The International Association of Virtual Reality Technologies (IAVRT) is planning an alternative Internet, the Neuronet, designed to meet the needs of Virtual Reality applications. Which begs the dual questions of what those needs might be and what does Virtual Reality actually mean these days?
Indeed, I spent quite a time poking around the IAVRT website and am not that much the wiser as to how the Neuronet will actually work in reality. Sure there are plenty of pie in the sky claims and hopeful visions of an immersive VR networked future, but IAVRT would appear to be lacking in the sensible technical specifications and detailed timetable departments. Simply stating that something is easy to create because there is a lot of dark fiber available which could be used is a long way from coming up with a viable real world plan to turn science fiction into business fact.
Statements such as “IAVRT recognizes that immersive and cinematic virtual reality technologies will play a fundamental role in the evolution of human experiences in cyberspace” are easy to make, and stating “there will be a need for a governing body to set VR standards and protocols that will ensure VR safety, reliability and functionality and the VR industry's long term ability to remain self-regulated” looks awfully like someone fishing for a money maker.
But it’s the doozy of “the underlying architecture of the Internet was not designed to support the data transmission requirements of immersive or cinematic VR” that really winds me up. Sure, it’s true in as far as it goes, but then the Internet was not designed to support the requirements of the world wide web, voice over IP, online gaming or YouTube for that matter. Doesn’t seem to have stopped them from working just fine using it though.
In fact, perhaps the biggest question that remains unanswered is not the how, but rather they who which in turn leads to the why? Who would want to pay a separate subscription for access to a Virtual Reality network of the type that IAVRT is suggesting, and to do what? Why would anyone need to when the broadband Internet actually does a pretty good job of delivering streamed immersive VR anyway? Is Second Life non-immersive and slow? I don’t think so. Can the same be argued for World of Warcraft even on dial-up for goodness sake? No, it cannot because the developers of these applications have understood that they don’t need endless bandwidth in order to deliver that immersive experience over todays networks. All they need is to reduce the data that needs to be moved around to a bare minimum through the use of clever client side software and sensible server side solutions.
Given the choice, I think I would rather stick to the Internet topology than the Neuronet architecture, in as far as it has been described so far, which is very sketchily to say the least. However, IAVRT has said it will be hosted on a single central server which can then be mirrored to metro servers in disparate locations. Supposedly this is to ensure the localized high transmission speeds required to facilitate immersive VR, but there is no mention of the security and reliability weaknesses in such a model.
Similarly, while IAVRT talk about the need for a Virtual Reality Over Neuronet Protocol (VRONP) comprised of a series of ‘special languages’ there is no hint as to what these might be, who or how they might be developed. The when is ‘live in 2007’ but, to be frank, I will not be holding my breath.