”. If you know these companies and have good ties to them, as sponsored by ESC, I’d challenge you to do the journalistic footwork required to go to them and ask them what the hard numbers of whom among their staff or customers or commerce circles really have been brought into SL.
I go to the L-Word sim occasionally and I see…Showtime and ESC staff building up the traffic numbers, dancing and hosting. It’s wierd. Many of the people in the “audiences” are…sherpa staff, creating the illusion of a full house, it seems to me. What are the REAL customer numbers from outside the magic circle?
I think the stream of people coming into SL now are coming in from the old-style RL media, especially in Europe. The articles in all kinds of European publications have brought in new people. Articles that continue to be seen in the U.S. press also bring in Americans. Media of the kind you can’t gin up in SL due to the 40-avatar limit, outside of SL, is still what makes SL work.
I also want to try to get a better handle on the actual physical problems of the scaling issues. Are Philip and company amateurs? Don’t they have degrees in engineering, physics, computer science. Have they never thought about the problem of how to pack sims into a refrigerator and keep it cool? This seems like basic stuff to me. What, they didn’t think of it, they’re stupid?
You might convince me that they are, without too much trouble, but I think there’s more to it. I think that the engineers who designed it and imagined it as infinitely scalable did so with empty servers, or servers containing a few of their own avatars or low-prim builds. They weren’t interested in really creating the world from the inside out, so they didn’t really conceive of eveybody having inventories with 16,000 objects like mine, with prim hair, shoes, swords, etc. that can sink entire sims with lag and prim overload, so high-prim are they (Reebok’s has sneakers that are 217 prims for each foot).
I also think they looked at their own usage for 3 years and saw people logging on for long hours, but just scripting or building — keeping a kind of “light footprint”. I think they didn’t really envision the heavy usage involved in people running islands with clubs, live music, scripted genitals, etc. etc. — these are people who became myopic studying intensively the usage and numbers of their own kind, and their own circle, and batted away in annoyance the facts coming in about how other kinds of people used the grid.
I think there are various ideological hobblers in the way of the Lindens that they can’t let go of to make things improved:
o socialistic ideals — they refuse to charge CPUS that would make clubs stop stealing it all on every sim, and they refuse to charge for inventory, letting people keep monster inventory
o hatred of commerce typical of the tekkie-wiki outlook — telehubs work to distribute loads of shoppers, and keep people from constantly using the map to find sims in the search list, or to use p2p which means pulling up the map. I realize it’s horribly politically incorrect, but when they removed telehubs because they “lag sims” or “trap avatars” (only a minority of them did that), they introduced other stuff that is even MORE laggy — map-opening to pinpoint teleports, or SEARCH PLACES opening to find stores by name, or the map list. These all involve expensive database calls.
o There is a horrifically huge volume of land for sale by people who can’t bear their laggy sims, or who have gotten fed up with SL and left it. If the Lindens did a *modicum of mainland zoning and steerage* chiefly by keeping clubs off waterfront sims, by labeling sims on the auction “residential use only” merely to create voluntary steering, they could eliminate the need of many people to sell land or leave land. There’s quite a few things that could be done to fix up poor distribution of mainland usage — but all of them involve soft human management, not the code-as-law toggle switches that the Linden game-gods favour. For example, if the Lindens didn’t have an ideological allergy to EVER deprecating scripts, in the name of ALWAYS providing MAXIMUM CREATIVITY RIGHTS to sandboxers and script-kiddies and NEVER taking into account property owners rights, they’d eliminate a huge draw on CPU that occurs with camping by *deprecating the Lazy Linden script and other camp scripts*. Wow, that’s radical, eh? Evil, huh? But that sort of management internally can reduce their load just as much arbitrarily stopping new sign-ups, which is ridiculous.
The Lindens hired 24 new developers in 2 months, but they didn’t hire 24 new community managers. That’s because they envision their job as getting out of the community management business and having everything go open-source and have people “host their own”. They are far from that place, however, and they need to see this as equivalent to hiring 24 oil-drillers and geologists and not hiring even one energy conservation specialist. The energy conservation involved in things like resolving the club problem, the camp problem, etc. isn’t just about client-side lag; it’s about literally cooling their own overheated servers in heavy use.
Sunday the 18th on CBS Sunday Morning will be a piece on SL, during David Pogue’s segment, Mark, might want to catch it. Not sure what the tone or vibe will be.
And great piece here, it’s also interesting in light of this EA buzz. Could EA do more damage to SL with the infrastructure they have?
All I know is that with my interest in building my sim-world with the XBOX kit this coming summer, and my new found love for accessing the web on my Wii, my metaverse just got bigger. It’s a faced paced race to survival, and I’m not even IN the goods or services content space, just the content-content space.
See you in NYC!
Great post - lots of insights and it pushed some of my buttons too. Some comments:
There are a lot of armchair enterprise architects and sr. business leaders :) Growing pains do not necessarily mean the sky is falling. It will mean change and you can’t keep everyone happy. Hell, you can’t keep people happy when there isn’t any change either. I have been in a company that went from 13 people to over 1300 in less than 5 years - you can live through that kind of growth but it is impossible (and a naive suggestion) to try and wrap a neat or clever management process around it. You need to something of course but you better hang on for the ride nonetheless!
Considering the above, 9 women cannot have a baby in a month and you cannot simply throw bodies at a problem to solve it. Outside contract do not work alone in a vacuum and show up at your door with a finished product. Contractors may have the necessary technical skills but will have little or no real knowledge of SL’s complex system. Getting contractors up to speed can be as costly as getting your own employees up to speed; sometimes more as the contractor may not be inside your infrastructure and internal processes. Then, like employees, you have to manage the contractors. Making a project even more matrixed can bring additional communication and timing risks.
Regarding metaverse companies fueling SL user growth. Though perhaps they tell their clients this, it is a big stretch to propose that a handful of metaverse companies are responsible for a significant influx of new users into SL. One would certainly expect to see the metaverse company built sims in the popular places listings if this was the case. A quick review of the traffic into the sims built by these companies contrasted against the growth of SL and the traffic of other sims reveals this is not the case.
Great blog Mark - I enjoy reading.
Peace!
Thanks, HatHead!
I don’t really think the sky is falling. I just thought I heard a distant creak.
As to new users: I do think that’s where most of the growth in new users is going to come from going forward, even if that’s not the main source at the moment. And if you look at some of the metaverse companies’ offerings, like Motorati Island or The L Word, you’ll see that these are far more sticky places than Linden Lab’s Orientation and Help Island. I would put money on the fact that users brought in by metaverse companies have a higher retention rate than those brought in through LL’s Web portal and marketing. That difference is only going to grow as LL steps back from marketing a particular kind of experience and metaverse companies grow more prominent.
Mark, thanks for your reply.
I most certainly agree that projects delivered by metaverse services companies typically provide higher quality user experiences and therefore a higher retention rate amongst these new users. This is a lot of, as I think you are noting, the value metaverse companies bring to the table; especially the imprinting of their client’s brand with the SL experience on new users (for however long that may hold).
Still, like everyone else I’m sure, curious to weigh the overall statistical significance of it all somehow. No easy task to commonly define measures and missing lots of necessary data to be all proper like about it anyway I think. Thus so it will likely remain for us outside of LL and even then sometimes things may not ever be possible as required data may not be captured or stored.
peace!
HH
Linden Lab Approaches A Crossroads…
A good article on 3pointD regarding the latest happenings at Linden Lab and what they could mean for the immediate future of the company (especially scalability):
1) the announcement of the position for a “leader of global technical operations”
2) t…
I’m a Linden Lab engineer, but these are my personal comments, yada yada. I’m not a policy guy — I just like writing code.
Scaling is hard. We made some design mistakes up front, most of which center around storing too much user account data in a central database. This stuff is moving to little web applications (which we can distribute to multiple machines) or to separate database machines (which we can just buy more of). It’s a slow, painful process though. I’m sorry it sucks for the community.
(Specific example: The list of users you are muting is stored in the central database. That table has millions of rows. It can be trivially partitioned by your user/agent id to the distributed set of databases we use to store inventory. I fixed that last week, it will ship with the next release.)
The code that causes the scaling problems is all in the client/server codebase, not the web site. We’ve beaten the web team enough that they don’t write expensive SQL queries anymore. :-) I don’t think that having them work on a better web site is going to hurt scaling work. We have different skill sets.
The open source client doesn’t help that much towards things like having external web sites or applications display friends list, group information, etc. The communication protocol we use to the viewer is UDP based and, well, how do I put this…. cranky. The right way to fix it is to build a web protocol based system to access that information that can be used by the client, the LL web site, and third party tools. I’m not on the web team, so I’m not privy to all the internal information, but my understanding is that is the direction things are headed. I am not an external develper (IANAED?) but I think work on a cleaner SL web site will provide easier information access to all.
Thanks for continuing to provide a reality check on our work. I really enjoy reading outside comments on what I do at work.
James
Thanks for checking in, James. And thanks for providing a reality check on 3pointD! Honestly, I do think it’s quite helpful to have detailed information like this out there, as it helps everyone to understand the direction a bit better than the occasionally somewhat terse LL blog entries do.
None of this should be a surprise to anyone in the industry who deals with high volume transaction systems. I posted in my blog back on October 9th that LL was in a crisis but didn’t seem to be aware of it.
This isn’t a technology issue, it’s a management issue. It has been painfully obvious for at least 2 1/2 years that scaling would be a major issue with SL, but instead of addressing it LL pushes out piffle about the infinite scalability of the model. This means one of two things; management doesn’t understand the problem (which current successful scalability model are they using as a benchmark?), or is trying to hide it.
Either answer will result in the same outcome - LL will get handed it’s lunch by the next up and coming metaverse company.
Linden Labs needs to take a serious look at how a company like GE handles crisis situations and use that model. Yes, scaling a company and a product is a heck of a challenge - but there are great examples in the industry of how to do exactly that.
There are tons of tech companies that worked very hard, had great idea(s), but never made it over the hump, or just plain failed. Many had dedicated employees who really believed in what they were doing, but were ultimately let down by a management team that couldn’t execute.
I’m hoping that LL can swallow their pride and do what it takes to get things under control. Personally I like SL in all its quirky glory.
I mean, where else can you watch newbies splat into glass that just rezzed in front of them?
Interesting discussions! FYI, I made some growth estimates and posted a spreadsheet showing how users may grow across rest of year — to 15million range! http://rezzing.blogspot.com/search/label/SL%20Growth
Seems that one of the main issues is the openness of the grid. Right now it is in general available to all people all the time. Is that really necessary? It would seem that some type of grid partitioning would allow for better performance as well as more stable user experience.
Seems that grid was okay at 20,000 concurrent, so why not just have 5 grids to handle the loading to 100,000? It would be interesting to see how the avatars could move between grids then. Maybe charge them for traveling and bringing stuff?
Some sims are not well organized for performance, lots of textures, particles, weird scripts etc. It is still unclear how a crowded, sub-optimized sim impacts other sims when they are load-balanced and sharing CPUs. Why should busy sims pay LL same for “land” as empty sims which cycle the grid less?
LL also does not have anything to say on service level agreements back to sim owners to outline performance objectives and sim availability. So why even make your sim more performance oriented when you still cannot get any commitment back that it will run better?
In terms of the unverified users, it seems that issue here is about age and access to credit cards. Certainly many sims have already blocked non-verified to protect from under-aged and griefers. The CC issue is more difficult to address as despite the aggressive marketing of CC’s not all people in world have them yet. Which on some level would mean that if you don’t have a CC you can’t earn money in SL which is an interesting drag on the overall SL economy.
Hubris indeed. The future of 3D, to the extent that has one, is VOIP and webcam. Not animated cartoons and floating chat bubbles. What LL has created is the Massive Multiplayer Online GAME of the future. Why do they not see it? Is creating ‘mere’ entertainment too trivial and demeaning for them?
It’s not trivial and demeaning, Mikyo, it’s just not what they’re doing.
(disclaimer: i’m not english mother tongue)
I’m on a “unverified” account, and I don’t feel discriminated by Linden Lab for that (instead, some “Verified” residents do discriminate, when they ban users on account basis and when they complain about griefers). I cannot buy land, but I can rent it if I need it. What else about “tiered service”?
Since I’m not paying, I’m not surprised that during overload emergencies Linden Lab is going to give higher priority to paying users. Doesn’t that make sense?
About “beef up the 2D Web interface”, I quote an insightful excerpt from secretlair.com (not that I’m fond of that website):
“Not only will the web-based tool set reduce the number of concurrent avatars in-world (and therefore help the entire 3D system run more smoothly), but it could help to make the user experience better by simplifying the standard-issue Second Life client interface–there’s no good reason we should have to log into a 3D world to perform 2D tasks such as checking instant messages, managing groups, or organizing inventory”.
About mainstream presence in SL (”a dozen of the most prominent corporations in America — including all three major television networks, several Hollywood film studios, AOL, auto majors in the U.S. and Europe and many more”):
it’s quite impressive to cite those names, if you refer to RL - but I don’t believe they really “have been responsible for directing a large stream of users and paying customers toward the platform in recent months”. I’d be interested in some more substance to this statement. Fact is, many of the sims they opened are often empty.
Mark, note that while I’m critic against this post, I really appreciated many other articles from you.
No worries, Signore, I welcome criticism here as well. Part of my purpose on 3pointD is to stir up comment and thinking about the issues we face in the metaverse. I don’t always get things 100 percent right, of course, although in posts like this it’s mostly just my opinion of what’s happening at the moment. I’m always open to other ideas, though, so it’s good to have people come along and share divergent opinions.
[…] Yesterday Mark Wallace pointed out Linden Lab is approaching a crossroads. I didn’t read into his article that the world was coming to an end because Linden Lab has set into place contingency measures to ensure the service of Second Life. I do see where Linden Lab is finally stepping up to the plate and admitting there is a problem which they are having more difficulty than they were willing to admit, and acceptance is the first step toward a solution. […]
Mark… one issue that hasn’t been mentioned here is the impact the new access policy will have on the education and non-profit sectors. There is some consternation on the Second Life Educators Listerv about how this will impact on educational programs, many of which are predicated on free and open access for students.
Introducing Second Life to administrators is tough enough without now also having to say that there is no guarantee that the platform will be available when required, or that students will have to get paid accounts - an option which is not viable for many.
Specific example: The list of users you are muting is stored in the central database. That table has millions of rows. It can be trivially partitioned by your user/agent id to the distributed set of databases we use to store inventory. I fixed that last week, it will ship with the next release.)
So, James, would it be fair to say that ideological decisions that are made about social engineering, that “it’s a good idea to mute people” and “we need to give everyone the maximum ability to mute as many people and things as possible” actually creates work for you, and creates millions of tables and lag and CPU draws and whatnot?
So that if you said higher up the chain there, hey, you know, let’s just stop this muting of people stuff. It’s pretty fascistic. I mean, this is so much more than just putting a spam filter on email, this is like Stalin erasing people out of photographs. Plus, it’s an expensive call on the data base. Let’s just suspend it for awhile because if people really are bothered, well, they can just not answer or ignore or log off or fly away and save energy.
That’s what I mean about energy conservation and community management versus technical solutions like “let’s figure out how to meet the huge growing constant demand how to mute the shit out of everybody and their objects”.
Sean, I’ll bet that just as the Lindens have already given non-profit status lower-priced sims to the educational community, they’ll figure out a way to put a flag on their accounts when bought in batches that will enable log in. Or they’ll accept the one professor’s payment on file for his entire class of last names or something. They can easily get around it. And they will make these kinds of political decisions.
The future of 3D, to the extent that has one, is VOIP and webcam.
The reason they call it “media” is because it mediates. Not everybody is ready for the real-life glare of webcam. People want filters, protections, anonymities, or at least choices. Also video is way more expensive than avatarization.
plus which, neither VOIP nor Webcams are exactly 3D, are they?
>> “These companies and their clients rely on the fact that Second Life is free to all.”
I’d say that’s a big assumption. Care to explain why you feel this is true, Mark?
Webcams as 3D..hmm food for thought/
Not to sure they arent mark…;)
What makes web3D a compelling mediatype is the immerssive nature of it.. its dimentionality and the reality of the broken Fourth Wall so to speak. Even qtvr or flash movies of a spinning object dont really offer true 3d since the seen object dosent truly exist in an x y z matrix with the viewer behind the screen.
A webcam can offer the exact same experience (x y z) without the need to “model” or simulate” reality in polygons or prims. Using the forgotten media type the - “natural world” webcams can move the viewer in all 6 directions and provide a base for a very immerssive experience in realtime… all presented to the viewer on the receiving end screen.
sometimes you have to look beyond the POINT. in web3d ..lol
larryr
What I meant by that, Hiro, and I’m sure this is true for you guys too, is that people like AOL would likely not be in SL if payment was mandatory. You don’t dip your toe in the metaverse if you have to squeeze money out of your users to do it. AOL, Pontiac and all the rest are only able to get in because they can say to their users / customers / audience, “Aw, it won’t hurt you to try, it don’t cost nuthin!” I seriously doubt they’d be there is that weren’t the case.
Data point on mute: The vast majority of entries in the user_mute table are people who have muted exactly one other person. The next most common number of people muted? Two. And so on. It’s a strict power law distribution.
[…] 就在我谈论这个话题时，值得注意：因为越来越拥挤（“拥挤”意味着大约38,000人－－很难是传说中的数百万人），Second Life正在陷入困境，林登实验室将转向“付费帐户”服务。更多信息见这里和这里。也许，现在是时候抛掉林登元了，呵呵？Randolph Harrison, Capitalism 2.0 February 19, 2007 [链接] [评论] […]
[…] So of course the SLogosphere is all a-twitter over the contingency measures announced by Linden Lab in order to prevent extreme lag. The method? Blocking logins to non-paying players. Plenty of commentary at Clickable Culture (Tony also has a nice reaction round-up post) and 3PointD , plus many more places I haven’t bothered to click through to yet… […]
But Mark, they aren’t there to reach their RL customers, they’re there to get the media hit to look cool in the newest thing. Isn’t that the recipe?
That’s a lot of it, Prok. And it’s those media hits that are bringing in the customers.
[…] “So of course the SLogosphere is all a-twitter over the contingency measures announced by Linden Lab in order to prevent extreme lag. The method? Blocking logins to non-paying players. Plenty of commentary at Clickable Culture (Tony also has a nice reaction round-up post) and 3PointD , plus many more places I haven’t bothered to click through to yet… […]
># James Cook on February 19th, 2007, at 7:13 pm:
Data point on mute: The vast majority of entries in the user_mute table are people who have muted exactly one other person. The next most common number of people muted? Two. And so on. It’s a strict power law distribution.
So James, why don’t you dump functions like “mute” if you want to save database energy, instead of endlessly chasing this feature? Why not limit mutes or even suspend them? Most people muted in fact don’t repeat. Many muted are unmuted. Muting seems like a luxury.
Walker, my problem with your claim that these are “their customers” is the idea that somehow, the hits that an AOL or a Sony or whatever gets in the media only brings in “their customers”. It just brings in *anybody* who reads the newspaper or surfs the web. It’s not as if these companies — most of them anyway — create special web pages and portals a la L-Word to drop their own cared-for customers into the world. They don’t.
First off - apologies for getting so late into this debate. But I was struck by the fact that Prokofy is the first person to mention The L Word in the context of RL pullthru into SL. As a rambler in SL, I’ve been struck by just how busy the L Word site is, and I would (perhaps wrongly) attribute this to their use of the (beta) Registration API. In other words, this is one case (the only 1 I know of) where there is a direct, clear pull-through from RL.
On the wider issue of LL’s current “strategy”, this looks (to the outsider) ill thought-out and reactive - not really a strategy at all. Perhaps most damaging is the high-handed way these snap changes of policy are introduced (the removal of First Land rights is the latest). This serves to destabilise, resulting in long threads like this in many blogs. If I were IBM,say, I’d be banging on LL’s doors, demanding an explanation and pointing out that policy change “on the hoof” was not only damaging their standing, but also that of their customers (specifically in this case, IBM).
This has nothing to do with ToS (which presumably say LL can make such changes on a whim), it is to do with LL’s credibility. Lose credibility with your customers [and prospects] and you will lose all that had-won market you spent so much of 2006 trying to acquire and retain. SL could sink back into the quiet, quirky backwater it inhabited in 2005 - (maintenance of 2000+ servers notwithstanding) and the 3D internet dream could be badly bruised into the bargain.
To sum up: I don’t have any particular issue with any of the steps taken by LL, but I do have an issue with the way they have taken them.
PS: I’m not with IBM!
[…] Millions of users, problems accessing accounts, slow loading times, and unable to see messages. Sound familiar? A lot of people feel that this would be the same issues that plague Second Life and it has been some of the beginning points as to why Second Life, instead of growing is collapsing on itself. Both Mark Wallace and Tony Walsh have both expressed their opinions about the recent contingency plans announced by Linden Labs as an admittance of their issues scaling the platform, a measure they have talked about many times in the past two years. […]
[…] This piece and comments talk about this in relation to Second Life; Linden Labs apparently said it thinks the world can scale infinitely. This seems doubtful - though at least the inherent limitations of each server processor mean that you can’t assemble too many people in a single place at a time. (Hard to plan a revolution within the confines of SL!) The game worlds all are grappling with the problem of abuse in the sense of farming and other exploits that allow people to cheat (buy a level 70 warrier account on eBay, pay someone real money for their magic sword, etc). […]