Nov 20, 2010

Skylight: The Zombie Apocalypse

Over the summer, I heard ideas thrown around and rumors of a web based version of #SecondLife. I’ve seen some alphas tested with Unity3D that could connect to a grid in a limited fashion, but something always didn’t seem quite right about the approach.

Fast forward some months and Linden Lab is going ahead with Skylight testing, of which some of us have had the opportunity to test it out in 60 minute increments. Over the summer, I remember a Smarter Technology talk where the subject came up, and even then I had voiced concern over a web based version. At some point one of the attending Lindens had asked me why I thought it was a bad idea, and I said that it wasn’t the idea but the implementation process itself may be.

Skylight - Teleport (Zoomed)

All this customizability! Unless you’re using Skylight…

Sure, Skylight is a step in the right direction, but I have misgivings about the intention of Skylight and how it is implemented which forbids me at this time to be able to say it’s actually good in the long run.

Let’s take into consideration that it’s running on a Cloud Computing back end in order to stream the environment to the web client. In theory this is smart, because you don’t need higher end hardware to run it, but when we look closer the cost is bandwidth. Many do not have a high end bandwidth allotment, and so this is a definite detriment if Linden Lab is targeting mobile and web users.

Cloud Computing generally costs per gigabyte of bandwidth and seeing as an hour using the Skylight embed uses roughly one gigabyte of bandwidth, I can see how this is going to add up in costs as it scales up.

How would Linden Lab actually recuperate those costs, you may ask? Well, one would assume with the increased exposure and usage of the Skylight viewer, and the possibility to embed that into things like a Facebook application or web sites, that they would benefit from increased transactions that would bring. After all, Linden Lab gets a cut from marketplace, and an order of magnitude increase or more for co-current users would prove to be an additional revenue stream in Skylight.

Skylight - Unable to Purchase (Zoomed)

Who needs a multi-million dollar revenue stream? Photo: Wurlitzer Seisenbacher

I don’t believe it’s a matter of “How much should they additionally charge to use Skylight?” because that seems silly in retrospect. Aside from this, charging to use the preview system would defeat the intended purpose of attracting more people. However, there is that Marketplace which Linden Lab earns quite a lot of revenue from transaction percentages.

We could also say that something like Skylight could also be a very good option for business use in that the business could embed the Skylight viewer in their corporate site and allow employees to log in for meetings instantly without the download (which to point is a major issue with business adoption).

The problem I see is that while these options seem to make perfect sense to not only recuperate those Cloud Computing costs, but possibly turn more profit, as of this moment, the Skylight viewer is devoid of Inventory, the ability to purchase items in-world, and an LM system (not to mention quite a lot of other things).

So what does it actually do?

From what I have seen, it essentially does what I argued it would over the summer during the Smarter Technology talk – in that it’s the absolute lowest common denominator and stripped of most of the things which would both make money and appeal to new users in order to give them an idea of what the possibilities in-world are. In short, it’s a limited time use 3D chat room that uses an obscene amount of bandwidth.

Sort of like trying to convince people to buy a car by only showing them the tires.

For people who have never seen Second Life, and are entering for the first time via the Skylight viewer, they are likely to look at Second Life and think “What’s the big deal? It’s a 3D chat room…”.

What then from the content creator side of things?

Destinations on the guide will definitely benefit from the added traffic from Skylight, but those Skylight users have no L$ or Inventory, and as we know a lot of those places have Malls and stores, not to mention tipping the DJs, hosts and venue. However, with Skylight, we’re adding a lot of dead weight.

Those Skylight users will see these clubs, and locations, and not be able to purchase anything, tip the DJs, tip the venue, or the hosts. They are just dead weight, especially for the people who can’t get into the sim when the limit is reached (potentially from Skylight users).

Maybe that last point is not true, and Skylight users don’t count against the sim limits. But it’s still dead weight, regardless. As venue owners and content creators, we’re happy when there is an increase in traffic, but this is a lot like Zombie Traffic.

Dead Rising - Zombies

It’ll be like this, except with music and hobos | Photo: Dead Rising

Of course, there is also the problem of security and enforcement concerning these zombies. If you are a venue owner, then this Zombie invasion is going to be a nightmare. For a moment, imagine trying to enforce a sim ban on an unruly Skylight Zombie, and realize that Guest [Random Number] doesn’t do you any good. Of course, we wouldn’t have that problem if people could actually just log in and be accountable at least to a unique account.

Maybe, then, Skylight is just a means to an end in getting conversions to real accounts via a “real” downloaded viewer? It may be ill-conceived all around if this is the point.

Linden Lab has a lot of incentive to monetize the Skylight implementation before it goes mainstream, and a lot of benefit from doing so goes to the content creators and venue owners as well in the process. If those zombies can have an inventory and LMs, as well as the ability to purchase L$ and marketplace items, then we increase revenue by orders of magnitude (which means more money for content creators as well as Linden Lab). Of course, only if you can log-in with the Skylight viewer as a real account and not guest [Random Number].

That extra money from the micro-transactions would subsidize the costs of using the Cloud Computing back-end and maybe even turn an extra profit boost, not to mention giving those Skylight users a better idea of exactly what Second Life is about.

After all, people are more likely to be attached to their inventories than a 3D chatroom, and that’s a better route for conversion to viewer downloads and “real” accounts.

Let us take into account that Skylight technology is a perfect solution to answer things like Avaya web.alive 3D embedding, used for business quite a lot. Instead of business users needing to download a viewer, they can log in with a web viewer and attend training, seminars and meetings nearly instantly. This has been a gripe as far back as I can remember for companies using virtual worlds.

But if we’ve removed customizability and even the ability to log in with a real account, and Display Names, then most of that business aspect was just tossed out the door, not to mention any hopes of revenue increases or conversions.

While Skylight is still in testing phases, I highly recommend that Linden Lab reassess their plans with it if they intend to keep it as stripped down as it is now for a preview. There is a lot of money to be made and sidestepping that doesn’t seem to make much sense… but then maybe zombies are good for the Metaverse?


  1. On issue with the cloud rendering model for enterprise use is the bandwidth problem that you rightly raise. Unless all the attendees at a meeting are separately located, the ones in the same building are going to be using a ton of bandwidth (1.5 Mb/s each) - one of the big advantages of the normal SL client when used for meetings is the very low bandwidth requirements once the environment is cached.

  2. Too early to call until we see exactly what features will be available when this rolls out.

    If it's just what we're seeing in the testing phase? I'm with you. But I can't make the call yet. Not enough information.

  3. @Avril

    "While Skylight is still in testing phases, I highly recommend that Linden Lab reassess their plans with it if they intend to keep it as stripped down as it is now for a preview."

    I'm on the same bandwagon, actually. Not saying they *will* keep it as stripped down as it is now, but I'm hoping they have more plans for this to make it more robust than it is.

    I'm kind of worried about the "Full Access" button at the top, because it seems to indicate that they are going to intentionally keep it stripped down to bare bones in an effort to convince people to download a client and register.


    I'm not sure if the bandwidth is a fair trade off for not downloading a client. I guess time will tell as it matures.