To be clear about where we are going I lay out the current state of technology in broad strokes. Then we will examine what one possible progression might look like. After all, we have to have a road map to begin. Once on the way, our path will likely be dictated in part by outside factors. What we can effect, we will. What we cannot, we will try to skirt. Engineering is after all, a game of compromise, a balancing act between what is feasible, and what is effective.

Today...

Today we have, and I'll cut right to the chase, the likes of SecondLife.com. An impressive, massive, world. This technology is what most consider the state of the art in virtual worlds. I couldn't agree more. I was stunned when I first saw it. Stunned by it's graphics. But also by the price tag of all the technology and infrastructure required to run it. If our world is going green, then lets be mindful of the megawatts dedicated to this behemoth and it's use.

I fully intend to make sure every aspect of this project is mindful of 10x improvements spoken of in Jeff Fox's  Thoughtful Programming and Forth article. If the entire system is examined we can see that there are opportunities here to retool the lot of it. With that in mind, here is the rough picture of today's systems and their issues.


The user platform:

This is the base of wildly varying PCs, Macs, Netbooks, Servers, switches, routers and the likes that are built into a given users location.
All this variability results in massive amounts of code and inefficiency associated with abstraction. We have bloated hardware and software that only
seems to get more complex and difficult to use. As processors grow faster, operating systems engineers freely gobble up the resources with extremely inefficient code bases. The hardware currently in homes is many orders of magnitude larger than that used by NASA to put a man on the moon. NASA learned long ago that when high density and efficiency are required Windows was not the first choice or even last choice. Resources are dumped by the ton into making these devices and spend most of their lives in landfills. Upgrades necessary to keep up with the ever larger more complex software base requires users spend billions of dollars yearly.

VR Platforms:

State of the art in VR development has progressed slowly in the last decade but has seen some interesting developments. 3D domes are cool, and super graphics engines really make for some stunning graphics. For the most part it has not become as ubiquitous as it could otherwise be. Currently the lowest cost unit is in the $200 range, and offers truly poor quality and comfort. The current high end devices cost in the neighborhood of $75,000 dollars and offer impressive quality, but are extremely bulky and uncomfortable. These systems are invariably tethered because of their power requirements. This prevents this type of interface being used in mobile applications. VR World development is at an interesting level at the moment with principal focus on graphics rendering and hobbled interaction due to a general lack of haptics. Current state of graphics rendering  is impressive. We now have the likes of the NVIDIA array core boards boasting near Terabyte bandwidths. Beautiful, but extremely power hungry. They're on the right track with array processing but the software and hardware architectures are still based on current processor advancements and this is where they fail to be candidate technology for mobile applications. All in all, apparently doomed to the executive and the basement, but not necessarily.

Hosting platforms:

Massive server farms dot the landscape and burn enough megawatts to be seen from space. It's crazy. We have a backbone based Internet system that is vulnerable to a minimum wage tractor operator. We had a major loss of communications on the west coast because some guy with an auger bit drilled into a major telecom line because he was off by a foot somewhere. Not good for businesses that measure profit in the millions per minute per mile of cable. Not good for emergency services, not good for anyone. We have billions a year added in infrastructure. The topic was chief among political pundits through the US 2008 elections. We have major duplication of systems but neither parallelism nor redundancy. The hardware required is also driven by systems and software architecture gone wrong, bloated and wasteful for the sake of time to market, and share. In today's more connected society, reliable, economical, and ecologically responsible alternatives will be required, and are not promoted in the current design. All these servers and expended energy are not going to continue to get the job done and communications industry leaders are looking to distribute communications through advanced peer to peer bandwidth sharing techniques. A rather stark admission of the need for change.

Tomorrow...

Tomorrow we will have, a system that is distributed and based on a common field programmable hardware base. Using tomorrow's array processing technology today, leaving behind the infancy of commercial software, we will forge a new future with a more sustainable technology.  To redress the system we will need to establish far more efficient and environmentally sound approaches.

The user platform:

The future user platform will employ technology already being eyed by large telecoms today. The "Cloud" as it is know is anticipated to be the place applications are loaded from and likely where your data will be stored. This would essentially mean a migration to terminal type systems. No local storage for large programs. In our plan we also do away with the heavy duty local PC and massive storage. Instead our platform will be a visor or tablet like interface. No local hard drive.

VR Platforms:

The VR Visor will be able to hold all the required data an average user would need. This would be supplemented by Thumb drives, and local solid state memory boxes. The unit will be designed to a single standard and the VROS will NOT be expected to handle anything but that. System upgrades will be controlled and be required to adhere to the standards or simply ail not be allowed in the market place. All code will be written in the most compact form using new architectures.

The hardware platform will be energy efficient and durable, making use of FOLED display technology, and new ultra low power MISC processor arrays. All will be standard. All will be made the way NASA would make it. Small, light and efficient. This will reduce resource usage in a sustainable technology. The new VR technology will employ multi-channel WiFi, and will support the EXO-NET, a parallel external mesh network consisting of all VR Visors, tablets and repeaters. This eliminates the use of a large number of resources, miles of cable, and billions of watt hours annually.

Hosting platforms:

Lets start with the massive centralized server farms, these will be just a faint memory with their massive drives, memory and miles of cabling. Getting rid of these will save billions over the years along with megawatts of power.

Then lets get rid of the backbone architecture and go wireless distributed. Today's servers are built the same way that PC and user systems are with layer after layer of software and security. Again this is all bloated and in need of pruning. Going with a decentralized wireless infrastructure prevents catastrophic failure in our systems. With such a system connected to a wireless mesh of repeaters, traffic is easily and automatically rerouted around trouble spots and congestion.

This also reduces the maintenance of millions of miles of cable. Adding mesh repeaters in newly populated areas adds bandwidth for Internet access but is otherwise unnecessary for basic communications and transactions. There is naturally the exception of mesh repeaters to access remote areas and connect larger population centers together. Either way the metro centers could easily be connected by satellite repeater, or by way of high throughput repeaters installed the length of major US highways. There is a little known maintenance network attached to the power line system that could be augmented to handle the trans-mesh traffic.

What will the future mesh look like? Lets examine the following drawing. Notice the coexistence of a commercial high throughput traditional wired network. The traditional Internet will be reserved for high speed traffic, While the majority of communications,Voice, IM, IRC, Tweets and the likes are run through the EXO-NET. This will slow the demand for Internet bandwidth, and divert much of it to the mesh, or the EXO-NET. Complimenting the EXO-NET will be local services provided by hosts that exist only on the Exo-Net. Most users won't need a PC, instead each Visor is capable oh holding the average users data and application extensions quite comfortably. For Artists, Musicians, Engineers and others that need a bit more storage or applications they can store these on local USB thumb drives. Or a larger WiFi repeater host that can accommodate a larger storage base, and provide local access to a private world or planet. These can be created for family's, groups, or small businesses that don't require linkage to the Internet for resources. Music, art, videos can all be streamed over the Exo-Net or uploaded locally into a private VR server. This server is accessed by way of a VR Visor or tablet device. No monitor, keyboard, hard drive or other peripherals will be needed.

The Exo-Net will connect to the Internet in a distributed fashion by larger service providers, and much of the existing services will remain in effect in their current incarnations. All of these changes will bring about a more efficient system and will remove many of the complexities that force the current software base to be exceedingly complex. New code will be fashioned over time that is extremely efficient and at least 100x smaller. Check the links page, and do a little reading. You'll find that an entire Forth system, compiler, interpreter etc... can be created in as little as 7k bytes. I remember my first Windows NT app, Hello World, it compiled to 1.4 Megabytes. Compare that to the same code written in Forth that takes under 100 Bytes. The savings of not carrying all that unneeded flexibility results in a 140,000:1 reduction in size. This is by no means atypical! Read Jeff Foxes Thoughtful Programming and Forth and you will understand why.

Now if we follow through and simplify, simplify, simplify, we can over time, replace the current failing infrastructure with high speed fiber, and create the EXO-NET. Below is what the system might look like.

Getting from Here to There, or How to build the bridge!

So how do we transition to this new system. Well one step at a time. I worked with a great guy by the name of Al Swain. He had a philosophy I'd like to borrow. It goes something like this; First make it work, then make it work well, then make it work cheap. In terms of developing a global information infrastructure we have something that works, we can now proceed, and re-engineer with reducing costs and improving bottlenecks in mind. We can't afford to continue to build on top of old garbage.

First Steps

Our first efforts will be centered around bringing up a full system and server. We need to establish a functional base from which to experiment.

Milestone 1 - Setting up

Software

1. Bring up Metatopia Server on an intranet. <-DONE! 1/12/10

2. Create test Metatopia world and verify minimal functionality. <-DONE! 1/19/10

3. Document the setups and configurations.<-In progress

4. Make Metatopia world accessible to outside world over the Internet.<-In progress

Hardware

1. Bring up Forth hardware development platform.<-DONE! 12/07/09

2. Setup GreenArrays hardware test bed.<-In progress

3. Duplicate Chuck Moore's current video hardware.<-In progress

4. Define preliminary VR hardware base.<-In progress

This is just to get the ball rolling! Once we have this functional base anyone recruited into the organization will have a platform to operate from. This will also give us an arena we can collaborate in. From there on out, all of the Metatopia and Metatopia Multiverse efforts will be coordinated in VR. To see where we are, or if there is some part of the Metatopia project you might be interested in helping with, go to the Software Components, Hardware Components, or Development Groups sections. Let the fun begin!