The 100th segment of The Drax Files Radio Hour went out on Christmas Day 2015, the last new podcast for the show for the year.
To make it special, the show was a live recording featuring members of the Firestorm Team and MadPea, discussing the Firestorm Gateway and the Lab’s trial gateway programme, and Oz and Vir Linden talking Project Bento.
The first part of the show focuses on the Firestorm gateway regions. These from a part of the Lab’s new gateway programme, which in a nutshell is a revamp of he community gateway programme which was operated up until September 2010. The idea is to allow groups and communities to build their own in-world and web presence, which they can use to bring new users into Second Life (including taking them through the account sign-up process), and then when they are in-world, help them with gaining familiarity with the viewer and accessing guidance and support from established users, etc.
The re-vamped programme is currently running on a trail basis, involving a number of communities and groups beside Firestorm, although they are perhaps the most visible. Their offering is quite expansive, folding-in their existing new user orientation island and their in-world support presence, both of which have been on-line since 2012, as well as providing a range of activities typical of those newcomers can find in SL, and free-to-play games, all designed to engage the new user and encourage them to return to Second Life and explore the grid as a whole.
Marketing such gateways isn’t easy; it requires a budget, and that’s something most communities don’t have. Firestorm is trying to address this by leveraging their existing user base and getting them to promote Second Life to family and friends. There is actually nothing wrote with this approach – I’ve frequently said myself that are no better ambassadors for the platform than those of us actively engaged within it. However, there are potential limits to how effective this can be over the longer periods of time, so broader-based approaches may be required down the road, but it is a good place to start.
That said, one particular advantage in leveraging existing users is that it might help further boost retention rates simply because it could led to some of those coming into SL receiving the direct support of family and friends already using the platform (although I would perhaps suggest they wait on the other side of the orientation island, and let people complete this under their own steam).
A crucial part in assessing the project will be the data that demonstrates things like throughput rates and, more importantly, retention levels; particularly when compared to the Lab’s own new user experience. The Lab, via Oz Linden, indicates this is the kind of data they’ll be presenting to gateway operators. Oz also indicates there have been some technical elements to be fully ironed-out, particularly in matters of compliance and data security, which might not have been so prevalent during the time of the “old” gateway programme.
The discussion with Oz and Vir Linden about Project Bento gets going at the 33 minute point in the podcast. This is the project, currently in open beta on Aditi and which requires the use of a project viewer (version 220.127.116.119171 at the time this article was written), which extends the avatar skeleton with a total of 92 new bones to allow for facial animation, finger movement, additional limbs and more.
One of the broader concerns raised about Bento is the potential load it may place on people’s computers. A contributing factor as to why such capabilities weren’t added sooner has perhaps been that older generations of GPUs may have struggled with the additional requirements, as Vir, as the technology lead for the project, explains in the show. That said, he also goes on to note that the Lab hopes that overall, the load won’t be too great – but this won’t be known for certain until Bento reaches a point where it is being extensively used in-world.
We did some kind of load testing with some test content rigged to additional joints and we see that the effect on load varies on whether you are using hardware or software skinning, and it varies on the particular graphics card you’re using … I can’t kind-of give one number to say what the impact is going to be, and the impact is really going to depend on what kind of content people create, right? The more joints you use, the more vertices your meshes have rigged to those joints [because] more complex content always affects performance; you make a more complicated mesh [and] it’s more work for everyone’s viewer to draw it.
So, we don’t really know the whole story on performance impact until we get this out into the world and get people testing it in groups together, so that’s one of the things; we’re going to be doing more testing on Aditi over the next few weeks [and] I hope we get the opportunity to do some kind of load testing.
But the overall numbers were pretty encouraging. We can see that adding additional joints has an effect, but it doesn’t really overwhelm the performance compared to a lot of other things, and we’re hoping we’ll continue to see those kind of results and people get a chance to pile on to it a bit more.
Above: Abramelin Wolfe of Abranimations fame has issued a viewer demonstrating the enhanced Project Bento skeleton using mocap finger data – watch the avatar’s fingers
And why Bento? Oz’s policy has always been to allow his lead engineers pick their own code-names for their projects, and Vir notes that in this case, the name many have been the result of hitting the local Japanese restaurant “a little too hard!”.
During the podcast, Drax asks if this may be a first step towards real-time facial tracking in Second Life. While there is a hint it might offer a first step, it is clear that no decision has been taken in pursuing that particular route for Second Life (although it seems likely such capabilities will be built-in to the “Project Sansar” avatar, given that is apparently an all-new, from-the-ground-up avatar; hence why it is extremely unlikely, as asked in the podcast, whether Bento could offer a means to “take” the SL avatar into “Project Sansar”).
The question did, however give Oz Linden an opportunity to explain how the Second Life team at the Lab determine the projects they carry forward.
We’re trying to be very responsive to what we think the most immediate need is, at the time we have resources available. So we’ve been doing twice-yearly get-together-and talk through – we meet with our customer support people and our operations people and the developers and the product people. And we fold-in the feedback we get from all directions, and we try to figure out what the most responsive thing we can do to meet the needs of Second Life and improve it in a dramatic way. And that’s where Bento came from; one of those meetings; and we do them a couple of times a year, and we decide on what we’re going to try to tackle next.
Bento still has a long way to go – as the Lab are first to point out. There has already been discussions on the fact that SL only officially supports bone rotations at present, and perhaps should support bone translations. Medhue Simoni has produced a video explaining why this is seen as important by animators. We can likely expect more such discussions in the New Year, hopefully with specific use cases presented to the Lab so they can see the issues and gain insight on how they might deal with them.
The show rounds-out (from the 44 minute mark) with a look at the highs and lows of Second Life’s 2015 as seen by the guests on the podcast. All told, another excellent segment, and one which leaves me looking forward to 2016 – I mean, Nick Yee!