We continue the debate on the merits and demerits of Second Life. In this contribution to an IDc debate, Andreas Schiffler counterposes Second Life and Wikipedia, in terms of costing, and calls for a truly distributed alternative.

Andreas Schiffler:

I want to throw in an comment about the technology dependence of these strategies.

What has arguably worked best in the past are systems that require a minimum of technology for the individual participant – allowing them to “plug in” easily. A good example is Wikipedia. Wikipedia allowed people to contribute with “just a browser” – even the text based “lynx” browser works. What’s more, the interface was designed so that one didn’t even have the hurdle of “logging in” – just click the [Edit] button and type.

The servers and software that run Wikipedia were similarly “minimal” at the onset and only needed to be expanded when traffic grew due to the popularitly (see the info on Wikipedia’s server park and also: http://en.wikipedia.org/wiki/History_of_Wikipedia).

Going back to the strategy argument: the simple fact that a system such as “Wikipedia” can run with these relatively modest hardware requirements helps greatly to keep the system operational through individual support and donations only – thus keeping it a relatively
“corporate and ad-free” zone. Currently the Wikipedia system is run on about 100+ machines which are mostly caches. Now if we compare that to the 4000+ machines of SecondLife mentioned on previous threads – about 20 times more – it is easy to see why a system like SL is only viable in a “for-profit” scenario.

One conclusion that one can draw from this observation, is that systems operating at the high-end of technological capability such as SL are not very viable to be open (although that can change over time, as technology becomes better). This is similar to and extends the arguments about our digital divide: access to the Internet requires a certain amount of $ leaving behind the part of the world that has only 2cents. Access to a Virtual World requires requires even more $$$ further skewing the economics of “free and participatory”.

Getting back to a strategy: What has to happen to facilitate a truly open virtual world? I think is likely best done as a massively connected distributed-computing system – a fragmented amorphous “Matrix” with minimal central server requirements similar to some of the P2P networks in existence today.

I could envision an open collaborative effort where participants contribute not just give their “labor” and their presence but also some bits and bytes form their harddrive, the idle CPU cycles of their screensavers and some connectivity to provide the resources that make up
the VR in the first place. Thus what would be needed is a software that allows participants to contribute “Micro-Matrices” to the whole pool. I could see this being build out of existing OpenSource software; Linux as the base to get the hardware to go, building on networking technologies such as BitTorrent and Tor (http://tor.freehaven.net/), enabling grid
computing similar to distributed.net, adding creative tools such as Gimp and Blender, supporting existing document technologies via OpenOffice like apps, providing communications via HTML, JXTA, H.264 and Jabber protocols.

Such a software might actually challenge the “Operating System + Deskop” metaphor sold by Microsoft and Apple. What if operating a PC means actually “plugging into a virtual world” in an equally give-and-take manner. If this would take hold, it might help free the Internet from
the stranglehold of the “Asymmetric Digital Subscriber Line” (ADSL) economy: the A means that most current broadband connections are too slow for uploads, disallowing individuals to operate servers effectively from home, hence inhibiting technologies such as the ones described above.