« Cory Blogs Cory | Main | A Walk in the PARC »

November 15, 2004


Siro Mfume

Physics. Real, working, physics. Collisions at precision levels better than a micrometer rather than .1 meters. Maximum speed limits more in line to real mass to energy ratios. I mean it, implement a physics book, gravity, forces... you name it we should be able to do it. Grid technology that completely overlaps from one sim to another so that there is NO lag when crossing at any speed. If this means doubling the servers to create a phantom grid for this purpose, it should be done. The ability to build above 700some meters. The ability to build/rez/place objects above 4096 meters. Lets make those space stations happen. I'll never get to space in my first life, why not my second?


Thanks to Cory for putting out this call for challenges. Holy cool new ideas Batman.

So, when challenging PARC it makes sense to pose mostly technical, code-based, software-blending puzzles. To help them frame those challenges in the best possible context, it might also be helpful to pose one super-social challenge, as SL is inherently people-based, bottom-up, a software platform now linked to by thousands of brains that have developed an attachment to this space and to the other brains. It's a highly complex mesh/ecology/economy. The particular social inhabitants of SL are quite literally part of SL. They themselves contribute changes to the environment. etc. (Are they really though? Are they?)

Some fun partially socially-oriented questions might be:

How best to to provide incentives for capital infusion while maintaing social stability?

How to provide social incentives for more complex behavior while maintaing system stability?

What are the best indicators for socio-economic activity in SL? (for example, Average Revenue Gained per Minute Played) Should these indicators be made public? Do new indicators need to be developed?

How best to demonstrate the socio-economic value of SL to both users and non-users?

Should SL view users as human resources and actively recruit value-generating individuals for its common spaces? If this is already a reality then how best to quantify the human resources?

How much control should SL users be allowed over the code on their own property? Does allowing users more control over physical and behavioral codes on their property allow for the more rapid creation of valuable SL sub-platforms?

What are the social structures that will best increase the $ value of SL? Deomcracy? Totalitarianism? Splintered rule?

Is it ethical to encourage capital infusion into all or some of SL? How much should entrepreneurship be encouraged? Business parks?

What cool new media products can be created through cordinated social interaction in SL?

When changing the form of SL, is it more important to institute direct changes of code, or to outsource the innovation in-world? What's mathematically more efficient?

word. sorry for overlong entry.

Strife Onizuka

Open the grid, make it so we can build our own servers and have them patch into the grid. Open Source. Better physics. FTP (or other) access to assets, external editor.

TTS wouldn't be all that out there. Windows and Macintosh both have TTS libraries. I'm not totally sure if Apple gives away speech recognition software but MS does. Realtime accurate language translation is a bit more complicated. And it should never be required, as many people have trouble forming thoughts in a linear fashion.

More hippos.

Jerry Paffendorf

This is great news, Cory. My 3 cents:

I would be concerned with the inevitable convergence of Keyhole-like geographic information systems (http://www.keyhole.com; note: Google just recently purchased Keyhole) and reality-based massively multi-user 3D worlds, like what Forterra Systems ((http://www.forterrainc.com) /There) is building (note: the folks at Forterra may be working on a military contract now, but they must have a broader vision of eventually building a public Metaverse on top of their geographically correct platform).

I look at Keyhole (check out the videos posted at the bottom of their site!--I recommend the $30 download) and I look at Forterra’s real world-based, to-scale geography and architecture (a very cool video demo at http://www.forterrainc.com/demo_mil.html looks just like There, but takes place in accurately reconstructed sections of the Middle East), and I can’t help but think that a primitive David Gelernter-style Mirror World (http://www.amazon.com/exec/obidos/ASIN/019507906X/qid=1100595694/sr=2-1/ref=pd_ka_b_2_1/102-4409856-0496951) isn’t so very far off. I do not want to see Second Life left out in the hyper-real cold here, potentially having too few ties to people's real lives, real world businesses, news, data and social interests! So as we look ahead, what are the ways to reconcile the Secondness in Second Life with the Firstness of meatspace? (Atoms are still thicker than bits.)

While it hasn't been the surface direction of Second Life thus far, it would be interesting to propose working with PARC on something like a “First Life Reality Integration Team” dedicated to thinking up and implementing ways to increase tracking and interaction between the digital world and the real world. Again, part of the promise that digital worlds like Second Life make is their ability to take immersive 3D environments beyond masked diversion to the level of real world extension. Is it possible for Second Life to take both roads--expressive fantasy and strict reality?

A short Second Life wish list to stir the pot in this area:

*3D printing (converting objects built in SL to wire-frame for exporting to 3D printing programs
(check out Z Corp.'s 3D printers: http://www.zcorp.com))
*3D scanning (automatically constructing a scanned 3D shape with the appropriate arrangement of prims
(check out 3D Scanners: http:www.3Dscanners.com)
*Creating animations by tracking existing video (so Philip can really have those Uma Thurman sword moves ;-)
(EyeToy (http://www.eyetoy.com/english/index.html) + Poser (http://www.curiouslabs.com/article/articleview/1156/1/281?sbss=281) ?)
*Better techniques for getting your avatar to look just like you (and ways to officially certify that your avatar does indeed look like you for better business, dating, etc.)
(something way better than this, but check out Tony Hawk Underground's "Face-In-the-Game": http://www.activision.com/microsite/thug/thug.html)
*Ways to export Second Life environments so they can be navigated in isolation from the mainland (useful for things like showing off a simulation without having to log into Second Life (check out where Adobe Atmosphere: http://www.adobe.com/products/atmosphere/main.html).


I like Strife's list. I believe the MBROLA project provides robust free TTS.

Wll this talk be open to the public?

Carnildo Greenacre

A real user interface. A minimum of head, hand, and upper body tracking, and preferably facial expression recognition as well. The SL gesture system is nice, but it's still very limited in what body language it supports.


For me a great path of investigation would be construction path questions -

I'd like to see thinking around a non-virtual building interface. 2nd Generation lego mindstorms in the SL primitive shapes which can be used to build structures importable into SL.

Jimbob Peltaire

While most technical challenges exist inside of SL, other challenges could clearly emerge out here in the physical space.

A scenario.

If we can assume that SL's platform is being copied and developed in one form or another, then we may assume the possibility of competition in the form of a console. One suggestion, likely to have already been shot down, would be to bring SL to the home so that users own the server(s) that are embedded in their console. Though SL could take a hit from existing users, this INSANE scenario would generate the grid that Siro and Strife are proposing, and more hippos.

Tiger Crossing

1) I/O plugin APIs
The ability for 3rd party developers to easily add hardware interfaces to the Second Life experience in the way of 3D visual interfaces ( http://www.stereographics.com/products/body_products.html ), 3D input devices ( http://www.essentialreality.com/p5_glove.asp ), and other future technologies. Scratch and sniff prims, anyone?

2) Advancement Paths
It's always hard to plan for the future, especially when it's so much easier to hard-code for today. But to keep an environment like Second Life going perpetually, it will need to stay with the times. Instead of having the disjointed experience of a separate 2.0 grid, then a 3.0 grid -- each time forcing the users to start over again when things change radically -- it would be to SL's advantage to have systems in place to allow for future advancements and allow old technologies to fade out gracefully. Two simple examples would be adding function overriding to LSL (so new functionality can be added by extending the parameters of exiting functions) and a pop-up at the bottom of a script's window to choose which version of the language to use. The pop-up would even allow alternate scripting languages to be added, such as a JavaScript variant or Perl. Imagine a LISP script running the AI for NPC-like objects.

3) Graphical Improvements
Related to #2, but deserving of its own mention, the graphical experience is what Second Life is all about. An IRC chatroom can do almost everything social that SL does now, but it's the physicality of the virtual environment that adds dimension and draw to the program. As it is the users that create the world, it is vital to keep up (as best you can, given the streaming nature of SL) with the state-of-the-art in computer graphics. New ways of creating structures, and shaders to add layers of reality (or unreality) are becoming commonplace in the gaming world. Can Second Life keep up?

4) Dynamic Motion
Second Life has a physics simulation at its core, one that will continue to advance in quality. It also has its scripting language that can move elements of the world around. But both are very limited, or limiting. There are so many motions that are not addressed currently by these -- motion needs to break out into other methods. An example could be the definition of spline-based paths through the edit interface or scripting (or both) such that position, rotation, velocity, and other parameters can all be pre-plotted, or even dynamically altered, all in a smooth and deterministic way. "Experience" is not adequately provided by static images. Motion is a must. (I suggest Adobe's After Effects as an excellent example of a simple paradigm that provides immense control.)

5) Static Knowledge – Data Storage
The World Wide Web is, first and foremost, a repository of static knowledge. While dynamic content is common and on the rise, most web pages are static and unchanging. They represent frozen knowledge in the form of words and pictures. In Second Life, it is virtual 3D constructions that are the static element. But text, knowledge, is not a static element there, at least not in a clean and well-integrated way. It is held at arm’s length in chat windows and “notecards” or roughly inserted into the 3D world as custom textures or tiny floating words. Likewise, dynamic data storage is almost completely absent. While the scripting language allows objects to react to events with cause-effect behaviors, more advanced behaviors will require “memory” of past events in quantity and quality surpassing what exists in Second Life today. While external data stores and processors can fill this gap, they add their own technical challenges -- most of which is passed off to the end user to deal with.

That’s all I have at the moment, off the top of my head. I’ll come back if I think of more. )

~ Tiger Crossing

There are 9 people in the world who care about this? Speak up!

Oz Spade

Great stuff mentioned so far...

Not sure how technicly feasable this would be, probably more off than you're thinking, but I've always thought it'd be cool to be able to run programs from within a virutal world, say a program from your desktop, that others could then see you using at the same time if you option to do so.

Croquet has things somewhat like this, but only works for programs running within the Squeak(sp) enviornment, which is limited, but still something.

It'd be fun to access your desktop on a laptop sitting on your desk in Second Life, completely unnessisary, but fun. It would also be a great use for confrence or showing someone how to do something, kind of like Remote Desktop. You could project your desktop on a large screen infront of a large group of people for whatever reason.

Again, not sure how technologicly far off that is to be able to display the desktop or other programs ona prim, but definitly something I'd love to see.

The comments to this entry are closed.