Cafe Bucky

It is the community that will win the Metaverse…

Overview

I see an inevitable migration on the horizon of residents from SecondLife to Opensim
based grids. There are two main forces at work:  the decline of Linden Labs, and the
ascendency of the OpenSim.

• Linden Labs has been making a series of moves that are alienating the users.  They are neither leaders in policy, nor technology.  Not even meshes will save them.
• OpenSim is the clear technical leader, is rapidly making strides to stability, and the variety of grids springing up will assure that every community will have a strong home.  OpenSim is much like X11 in the sense that it provides “mechanism, not policy”.   Grid Operators take on the role of policy.

The Decline Of Linden Labs

Just since the time of SLCC (August 13-15, 2010), I can point to no less than four Linden Lab missteps which have upset various parts of the community:

1) Display Names – with over 1000 comments on blog, the Lindens appear to be intent on going live with a very ill-conceived plan to allow anyone to choose almost anything as a display name. What’s the problem with that?  They are not providing a means for users to protect their names from abuse.   While it may be true that a click or two will always reveal a true user name, it is still not clear what (Display Name?  User Name? Both?) will be written to chat and IM logs.  It is also not clear what the IM logfile names will be.  This is a huge worry for anyone who is concerned about identity and reputation.

The Lindens take the view that people can file Abuse Reports.  I note that this approach
attempts to address a symptom, but is no cure.  With all the layoffs that the Lindens have been through, and the recent reduction in support hours, do they have the staff to handle the increase in ARs that will be generated by Display Name abuse?  It would cost them less
to provide a control to prevent others from using your name as their display name.  It would also go over with the community a lot better.

2) Teen Grid Closing - I can understand the economics of closing the Teen Grid.  It costs a lot to run, and has a mere sliver of the main grid population.  What has a lot of people up in arms is the misguided idea of migrating 16 and 17 year olds to the main grid.  They would be better served by migrating to a grid that has been proven to be more focused on education (Reaction Grid comes to mind).

3) Customer Service Cutback – The short of it: same tiers, less service. Combine this with a slew of new problems (numerous ARs occuring from Display Name abuse, and legal/customer service issues arising from the presence of teens on what used to be a strictly adult grid).  The hours are also incredibly USA centric.  This will be one more factor that drives European and Asian users towards grids based in their regions.

4) Snowstorm, Do We Care?

Wizard of Oz: [in a booming voice] Step forward, Tin Man!
Tin Woodsman: [terrified, steps forward] Ohhhh!
Wizard of Oz: [still in a booming voice] You DARE to come to me for a heart, do you? You clinking, clanking, clattering collection of kaligenous junk!

Although SnowStorm is a great concept, in that there will be one viewer source code stream internally and externally, I sense that much of the developer community will focus their efforts elsewhere.  The Wizard of Oz Linden (who, as far as I can tell [1], has only
been in SL for 4 months) has declared:

“Don’t waste everyones time suggesting that we throw away Viewer 2, or that we revert the UI to Viewer 1. It is absolutely not going to happen, and any suggestion to that effect will be ignored.”

Yes, Scott really said that: http://www.mail-archive.com/opensource-dev@lists.secondlife.com/msg02459.html

Since the overwhelming majority of the community uses third party viewers, and does not like the 2.x UI, sensible developers are going to focus their efforts on giving the community what it wants.  The Linden Lab attempt at outsourcing some of their viewer development for free, while at the same time insisting on the 2.x UI, will not get much traction.  There is some limited cachet in being able to say that you contributed to the mainstream SL viewer, but if its usage is eclipsed by the offerings of several third parties, is it the best use of your valuable developer time?

The Ascendency Of OpenSim

I’ve touched on just some of the factors that are upsetting the SL community.  It’s time to shift to the positive.  We should be very glad that we are seeing the rise of a viable alternative!

As I have written before, and as others (such as Gwyn and Maria) have noted, OpenSim is showing itself to be a serious metaverse platform:

Technical – HyperGrid, MegaRegions, choice of voice (Whisper, FreeSwitch), more flexibility, compatible with existing SL viewers, archiving of regions and inventories.  Willingness to evolve past LSL.  Region scripting as well.

Policy – widespread, dependent on Grid Operators, if you dont like it, run your own grid!  Opensim grids win because none of them has to address the issue of trying to come up with policies that would satisfy everyone in a large walled garden.  We can easily move between web sites with varying policies and narrow niches.  The same thing will be true in the Metaverse.

Cost – a fraction of what LL charges –  Is there a market for high end grids with hand holding support, as well as a slew of specialty bargain basement offerings?  Yes!

Currency – yes!  OMC and Cyber Coin Bank These work across grids.   Maria writes in depth about this.

Scaling – with one fell swoop, HyperGrid does much to address the issue of scaling.  How?  If people can easily get from one grid to another, then there is little reason to try to shoehorn them all into one place.  As we see with some web sites, there will be some huge grids, possibly with millions of users.  But balanced with that, there will a large number of grids that address niche communities, with 10’s of thousands of users.  If they choose to be isolated, they can turn off hypergrid access.   Within 5 years, the combined scale of OpenSim based grids will dwarf the walled garden of SecondLife.  It is like the old AOL service trying to compete with the entire web.

Third Party Viewer Community -  Not only will the viewers based on 1.x and 2.x source streams become the preferred entrance to OpenSim based grids (easy grid management in them), there will be a movement towards more types of front ends.  HTML5 is one possibility (resurrect AjaxLife?).  Unity3D holds a lot of promise.

What Could LL Do To Stave Off The Inevitable?

At this point in time, the biggest technical innovation LL has on their agenda would be meshes.

On the development side, they would have to show that they are open to some emulation of the 1.x interface, in order to gain traction.  They are competing for something valuable, developer attention, and are making a lot of wrong moves.

On the policy side, they need to show that they will provide adequate controls for display names, so that people can be proactive about protecting their identity.

On the support side, they would need to make a case as to why people should pay the same amount of tier, but accept a reduced level of service.

On the interopability side, LL finds itself in the position of being a large walled garden, a la the AOL service of the 90’s and 00’s.  LL should ask itself it wants to follow the examples of AOL, CompuServe, and Prodigy.

The Tipping Point

Having said all this, it is quite clear that the majority of Metaverse usage hours are still being spent in SecondLife.  I dont see any single event from the sides of OpenSim, or SecondLife, that will radically tip the scales.  I see something more organic going on: friends will drag friends to OpenSim grids.  Any one of several LL policy decisions will be a straw that breaks a residents back.  OpenSim will make some technical advance, etc.  There will be as many reasons to switch as there are metaverse residents.  The Tipping Point occurs when a sizeable number of people realize that their personal puzzle pieces have fallen into place for them to make the switch.

---

[1] I did write to Scott, wondering how long he has been in SL.  No Reply.

Let’s Take It From Here
Time for the OpenSim / Viewer Community To Assume Leadership

My premise

We can no longer assume that Linden Labs will provide leadership.  We know what needs to be done, and with some Cooperation, we can take it from here.

• There are many separate pieces to the open metaverse.
• There are many competent folks involved with them.
• It is time for them to band together and chart a course forward.

It is time to go forward, and no longer worry about compatibility with the SL Viewer and server.  That being said, the key to success is to not go forward in 20 different ways.  Cooperation will be the key.

The Linux and Apache Communities are a success.  Their example can be followed in the Metaverse.

What do I mean by pieces?

Grids – Inworldz.com, Reaction Grid, and others are OpenSim based grids that are easy to get started with.  Think of them as Metaverse ISPs.

Server – OpenSim 0.7 shows promise and technical leadership (HyperGrid, MegaRegions, flexible architecture, archival capabability)  Think of OpenSim as the Apache of the Metaverse

Viewers – Emerald, Imprudence, Hippo, and others.  Think of viewers as you would think of Firefox, Safari, and Chrome.  How does the SL Viewer 2 fit in?  Think of it as Internet Explorer.

Why do we need cooperation between OpenSim and the Viewer Folks?

There are certain technologies that would be of great benefit to the Metaverse as a whole.  The shining example would be Importing 3D Meshes.  The community cant sit around and wait for an implementation to come from Linden Labs.  However, it is a major feature that requires a protocol between server (OpenSim) and Client (Emerald, et al).  The community needs to agree on protocols, so that multiple clients can take advantage of new features added to the server side.

What role should Grid Operators Play?

They need to make it easy for users to get to compatible viewers.  If the community sees great benefit in certain features (such as meshes), and the SL Viewer lags behind, then it will behoove the grid operators to a) educate and b) point users to the resources they need.  We need to get out of the mindset of having to maintain compatibility with the SL Viewer.

Does this mean that the SL Viewer may fall by the wayside if it doesnt keep up?  Yes.  So be it.

Is that pretty much it?

Yes.  Stakeholders in their respective domains need to bear one thing in mind:

Dont wait up for Linden Labs to implement what you want.  Cooperate amongst yourselves, make it happen, and try to follow the great examples of the Linux and Apache Communities.

To the Viewer and OpenSim contributors, and to the Grid Operators, I would say
this:  band together, assume technical and policy leadership, and forge a course together.  In unity, you shall thrive.

[this is a sketch about how I see a couple of trends merging to the point where something very new is created]

We are going to hit a point of blurring.  It’ll be great …

I’ve paid attention to Virtual Reality (mostly in the form of SecondLife, but a little exposure to X3D as well) for a few years now, and am about to jump into Augmented Reality …

But I can look ahead and see what’s really going to happen in a few years.  We’re going to get a lot of mixed scenarios where it’s not so clear how to classify what happens on screen:

• Virtual Reality is about synthetic worlds, where content from the real world is brought in.   A couple of examples I have personally worked on would be: bringing photos and video into SecondLife  – sometimes static, sometimes live.  I also have personally worked on objects for Vivaty that bring audio, video, and images from Flickr into a Virtual Space.
• Augmented Reality is about the Real World (as in realtime video), with data from the internet and local storage dynamically overlaid.  The classic example is to take video with a smartphone, and dynamically overlay information for nearest subway stations. places to eat, and real estate listings.

Virtual Reality and Augmented Reality can be seen as the flip side of each other.

You start with the synthetic, and bring in the real / web.  Or … you start with the real, and you bring in the web / synthetic.

For the time being, the worlds are largely going to stay that way.  It takes a lot of processing power and a lot of code that doesnt readily exist at the moment to put the two modes together.

But I can see it — think the present boundaries are going to just stay put? :)

Blended Reality will give the ability to flow seamlessly between real video with synthetic overlays, to the scenario of completely immersive environments, back to straight live video.

Scenario: I walk to Central Park and pull out a decent sized tablet (jam 4 iphones together — That’s a good screen size to envision).  I see live video, and overlays of points of interest.  I may pick a point of interest (let’s say it’s winter, so we pick the ice rink).  I get a feed in-scene of the latest photos / writeups of the ice rink, and I can segue into a virtual world recreation of it.  I meet with friends who may be elsewhere, but who are enjoying the get-together in the virtual sense.  From their point of view, perhaps they see live video of what’s happening at the ice rink.  Maybe their friends in real life have a clickable visual tag floating on them from the viewer point of view (if the RL friends choose to make their location info available to some of their contacts)  Perhaps some of the video comes from my tablet! The gist of this is, mixing modes based on location…

I’m just doing the flat-screen writeup of this for the time being.  It’s enough of a start to envision stepping into and out of virtual and augmented spaces, whether on the go with some tablet, or behind a desk.  Think not so much of overlapping hard window definitions, but of smooth transitions from between flat / augmented / virtual areas of the screen.

As I said, this is just a sketch — thoughts that have occurred to me in the last day.  The basic idea is that I dont see the hard boundaries of Virtual and Augmented Realities staying put.  They are going to blur and meet in the middle ;)

I recently marked 3 years in SecondLife, and have also been spending time using OpenSim.  I think OpenSim is the virtual world equivalent of Apache, and I think it’s going to catch on in a big way …

Backing up a bit …

So, OpenSim could be seen as an Open Source implementation of SecondLife.  But It’s more than that.  It is a platform for creating your own virtual worlds.  It’s a 3D environment where the users create content, can meet each other, and interact in real time from anywhere in the world.

Oh oh .. blank stares …  Some from tech people in the audience who think they are all done learning ;)

I recently had three encounters with a couple of friends and a recruiter, and each had the same sort of skeptical look or response to the idea of SecondLife and Virtual Worlds.  Pretty much 3 in 24 hours …

If SecondLife is off the radar for many in the tech world, OpenSim is farther still.

So let’s meander along.  I’ll explain.  I’m good at this.  I’ve been on the web since 1993, and in Virtual Worlds (VW) for three years.  Relax, this will be fun. Get your popcorn.  Keep your frikken butter off my sofa! Sheesh… some people …

In 1993 and 1994, the web was still way off the radar screen for most people.  I was at Autodesk, and I can say most managers there Simply Did Not Get The Web.  I went on to AOL to work on AOL.Com.  I made and lost a fortune, but that is not today’s topic.

So I am used to the blank stare thing.  Y’all will get the Virtual World thing… eventually.

Enter SecondLife and OpenSim.  SecondLife is a great VW platform.  It’s controlled by one company, and the server side of it is proprietary.  A few of the strong points of SecondLife are:

• immersive 3D environment
• user created content
• an economy
• great place to have meetings and trainings
• it is what you make of it
• strong creative and educational community

The client side of SecondLife is a viewer you run on your computer that gets you into the immersive 3D environment …

… and it so happens that the viewer (and its derivatives) work fine with OpenSim servers.

.. Where can an OpenSim server run?  On your Windows, Mac, or Linux machine.  Yes, you can have a self-contained virtual world on your laptop.  This is fine for some.  They’ll run a server, tell their friends how to connect, and that’s that.  Just like a private web site but in 3D.

Where it gets really interesting is to survey the publicly available grids out there (collections of one or more OpenSim instances), and to realize that companies and organizations can have their own private ones.

1994.  Apache.  Web.
2009.  OpenSim. Virtual World.

I’ve thrashed through some of the basics.  You can go to SecondLife.com and Opensimulator.org sites to get more background.

OpenSim is becoming to Virtual Worlds what Apache has been to the web.  It’s Open Source, there are brilliant people from all over the world contributing to it (echoing the development model of Linux, Apache, PHP, Perl, and some other high profile successes).

And we are at 1994 all over again.  OpenSim is at version 0.6.4, which means it is 64% of the way towards implementing the functionality found in the SecondLife server.  It looks as if it could reach 100% parity by the end of 2009.  There are already many organizations getting real results from their initial explorations (such as IBM), conducting meetings and trainings, or using a virtual space as a museum (ReactionGrid.com is recreating the 1939 Worlds Fair).

Do you think I mention meetings and trainings too easily?  Would I really do that?  Of course not.  Amada Linden did a good writeup entitled “Working in the Virtual World“.  Amanda says:

“I believe that the only good alternative to virtual meetings is a face-to-face meeting. It would be a hard to argue the teleconference calls or WebEx can create as immersive an experience”

If you want to drill deeper, see Caleb Booker’s post: Why Webcams Fail

I had known about OpenSim last year, but dismissed it as too early.  The wake up call for me was an article “OpenSimulator: The Choice for 2010” by Gwyneth Llewelyn.  She analyzed the state of OpenSim very well, but more importantly, she has a great handle on what it is going to take to succeed as a VW platform.  It got me thinking, and together with my partner Kim, we started checking it out.

Without going into 20 reasons why we personally love it, I will just paint with some broad strokes:

• SecondLife provides a great reference example
• the OpenSim developers are very capable, and there are organizations such as IBM committing real resources to the effort
• the developers of OpenSim do not feel constrained by the Linden Lab efforts
• there have already been very real advances in the OpenSim platform that cant be found in SecondLife, such as scripting at the region level, integration with skype, dynamic text on prims, arbitrary images on prims from URLs, and HyperGrid (teleport from one grid to another)

In a nutshell, OpenSim is evolving into the sort of effort we have previously seen with Linux and Apache.  A very real community is forming, and there’s even some tutorial material out there …

It’s 1994 all over again, and it makes me smile.