What Would Lum Do (If The Lindens Listened)? Veteran MMORPG Developer Scott Jennings On Improving Second Life

Scott "Lum the Mad" Jennings is one of the most respected analysts and developers of traditional fantasy MMORPGs-- on that subject, in point of fact, he quite literally wrote the book-- so when he recently told me for GigaOM that the industry's standard subscription MMO model is “broken, an arms race that few can even hope to compete in, much less win,” it provoked a flurry of passionate conversation.  (As here on Slashdot, here on Massively, and perhaps best of all, on his own influential blog, Broken Toys.)

As it happens, Jennings is also an occasional Resident of Second Life, where he's known as Lum Lumley.  His 2007 post on the metaverse, "Utopia Hidden Underground: Another Look At SL", is based on his in-world observations, and experience in mini-MMOs like City of Lost Angels, and to my mind, it's one of the most incisive essays on Second Life and its ability to foster a network of in-world communities that succeed and thrive (especially for women) despite the world's many painfully obvious flaws.

Many of those problems still exist, of course.  So, what top three things would Lum advise, if Linden Lab asked him?  And how could the Lindens best help developers of the many in-world MMOs?  I asked Jennings both questions last week, and he laid it out thus:

"1) Invest heavily in customer service. While the libertarian nature of SL is responsible for a lot of its growth, the fact remains that online worlds are services; services require support. The perception is that SL is lacking in customer support-- to rectify that there needs to be a visible and recognizable CS presence.

"2) Clean up the mainland. Ad/spam farms give new users a horrible impression of SL. Again, this goes contrary to Linden's libertarian impulses but is part of community stewardship. There are plenty of private island/landholder use agreements that Linden can crib from, and of course, enforcement of this will require more from customer service, which is why this is #2 and that was #1.

"3) Partner with a third party to provide a services directory. Even with the search improvements (and yanking 'Popular Places' may have been the best quality of life improvement Linden's done this year!) it's just too hard to find things if you don't already know what you're looking for. A Lonely Planet Guide to Second Life, if you will. I'm somewhat surprised someone isn't already doing this."

I pointed out the Lindens did announce a policy to end ad farms.  "Both the added customer service and crackdown on errant zoning needs to be visible and accountable," he answered. "Simply announcing a policy and then not enforcing it (as is the case with the ad/spam farm takedown notices) merely reinforces a sense of bad governance."

As for the many mini-MMOs based in Second Life like City of Lost Angels, which Jenning's played (seen here as "an insane angel from the future"), how could the Lindens best serve them?

"[O]ne thing that would be absolutely required," he told me, "is improving the client/server platform to bump up the number of avatars that can be in a given sim at once. That's what seems to be COLA and other games' biggest roadblock, and dwarfs scripting needs and payment collection (both of which, oddly, SL seems to have covered already)."

Mixed reality images courtesy of Lum.

Comments

[Robustus Hax]

That really is the biggest issue I think and will continue to be until they can fix it, the amount of avatars in a given area. A big social event gets cut off, or a concert unseen. Even when an area is full, its almost a terrible experience with lag and such. It was a big reason I started up Metaverse TV, so I could broadcast events to a wider audience than was allowed in a sim. I was unable to get into the SL 4th Birthday opening keynote. The Lindens announced (to over 30,000 people online) to come to the sim to watch the keynote.

[Kate Amdahl]

"I'm somewhat surprised someone isn't already doing this."

I'm working on it...! But even more important than finding services, if you ask me, is finding *people*. For a place with tens of thousands of people in it at any given moment (when the grid is up), Second Life is no good for finding friends and activity partners, because you can only cram a pretty small number of people into any particular location, and then the best you can do to get to know them are to start looking at profiles and try to start conversations if they're not already swamped with IMs. Solution on the way! (Who wants to beta test?)

After people to do things with, next I'll tackle things to do, and I've been taking some preliminary steps already. :)

^^^\ Kate /^^^

[Sophrosyne Stenvaag]

It seems to me (3) would be nice, but (1) and (2) are critical to SL's continuing. The blight and anarchy of the mainland defines SL in the minds of a lot of people who might otherwise come here to live or to do business.

LL's model of refusing to manage their world, and refusing to provide Residents with real management tools, is not only bizarre, it's self-defeating.

In every other community in history, either the people who built it *ran* it, or some sort of system of management and justice emerged to fill the vacuum, or the real estate developers sold or rented to someone who *would* run it.

SL and earlier social digital worlds have been an interesting experiment in enforced anarchy. It's been important to try, but it seems to be time to move on to a wider array of social solutions.

Lum's really called it - is anybody listening?

[Kate Amdahl]

Well said, Soph!

^^^\ Kate /^^^

[Nexus Burbclave]

I would respectfully disagree with the idea that bumping up sim concurrency is an immediate need. While this is certainly a nice thing for putting on larger events, I can count on one hand the number of times that I've been unable to teleport to a full sim, while I've long ago lost count of how many times I've teleported into a sim to find that I am the lone dot on the map.

I would argue that better tools for promoting and finding places and events far outweigh the need for increased concurrency. There are (admittedly imperfect) technical workarounds for the concurrency issue that can be leveraged for large events, but there seem to be fewer tools for growing events into ones that need such workarounds.

[Ann Otoole]

They have a very long way to go to increase how many avatars can be effective in a sim and have a good experience given in the last year the number has dropped from 70 to about 10. So they have to go through the code and find out what is causing all the erratic timing in scripts with the latest mono version sim code and whatever else is slowing it down. Hair fair was a great example of how child agents can kill a sim.

And don't you think it would be wise to move to a real transaction manager system to end failed transactions and beef up sim capacity before worrying about making it easy for 5000 people to try to get into one sim that croaks with 50?

I would love to see SL handle 250,000 online and the Lab then be able to do some google ad sense campaigns.