So obviously this first version of the Character Customizer that we're intending to come with 3.1 won't be the final version. What will it have versus what we've talked about that it won't have in this first version?

It just won't be sculpting, that's the thing. It won't have the DNA.

We're talking about the first version.

Yes. I've talked about it before, but there's complexities because of the rig logic and because of things like we want to drive the face with Faceware, we'd like animation shared. And then even further, we want to be able to update the runtime rig code implementation, DNA as it's called, at runtime and be able to save this for your given character. So there's a lot of reasons why it's not just as simple as putting a couple blend shapes in and going between them. Because we could absolutely do this, but it's actually not going to get us to where we want to go with this system. So you're going to have head selection, you're going to have hair selection, and you're going to have eye selection, and colors between the hair and eyes. And I mean, we're even beating to death some things like, okay, the workflow right now is that we're going to have to have images for each of the hair, but is that really the scalable way to go? Shouldn't we do a render-to-texture sort of thing of the hair asset? And then, oh wait, hair assets don't render super well in a render-to-texture kind of setting, so how do we set that up? There's minutiae in here that has to be addressed. So that's why it's getting addressed on the feature team. But again, to go to Gaige, he's already set up all these assets like this, because we had to support this for Squadron 42. We had to support this for the PU and for all the designers, because we've got that design-facing tool, a Loadout Editor, because that's how they make all the NPCs.

Yeah, so technically we've had customization for at least two years now, just in different phases. What's cool about the phases too is, with our back-end setup and our internal Loadout Editor, we have all these dynamic tag setups so that when you switch a head, it'll switch the body color and it'll find all the appropriate items to fit this character. That's why we can have all these modular setups, as they all dynamically switch to each other and piece it together based on how we set it up internally.

Right. One of the interesting challenges, even for Squadron, and we can talk about how we solved it for the PU as well: in Squadron, for example, all the heads, they've got different size necks. And I think in an ATV we're going to talk about how we actually scaled that to Gaka. He's one of the cool characters within Squadron 42.

Like six feet tall or something?

Yeah, six-six or something like this, really big. Where we were going with that was the necks were actually different sizes between the heads. And this totally makes sense, right? Because a human is just different proportions. But the problem is when you have a modular system and a shirt, the shirt doesn't just stretch. It just doesn't move for it. And then you can't use the same shirt across a bunch of different guys. So that's why we came up with this tag system, so that the tier one and tier zero, really high-end actors, get custom clothes that fit them specifically. And this is going to be the case for... actually, I'm not even sure we've talked about this character, so I probably shouldn't. It's a very specific character. Nope, I can't. Won't do it. But whoever they are, they're custom. But how we solved this in the PU was actually we unified the neck shapes right down. There was a range of verts that we actually unified.

It was a longstanding discussion too. It was a big decision to start unifying things, because you start to lose some character when you unify or standardize anything. That's the trouble we've had with clothing too. You can't standardize so much that everything looks the same. So of course, did we want to standardize skulls and head shapes and necks, and how similar would everyone start to look after that?

We wanted to put the ears in the same position. There's so much detail.

So they started testing them. Vineet in the UK did some tests to figure out how far we can blend before we start to lose the character of this person. And then once we finally decided, we made that the new standard, and then all clothing up to that point on the neck would then fit the same size, and then they can do whatever they want up here.

I mean, we literally would just walk behind his desk and Gaige would be like, does this look weird? And if it jumped at you, like, this is now a different character, then we've gone too far, right? Because you've got to shape the skull, because you've got a hat you've got to put on, and some people just have giant heads.

You don't think about it, but if a hat is stretched to fit people's heads, you can't just shove the same thing on there. It doesn't work.

And to talk a tiny bit more about it, there's a lot of logic in each of these items, like he said. The head carries logic with it, because there's components on that gameplay entity class of that given head that trigger a bunch of other things on the body. So that's where the complexity comes in. And again, sometimes I get jealous of other games, because it's a difficult thing to have a system that does everything. You usually can make a system and make some concessions based on your gameplay design. If you're making an MMO, you're going to make different concessions than if you're making a single-player scripted experience. So now you see my problem. Because we need a system that, one, has to coexist, so they coexist together, and two, has to do everything.