OpenAvatar FAQ¶
Q: Does OpenAvatar lock you into using VastPark?
A: No, we designed OpenAvatar with the expectation that most users would not be using it on a VastPark deployment but could be using any of a number of platforms. It also has no lock-in at all: it can enable platforms and world owners to set up content marketplaces or even allow a distributed open marketplace for 3D content where the world owner controls or offers the avatar generation service for their users but doesn’t need to dictate the avatar body components.
Q: Is OpenAvatar a standard?
A: No, OpenAvatar is not pretending to be a standard. The SDK contains some XML, art assets, documentation and a generator that will become permissively open source. People can use it however they wish. It's there to solve the problems that every platform faces when adding avatars: how to generate them. It certainly doesn't dictate the style, skeleton, rigging, components, etc.
Q: Why would we want a mesh generator when in the years to come much of this may be handled procedurally instead?
A: While the generator we’re supplying works with Collada mesh, OpenAvatar’s specification is designed to be media agnostic. How so? Avatars are described in a simple high level XML markup and reference the media components (at this stage, Collada files) that snap together. Only the avatar generator cares that Collada files are used. You could swap the Collada generator out for a procedural-based avatar generator that processes the same XML and creates new procedural description files. But procedurally based avatar technology that doesn’t need mesh at all doesn’t exist today in 3D games and virtual world and may be years off. So we're building something that fits today's use-case (most platforms are focused on Collada being the interchange and archive format of choice). At the same time, the user’s decisions are captured in our DNA metadata file format (Based on Metaforik XML) so swapping generator technologies could be handled seamlessly for end users.
Q: You say OpenAvatar is agnostic and doesn’t lock a developer into using VastPark but no one in business gives something away free of charge: This is an attempt to create a default standard that benefits only one company, right! Somehow you've slanted this to help VastPark, right?
A: Yes you are a cynical person, but no, OpenAvatar is not slanted towards VastPark. Firstly, our very first use was for a Unity3D project that had nothing to do with VastPark. Just BTW, VastPark doesn't have the concept of an avatar. If you delve into the VastPark platform you'll find various open metadata specifications but nothing that locks customers into a certain way of doing avatars!
We're going to be using OpenAvatar ourselves because it's better than what we've seen on our platform before and because it's based in open source code that we can deploy in whatever way suits our clients. The fact that we coded the source code is nice, but think about it from our development point of view, it just adds extra work to our limited developer cycles and so I'd be very happy if someone else had done this for us. The problem was no-one did, so we used our accumulated expertise to invent something simple, reusable and open: so now we have what we wished we could have found online.
Q: I don’t like the idea of adopting anything that came from a vendor rather than a standards body
A: Our core hope is that OpenAvatar is useful. We don't think of this as an attempt to create a standard, because this isn’t a standard. If it one day becomes one and is ratified by some standards body, then we hope you'll consider it. As far as standards go, our view is that useful standards can grow our industry tremendously and it shouldn’t matter where they originated from.
Q: I love it. It’s simple, flexible and could work on any platform supporting mesh. Could it be a standard?
A: Thanks for your understanding. We hope platform developers find it useful over the next few years, because then they may want to turn the XML specification into a standard. The specification is under the open specification promise.
We are very keen on interoperability. We tend to believe that specifications that gain broad adoption are good starting points for standards and hope that this happens with the OpenAvatar specification.
Q: I’ve already got xyz specification for avatars, why would I consider looking at OpenAvatar?
A: OpenAvatar is a high level abstraction that makes it easier for others to produce avatars and components that fit your avatar specification.
Perhaps what will make OpenAvatar easier for other platforms to adopt it is that it doesn’t conflict with specific avatar specifications. An example is the Web3D consortium’s H-Anim, which could be described in OpenAvatar and then H-Anim compatible avatars would be created by the generator. That should be a solid indication of how agnostic we’ve tried to be with this approach.
Q: What’s VastPark’s strategy on other standards for Virtual Worlds?
A: VastPark intends to adopt any emergent lightweight standards for virtual worlds. Right now, Collada is about the closest thing we have to a content standard, so we’ve worked hard over the past year to change our pipeline so we now use it. If someone comes up with a standard for procedural objects like primitives, grass, etc, again, we’ll want to adopt it.
Q: What if competitors offer new specifications, will you consider them?
A: Of course! We’d love “competitors” to offer field tested battle-hardened open specifications, then all we and others need to do is try it out and adopt it if we like it and straight away it can become an industry standard by the fact that multiple vendors are adopting it.
Most larger organizations are absolutely unwilling to openly share what they’ve learned so the rest of us are left remaking the wheel over and over. Let’s not do that. Let’s get virtual worlds into mainstream usage as soon as we can.