The Drax Files 35: exposing the bones of Project Bento

Project Bento is an ongoing project running as a collaboration between Linden Lab and content creators / animators to extend what is possible with the standard Second Life avatar. It’s also the subject of a series of reports and updates in this blog and, more importantly in this particular context, the subject of segment #35 of The Drax Files World Makers.

At just under 3.5 minutes in length, the segment is somewhat shorter than previous editions of World Makers, but what it lacks in time, it more than makes up for in terms of content. Anyone wishing to grasp the intent of the project, its complexity, and its collaborative nature can do no better than to sit down and watch World Makers 35.

Bento’s roots as a real project go back to around the start of 2015. The Second Life team at the Lab (which is probably larger than many people are prepared to give credit, comprising as it does product managers, engineers, designers, coders, the viewer team, etc.), periodically get together to discuss how and where they might improve Second life and offer new features.

Improvements to the avatar skeleton is one of those things that has long been requested by SL users and content creators, and thanks to work previously undertaken under-the-hood within Second Life, both within the servers and the viewer, it had, by early 2015, reached the point where it was felt it could be undertaken in a manner which would both yield a positive outcome.

Vir Linden has been the engineering lead in the project, which is by managed from a product perspective by Troy Linden and from an engineering standpoint by Oz Linden. It has over the months involved many from the SL team at the Lab including Aura Linden, Grumpity Linden, Dan Linden, Simon Linden, Rider Linden, and Coyot Linden, many of whom feature in the World Maker’s Video.

Perhaps most importantly of all, from a user’s perspective, is that from very early on, Bento featured – one might say almost driven – by content creators themselves, including Cathy Foil, Siddean Munro, Flea Bussy, Toady Nakamura, and Matrice Laville.

These are the people who focused on what bones should be thought about in order to extend the avatar skeleton, and who undertook a lot of the work testing ideas and feasibility, options for integration into tools already available to support avatar creation (such as Avastar, Mayastar), and so on. Several of the LDPW moles were also involved in the work, offering input and ideas based on their long-term experiences as content creators and developers.

Meetings for the project were held on a weekly basis (transitioning into the public Bento User Group meetings on Aditi once the work had reached to point of being available in a project viewer), and both Drax and I were invited along to witness the collaborative nature of the work between residents and Lab staff (although I admit, my attendance was sporadic, as the scheduled meeting time ended up being awkward for me to make in the latter part of 2015).

The complexity evident in Bento can be summed up in the fact that there are now 106 bones in the SL avatar skeleton, plus the original 26 collision volumes of the basic avatar, giving a total of 132 joints. These new bones include:

• 11 extra limb bones for wings, additional arms, or extra legs.
• 6 tail bones
• 30 bones in the hands (all 10 fingers!)
• 30 bones for facial expressions
• 2 other new bones in the head for animating ears or antennae
• 13 new attachment points associated with the new bones

Getting to this point alone took time and effort – and no small amount of testing. How many bones could the avatar realistically support within Second Life? What would be the data load placed on the simulator (allowing for a considerable amount of work the simulator used to do having been moved to the CDN service, something we’ll also be seeing more of in the future)? What happens when you get a lot of animated avatar appendages all operating in the same space?

These are just some of the questions which had to be addressed by those initially working on Bento, and are still being considered now in the more open Bento beta. Nor is it just a case of providing the bones and the options for animating them: there’s also the matter of ensuring the data relating to  bone movements, etc., can be reliably managed, tracked, communicated and visualised by both the simulator and the viewer.

All of which adds up to a complex project, but it is one seen as genuinely important by those at the Lab working on it, as Troy, Coyot, Aura, and Vir note in the video:

[Troy] The avatar is an extension of the resident. It’s an extension of their personality in a Very detailed way. [Coyot] Enabling  that imaginative element of ourselves is really important and not to be under-rated. People can go, “ah well, it’s fantasy, whatever,” but the imagination is an expression of who we really are. [Aura] By allowing avatars to have these new expressive modes, we can really increase this visual communication that people have. [Vir]  That’s going to increase the sense of presence in the 3D world.

As noted, collaboration between the Lab and content creators has been an important part of the initial work on Bento, and it in something which didn’t stop when the project was officially announced in December 2015. The Lab has continued to engaged with content creators and animators, seeking way in which the Bento work can be further improved and enhanced without expanding too hugely in scope (and thus risk remaining undelivered).

The primary mechanism for this is the Project Bento User Group meeting, which is held on Aditi every Thursday at 13:00 SLT, and which is supported by the Project Bento forum thread.

This continuing collaboration has already seen the Lab reconsider some initial decisions around the project, such as in the issue of bone translations and rotations, and offering the opportunity for further bones to be added to the Bento project (or even swapped / removed).

Bento still has a good way to go before it reaches the main grid. The lab is determined to make the initial phase of the work as inclusive as possible – hence the public meetings, the recent bone survey, and the  general willingness on the Lab’s part not to insist Bento must be “finished” by a specific date and thus curtail elements of the work seen as necessary for this part of the project.

Of course, there are limits to how far the project will go with this initial phase. There will not be any ability, for example, to upload completely custom avatar skeletons at this point in time. Providing such a capability within the initial Bento work is seen as something which is a step beyond offering a new avatar skeleton, and something which could add considerable complexity to the project at this point in time. But, that doesn’t mean such a capability won’t be considered down the road, once the initial work on Bento has been fully released.

While there has been plenty going on with the platform through the likes of Project shining, the HTTP work, the introduction of CDN delivery mechanisms, all of which have contributed to improving the service, they tend to be largely “invisible” to the majority of users and all too frequently overlooked.  Project Bento is a project that is very visible and could offer up one of the most extensive changes yet seen within the platform.

As such, it will hopefully not only demonstrate what can be achieved with the Lab and users working together (and long may this continue), but also stand as a positive example that when all is said and done, the Lab aren’t simply “giving up” on Second Life in favour of something like Project Sansar.