Second Life Enterprise was a costly mistake

[Editor: A couple of years ago, Chris Ravensoft was an actual paying customer for Second Life Enteprise, a behind-the-firewall version of Second Life, designed for companies who needed a secure, private space for internal collaboration, training, and rapid prototyping. In the summer of 2010, Linden Lab shut the project down.]

How bad was Second Life Enterprise?

I have had positive experiences with virtual environments in an enterprise setting. But in this column, I’d like to talk about one project that did not work out for us.

Second Life Enterprise was a failure, in my case, for several reasons. Note that this experience is my own, maybe other beta users had a better one. Linden Labs announced the end of the experiment in August 2010.

Let’s see what really did not pan out despite the promises of the platform.

Ideal conferencing – not

Features that would really have brought it closer to the ideal conferencing tool were still far in the future.

Face capture and embedded video feed would have allowed other users to measure one’s state of mind through their facial expressions.

We were sold on the idea when we saw a cool video of Ashton Kutcher’s face mapped to this avatar’s “head” showing him interact with Second Life and expressing obvious delight at what he was seeing.


Another feature we would have liked to see was seamless integration with Microsoft Office — what a leap in the future this would have been! The only alternative available when giving a presentation using Powerpoint was to take screenshots of every slide, convert them using suitable jpg settings, then map them as textures to wall surfaces in Second Life.

Supposedly, this painful exercise was about to become a thing of the past “anytime soon.”

Not worth the cost

“Cost” can be interpreted in – at least – two different ways, both of which apply here.

First, the entry cost to simply be a part of the beta, was between $50,000 and $100,000. For the first year.

For that money, you would receive two servers and access to an online help desk. Tickets only.

Second, the ownership cost was just too high. So many prerequisites would pop up as we progressed through the setup process that I reached a point when I became quite convinced that the Linden people had never talked to anyone who works in a big company. I could not imagine a use case that their process would satisfy.

For example, customers did not get root access to their Second Life Enterprise machines — that part I can understand. However, in order to perform simple configuration, I needed to be able to get our IT department to poke a hole for incoming SSH sessions in our firewall at the most random times. The Linden gremlins would then come in through this wall and type whatever one-liner was required.

Next, software upgrades were performed by swapping the existing hard drives with new hard drives they sent in the mail. In order to perform the swap successfully, I needed to work with our IT guys to have unfettered access to the data center for two days. Alternately, I could ask for an IT guy’s time to be dedicated to this task for two days. The Linden people expected us to have someone continuously available to be on the phone with them as the upgrade went through.

>Note that, even if I did the upgrade myself, basic security rules required for IT personnel to be present in the data center at the same time as a non-IT person –me! We originally thought this would be a non-issue as the tool was being marketed as a “turnkey” solution.

Then, support tickets could go unanswered for days and, in some instances, weeks. After several weeks of this frustrating dance, I finally realized what was going on: they were using the same ticket system for both enterprise customers and public grid Second Life users. Now, I do not know what their triage policy was but considering the delays in getting any kind of reply, I suspect it was not what an enterprise customer would have expected.

I was not trying to do anything fancy. I just wanted to get the servers up — which I eventually did — and run a few grids. Some of my basic questions left them dumbfounded. Questions I would have expected other beta testers to have asked as well.

IBM Learning Commons in Second Life, a high-profile example of companies using the public grid. IBM has since moved over most of its virtual operations to OpenSim.

Finally, there was no tool was available to automatically export our existing builds on the Second Life public grid and import them to our private grid. Instead, we had to compile a full inventory of every single prim, mesh, texture, animation, and so on, and send them this list along with proof that we had all the copyrights cleared for all of them. They would then, after an indefinite period of time, send us an archive containing, hopefully, our stuff.

That’s when I started looking into the banned tools that some enterprising users had created to “steal” content and transfer it to OpenSim — or sometimes for quite shadier purposes.


An important aspect of owning our private grid was using it to provide training to our customers. We already had at least one major account asking for virtual training since it would allow us to train geographically dispersed technicians. We were also considering the potential for internal training and this was obviously the first direction we would have taken the grid’s design.

Unfortunately, educational tools such as Sloodle were only available for OpenSim, the competing free and open-source virtual world platform.

Worse, Linden Labs added a paragraph to their Second Life Enterprise agreement that stipulated that we could only use the grid internally, which meant keeping customers’ employees off the grid. I talked to the salesperson about my concern with that clause and got verbal assurance that we would not have to worry about it. The agreement was not amended, though.

That’s when I decided that it was definitely not worth it.

The aftermath

It is now obvious that Linden Lab’s proverbial heart was never really in the whole endeavor. Unfortunately, this means that the technologies I listed earlier, such as real-time face capture and seamless office integration, were never deployed in the enterprise.

Instead, we spent time and resources attempting to set up a product that did not offer the minimum set of features that would have benefitted our company. Linden Lab, likely realizing that they had waded in quicksands they knew or cared little for, threw the towel not long after I had decided to cut our losses.

An interesting take away lesson here is that, had we gone with an open-source alternative, we would now be working with a product that’s more current, was not end-of-lifed and for which plenty of support is available. And the return on investment would be, in comparison, astronomical.

(This column adapted with permission from Nexus.)'

Chris Ravenscroft

Chris Ravenscroft is an open source advocate, and founder of the Nexus technology blog. He has been building online communities for various companies since 1989.

13 Responses

  1. Revel Peters says:

    Is it okay to laugh at others pain? because I have to seriously say by the end of the article i was laughing hysterically.  I started laughing at the hard drive being sent by mail and just well never stopped..

    • I know, right? 

      I always figured that SLE was distributed the same way most commercial software is distributed — as compiled, closed-source code. I know there’s companies out there selling appliances — pre-packaged servers with the code and everything else needed pre-installed — but those are usually for the customers’ convenience. Not inconvenience!


  2. Sarge Misfit says:

    Second Life reminds me of those characters played by Don Knotts that involved him handling a gun.

  3. Thanks for mentioning sloodle. But if you want an example of an educational tool that you can’t get in Second Life, it’s not us. To date we’ve always worked primarily Second Life, and IIRC in the period in question we barely supported OpenSim.

    In future we’ll probably build mainly in OpenSim then port finished tools over to SL, but only because the creator export limitations make getting stuff out of SL such a PITA compared to doing things in the other direction.

  4. Paul Wilson says:

    Customer service: Were not satisfied until your not satisfied…

    LL really didn’t think that through properly. Basic Customer service and systems design should have picked all those problems up before they went live with anything (even a public beta).

  5. I could only contact LL’s Enterprise person through my Second Life name/email, and explicitly asked them to reply to my work email, as government correspondence is all public and subject to Freedom of Information law. The reply came to my Second Life gmail account, and that’s when I knew they weren’t serious.

  6. Various United States military entities approached Linden Lab, very early on, and a rather large contingent from the US Army Office of Simulations (Team Orlando now) was sent in, to check out the possibilities that Second Life offered. On the whole there was a lot of enthusiasm for the product. One of the results was that Linden Lab was approached, against my advice, about providing secure virtual environments. Another result was the Jessie War, but that’s another tale. Anyway, Linden Lab could have had hundreds of millions of dollars from military contracts, but their representatives expressed such disgust at the idea, disgust that maybe should have been reserved for the child molesters which they continue to provide with safe harbor. Anyway, SAIC wound up buying THERE, a totally less desireable tool, though it is quite good now, as OLIVE, and wound up making the money instead. Never let personal politics stand in the way of a better paycheck.

  7. Chris, a rebuttal from someone who has used SL Enterprise:

    “Ideal conferencing” – when exactly were you promised this? I never remember any Ashton Kutcher video – and you’d have been foolish to bought technology based on a video and not seen it actually performed on the real server before purchasing.

    Lack of Microsoft Office integration – was never promised. How can you criticize a software on a feature that was never promised? In the meantime, web-on-prim was rolled onto SLE and that did much more than being able to stare at Excel spreadsheets could have done.

    “Not worth the cost” – Bullcrap. Every other behind-firewall solution – Forterra, web.alive, Protosphere, etc, etc, for an enterprise level package, cost in the 40 – 100k+ range. Not only was the pricing competitive – but Linden Lab *SOLD OUT* their Beta target. They MET sales numbers.

    “and access to an online help desk.” – Did you never get the phone number? That’s how we handled support, and Linden Lab was fast.

    Oh, speaking of fast? SLE was LIGHTNING FAST. Rezzing, teleports, everything! With a local database server, everything was super responsive.

    “Second, the ownership cost was just too high.” because:

    “However, in order to perform simple configuration, I needed to be able
    to get our IT department to poke a hole for incoming SSH sessions in our
    firewall at the most random times.” We said no, and Linden Lab obliged. Guess you weren’t as persuasive?

    “software upgrades were performed by swapping the existing hard drives with new hard drives they sent in the mail.” – ours was done by delivering files and we did the upgrades with Linden Lab supervision. I had never heard of anything as silly as a hard drive swap. I can’t believe you agreed to that.

    “The Linden people expected us to have someone continuously available to be on the phone with them as the upgrade went through.”

    You accuse Linden Lab of never having worked with a big corporation, except you then say this. Supervising a major hardware firmware upgrade is generally a “you should have someone dedicated for that day” sort of task. And even at $200/hr billable, we’re talking $1600. Is that REALLY such a big cost for an Enterprise?

    “Then, support tickets could go unanswered for days and, in some instances, weeks.”

    Here’s where I see your first valid criticism, but you completely misread what was happening. Linden Lab decided to prematurely pull the plug on SLE, and so the support disappeared.

    “Finally, there was no tool was available to automatically export our
    existing builds on the Second Life public grid and import them to our
    private grid.”

    There was, however, it required, as you described, using another tool that checked all permissions in a sim. I’ve long since criticized Linden Lab’s inability to allow reasonable export – for example, something could be fully permissive, but if it was created by someone else, we STILL had to get permissive.

    However, I write this criticism off to problems with Second Life in general, and not SLE.

    In conclusion, SLE was priced right, was lightning fast, had good support … until Linden Lab pulled the plug. The real problem was that Linden Lab did a 180 on their support, which echoed through the company as they would, in the next year, lay off half their staff. SLE was just collateral damage in a much larger internal war at the lab. It’s sad, and THAT is what was wrong with SLE; the platform and the way it was marketed was just fine.

    •' Chris Ravenscroft says:


      First, thank you for your well researched reply. As I pointed out in my original post, your experience may differ and I’m glad that, in your case, it did.

      I suspect a major discrepancy between our experiences is that we may have been one of the very first beta testers for the product; what I relate in this post is our experience as their support for the product was still ramping up rather than dwindling down (unless I, indeed, misread the situation and support started being phased out early on). I believe that it’s one of the reasons for the hard drive dance: too much of the infrastructure was still unstable and had to be replaced wholesale during these early days.

      Regarding promises: of course, I mention the video as an anecdote. The promises were nonetheless real and made personally. You may argue that I should have demanded these features in writing and you would be correct.

      Web on prim was a nice technology for viewing web content; less so for my stated intent, which was displaying Powerpoint presentations. I may be missing how web on prim would have helped us.

      $1600: is it a big cost? No it isn’t. Is it difficult to get? Yes it is, when you buy a turnkey solution. The cost for that product is in someone’s CAPEX so, whether it’s a thousand dollars or a hundred thousand, when it’s in, it’s in.
      That meager $1600, however, is an exceptional expense that has to come out of someone’s budget. The fact that the product was not acquired by IT itself in the first place doesn’t make it any easier. Again, your mileage may vary. In smaller companies, it is typically easier for IT to be less fussy about paperwork.


  8.' KimRuferBach says:

    I liked the SLE product a lot. In fact, I was the first Solution Provider to bring a client to the SLE program.

    Yes, the process of transferring inventory from SL to SLE had bugs to work out. However, SLE itself was, as Hiro said, a fast and solid platform for my clients. The real problem was the way in which the Lab ended the SLE program, not the product itself. I believe LL’s great concern about protecting IP was what made the inventory transfer from SL such a trial, and why they didn’t allow exports from SLE. Along with potentially becoming liable for someone’s infringement if they didn’t protect the IP on SLE, one can imagine the fallout if there was infringement and SL Residents learned of it — the Lab would have had a PR nightmare on their hands, with panicked content creators concerned that their work might have been exported from SL without their permission. While it was a trial getting content from SL to SLE, as a content creator I did have to appreciate their extreme concern about protecting IP, and I imagine if the program had continued this process would have improved.

    Again, I thought SLE was a good product, and would have become a better one, and my main issue with it was the way the ending of the program was handled (which I can’t discuss here). I preferred it in many ways to working with OpenSimulator, although of course OS has come a long way since then.

  9. In a Karmic kind of way it’s good to see that Linden Lab’s failure to deal equitably with people’s wishes to get their content _out_ of Second Life – for example by giving creators an export flag along with the copy/mod/transfer permissions – seems to have hurt them as much as it hurt us.

  10.' Valiant Westland says:

    While an “Open Source” product like #OpenSim might be a superior solution to the #EPICFAIL that was Second Life Enterprise; an “Enterprise-grade” channel is the key!

    Like you Chris, I had high hopes for the Second Life Enterprise platform.  I came to Second Life hoping to find a radical, game (no pun intended) changing collaborative platform I could build profitable, next generation solutions on. 

    Instead, I discovered tremendous potential, controlled and hindered by a Geeky, OpenSource-oriented, self-absorbed (Phillip) bunch of social activist programmers, who knew ZERO about building an effective Channel of independent Developers, Consultants and VARs.  LL was/is also completely clueless about how to position / market their product; insisting on targeting complex, high-expectation Enterprise Clients or penniless consumers, when a huge SMB market could be captured; by a motivated bunch of channel evangelists.

    I don’t share the love of OpenSource software that some do.  What I want is a Second Life or similar platform that has a 1-3 user version that runs on my own system (PC or Mac… sorry, Linux doesn’t have the market share) or in the Cloud (ala and is connected to a master Virtual Commerce and Communication Grid in the Cloud, that allows me to securely “travel,” share and sell virtual goods, grid-wide.

    I’m still waiting….

    •' CandleFOREX says:

      Would you mind expaining why you are against Linux for what you have (had?) in mind?

      There is a good reason why linux dominates the webserver market and cloud market.