Is Second Life the Paris Hilton of Virtual Worlds?

An anonymous reader writes "Second Life appears to be suffering a bit of a backlash from its PR efforts. Matt Mihaly over at The Forge, newly-returned muckracker Peter Ludlow at the Second Life Herald and Tony Walsh at Clickable Culture have all recently taken Linden Labs to task for their non-stop, arguably deceitful, PR machine and frequent downtime. Further, over on Terranova a veritable cornucopia of long-time, experienced virtual world developers, including Raph Koster, Mike Sellers, Randy Farmer, the aforementioned Matt Mihaly, and Daniel James, have piled on, calling into question the fundamental utility of Second Life. Does Second Life have real utility, or is it simply an endless exercise in unsubstantiated public relations? What do Slashdot readers think?"

Comments

Utility?

[Anonymous Coward]

Does Second Life have real utility

If by "utility" you mean does it provide enough enterainment to warrant thousands of people paying a monthly fee to engage in it, then the answer would be yes. If by utility you mean, "does it serve a bigger good", then I'd say, how many of the other MMORPG serve any more utility than what I first mentioned?

Does Myspace?

[Bieeanda]

Because seriously, aside from the graphical angle and the constant downtime issues, the two services are extremely similar. Neither are particularly well policed, both allow for truly godawful site/plot designs, and both are increasingly heavily occupied by corporate PR interfaces.

The whole reason for the existence of both is social networking. Making money is clearly a big deal for some SL users, but without other users to actually buy their virtual goods or rent space to build upon, the creators/sellers wouldn't have a market.

Personally, having been in SL off and on for over a year, I think it's a product with limited shelf-life. The developers have been promising big things, like better physics, rendering and interface tools for next to forever, but between community resistance to change and their own middling competencies (not to mention popular interactive items that depend on bugs and bad scripting to function), their efforts have dwindled to very basic bug-fixing and quality of life tweaks, while doggedly chasing after investment capital. Major changes risk forcing the users to re-learn or rebuild their projects, but at the same time other outfits are developing similar applications that leave SL in the virtual dust.

Second Life is totally non-scalable

[Morgaine]

>> it's too early

It's not too early to understand Second Life's implementation and to call it a disaster, because that is not changing. And it won't change. We've explained the problem to them repeatedly, to no avail.

I've been on Second Life a couple of years, and I still am, because the concept of Second Life is fantastic and I would love to see them succeed. But it won't, it can't possibly, because it's designed like a toy instead of for growth.

The problem is simple: SL's servers are mapped physically and logically into a static grid, where each server implements a fixed number of zones (called "sims"), usually just 1. This server does all the processing for everything in that zone (excluding database), and that includes all objects, all land-related storage, all scripting, and all handling of people in that zone.

Now early on in SL's life, some incompetent designer convinced the CEO that this is scalable, simply because you can extend the grid north-south and east-west as much as you like. Unfortunately, he or she failed to see that this is only scalable as long as all people and all objects stay in their home zones. Needless to say, that kills any prospects the world may have had stone dead. No crowds, no major sporting events, no well-populated pop concerts, no nothing beyond nightclub size, because 1 machine per zone (no matter how powerful) simply cannot scale that way.

Replacing each sim server by a cluster can't help, because SL zones can't be processed in a distributed manner. Huge multi-core SMP machines operating on a single server image might work, but then their entire business model of "one cheap machine per zone" would break down. And they can't put just a few big-iron machines in and restrict the large events to those zones, because anyone can hold an event on their own land, and that would discriminate between zones.

Another way of explaining the problem: processing people takes up most of a zone server's CPU in Second Life, but when people move from their home zone to another, the CPU power of their home zone does not follow them. So the server at an event is massively oversubscribed, while the one at home is now idle. It's inherently non-scalable for events and for objects that move between zones.

I've told their CEO and lots of other people there about this many times (and given them dynamically scalable solutions too), but it's bad news so the message is accepted politely and then ignored.

And yes, it *is* very bad news, because not only does it mean that Second Life has no future as it stands, it also means that there will be a revolution should they try to retrofix it. Because you see their business model is based on people paying for computing resources, and the economics of a dynamically allocated design are radically different. 400K+ landlords will suddenly find that their "investment" is now worthless, because land acreage is merely inactive storage in a dynamic architecture, and will cost almost nothing.

Which is almost certainly why Linden Labs haven't bitten the bullet and replaced their static design. It will be too painful. And now it may be too late.

Still, I wish them luck. The concept of Second Life has huge potential. It was just let down by a system architect who didn't understand scalability in a living virtual world, where people actually leave home and want to gather in events.