Sean Tracy · Technical Director of Content

Have you guys talked about the player hostility stuff?

Jared Huckaby · Host, Community Manager

Only mentioned it. Just that it exists.

Sean Tracy · Technical Director of Content

Right, okay, so player hostility exists. There'll be a big write-up on it, and I don't want to try to explain it and get it wrong. Mostly it's like a reputation. It's not a very granular reputation system, but it's really meant to...

Jared Huckaby · Host, Community Manager

It's the very first iteration of it.

Sean Tracy · Technical Director of Content

Right. And again, with Star Citizen, we always want player-generated missions, because that's what gives us the emergent gameplay. So it's things like having a distress beacon and dropping a distress beacon. Now, maybe a player's actually in trouble and he drops a distress beacon. Maybe a player drops a distress beacon because he wants to bring people in. This is what I mean about emergent gameplay. So with the reputation system, if there's somebody griefing other players, they will go up in sort of a red reputation level to a point where there will be a bounty put out on them, I believe. I don't know exactly on the design side what you give the players for killing these guys, but they're basically bright red. They become a mission in themselves, basically, to go get rid of these guys. And it's not to get rid of them, but it's more of a challenge if somebody wants to put themselves up to a high reputation or a high... What's the wording we're using? I don't even know off the top of my head for the design side. But a high hostility level, basically. There'll be a mission and you can go hunt these guys down, you can group up with a couple buddies. Most MMOs have something to this effect, where a PvP player is going to be bright red. I played Lineage 2 for a number of years, and if you ever saw a bright red guy coming your way, you're going to take off and run, or you're going to try to group up with some people and try to go fight them. So that's pretty cool. It's a great system, gives a lot more gameplay, so that's fun.

There's a lot of stuff on the tech side that we fixed up for 2.2. Tons and tons of bugs that we've rectified. Some of the bigger ones are surrounding EVA, getting in and out of the ships, transitioning between different physics grids. These are all kind of unsexy things, but very, very important to be fixed.

One of the other things that we're working on: even just two days ago, we were running a bunch of our assets through a build process. We use something called a resource compiler, and it's kind of a magic black box in the build system that takes all of our source assets and converts them into a target format. So you export a model from Max, it then goes up on Perforce, the build system pulls down from Perforce, compiles the asset, and it may or may not have errors, and then it ends up in the build. Now, what you showed very nicely on your Twitter the other day was a tomato man. So tomato man is actually just replace-me balls, is what they are. It's a really simple sphere that has one texture map on it that literally says 'replace me'.

So something interesting with that is that we have wanted to convert our mesh format for a while for most of our assets. In video games and real-time rendering, we use a certain floating point. On consoles, you kind of have to use a 16-bit floating point, whereas on PC, you can allow some assets to have a 32-bit floating point. When it comes to an asset, what that means is that the position of that vertex is accurate to a 32-bit floating point versus a 16-bit floating point. And as everybody knows, if you've ever converted an analog signal to a digital signal, there's quantization that happens. And quantization means converting 32-bit down to a lower fidelity, basically. What I mean by fidelity is the accuracy of the vert. So what happens to the artists is that they'll run it through 3ds Max, they'll export it as 32-bit floating point, but the build process will then quantize it down to 16-bit. And what that can do is it'll take a mesh like this microphone and it'll crumple it just a little bit. This vert will be just accurate to here. It's like rounding, is what it's doing. So when you round a mesh, it can destroy it. And our artists spend so much time on every single vert being just right, so when they see their asset in the build, they freak out. They're not happy, for obvious reasons: hey, I made this asset, it's really gorgeous, and now you've run it through your build process. You've made it optimized, great, but you also made it ugly, or whatever has happened to it. So basically, Carsten had worked on a new mesh format that we want to support, and this is a flat integer 16. And we're running all the assets through that. But of course, anytime you do that, you end up with errors. So we have about 300... QA was super excited about that yesterday, when I said, put a Jira, put a task for every single one of those issues.

Jared Huckaby · Host, Community Manager

And Jake Ross came in this morning and saw a list of 300-some odd.

Sean Tracy · Technical Director of Content

Right. And then they tried to tell me, maybe we'll just release with that. I'm like, no. We're not going to release with that. We're going to fix the issues. And here's the thing: they're all super easy fixes. It's not that they take a long time. It's just that QA is not happy about sitting there making 300 bugs. But I timed them. It took them 35 minutes as a group to create all those. So I'm like, you guys are doing fine. Didn't take all day. So we just got to get those assets fixed now. I think the idea is that we might not get through all 300 of them by the time PTU goes live. So there may be a missing asset or two. If there's a wall not there or whatever, then that's what's going to happen if we want to release it with the asset errors now. Will we get them fixed? Absolutely. It's just how long that's going to take, I don't know, and that's up to production.