Banbury OpenSim Grid Offering: Cheaper than "Nebraska"

Linden Lab is still marching along to deliver the inside the firewall Second LIfe with secret code name "Nebraska" Says the SF Chronicle: "The so-called “Nebraska” version of Second Life, which is run on an institution’s own servers, will get wider testing this summer and is scheduled for general release by year’s end. The pricing for the private version has not been announced." Hard to imagine an enterprise project price lower than $150,000USD though for licencing, set-up, installation etc. Then you would need servers, some people in charge of it, so probably the project cost fully loaded is $250,000 to $500,000 per year minimum?

And what is the ROI against? Lower travel? Better teamwork? I guess we will start to see some case studies from the war machine test guinea pigs at the Undersea Nuclear Naval group or Northrop. Yes, we are killing civillians more effectively now...blah blah blah. (more of my snarky, cyncial posts about LInden Lab enterprise offerings here)

The team over at the ReactionGrid is trying to beat Linden to the market with their OpenSim spin on a server solution called Banbury . For $9kUSD  you get a stack of opensource software, speakers and a shiney red, wireless, laser mouse. More importantly you get some training and 1 year of upgrades (although I hope better than the April's fools trunk releases). 

My estimate is that with the 8 Gig solution they are offering you could/should be able to run 16-20 regions on that server. Depending on your internet speed off that box you could be able to support 300-500 concurrent users.  $9k would barely buy you 2 Second LIfe regions (33,000 prims) for a year.  So if you care about price, well this looks interesting.

We haven't been that impressed with the OpenSim voice solutions. But potentially you could run Vivox and we also are seeing some rumblings on a Skype bridge for OpenSim (ooooooo....). Billions more real minutes into the metaverse!

OpenSim isn't the same as Second LIfe (we think it has MORE potential still). Zonja has a nice piece about her views also here.

I guess for a corporate buying decision the other main factors are/should be:

1) what is the value of the solution and specific ROI. Mainly this around lowered costs and improved productivity. I still find SL to be a massive time drain for day-to-day work, but clearly it is more effective for conferences and meetings than having people travel around. But how often would it be used inside a firewall? So ROI is gonna be price/risk driven on that point. And also there is a cultural issue about having corporate avatars (which is kinda a cool thing tbh) but not every corporation is ready for furry accountants?

2) is the technology partner reliable. While the SL stack is pretty battle tested, do you really wanna sign up to LInden Lab to support a high profile project in your enterprise? They really have zero credibility with OpenSpaces, Customer Relationships, SLA's, Roadmap, Invoicing and are generally an eccentric, self-obsessed organization. Ok, so they make software and are generally nice than Oracle ;p . But is Linden Lab a serious technology partner for a serious organization...? No way. At least ReactionGrid's offering is low risk, low cost.