Sandi & Chris: Checking In

Well, it's around the corner. I think probably middle of July is when it'll probably release. We're in Evocati right now. I think we'll be going to PTU in the near future. And it looks beautiful. I mean, I think we just released an Inside Star Citizen that shows off Orison, which is beautiful. And it's got the first generation of the cloud tech in that we're still continuing... I think there's a lot to work on that I think is pretty stunning and it's going to get better as time goes on, which is volumetric clouds, and it's a bit more of a challenge for what we need it for compared to, say, what you would see in Microsoft Flight Sim or some other games, because we have to deal with scales that they don't have to deal with. So we can pull all the way back and see the whole planet and have to see the cloud layers in a way that you wouldn't do, say, in Microsoft Flight Sim, and have it work as you come all the way in and fly through it. There's still some optimizations to happen and visual improvements and a lot of layering and detail on effects, but already I think it looks beautiful. It basically combines the rendering and the lighting all together with the atmospheric model. So we raymarch through all the clouds, and we'll have it on Orison first, and then as we go forward, we'll start to roll it out on some of the other planets and then start to work on dynamic weather. But that's down the road. But I think it's going to be pretty brilliant. And then Orison in itself, it's a floating cloud city. I mean, I'm sure I'm not the only sci-fi fan out there that's wanted to step foot on a floating cloud city in person. And it's beautifully done by Ian Leyland and his environment teams. They've just done an outstanding job on it. So it looks beautiful. So if you haven't seen it, you can check it out in the Inside Star Citizen, or obviously you'll be checking it out in the PTU or in the game in a few weeks, two, three weeks.

Okay, well, server meshing we're working really hard on. We have put together probably our biggest engineering strike team. So we have a whole bunch of the engine team that's working on it as well as the networking team, led by Clive. So on the engineering team we have Chris Bolte leading up server meshing work and Steve Humphries is also contributing a whole bunch to it. And we have a whole backend setup that we're doing here internally, but also some of the backend services and connections are all being engineered partly up in Montreal as well with Turbulent. And there's about 20 people, mostly engineers, working on server meshing. So it's probably our biggest overall technical initiative. It's super critical. I mean, we've talked about it before. It's the bit of tech that will allow us to open up the number of players, but also just help a lot with our performance issues that we have. Because when you play right now in the game, you play in the live build and you're running around Stanton, if everybody's in one location, it's a lot better than if people are spread all over the system, and typically what happens is as a server stays up for a longer amount of time, people spread out. You'll have people in Area 18 and ArcCorp, you'll have people in New Babbage and microTech, you'll have people in Lorville, you'll have people flying around in space, you'll have people at space stations. And so the longer the servers stay up, the more you've got different actors and different entities that are all coming around the star system. And as we build out the star system, more and more entities and objects get created. So like, we're going to have Orison coming out, and that's hundreds of thousands of new objects and entities that just come in with that. So if someone goes there, now all of a sudden, that's hundreds of thousands more entities and objects that the server has to simulate. So the key with server meshing is that this is allowing multiple servers to handle different areas inside a star system. And even later stages of it, multiple servers working together to handle even one area, an Orison or a New Babbage or something like that. And allowing players to seamlessly travel around and be passed under the control of different servers. So the authority passes from one server to another server as one goes. And that allows us to have a much larger amount of players in the same space, in the same instance. It's also incredibly critical for transitioning to new star systems. So if you go through a jump point and you go to Pyro, for instance, we need to transfer you from the server that was handling you in the Stanton system to the server that will handle you as you enter. And so a lot of the core features of server meshing we've been working on for quite a long time. So the authority transfer, there's a whole bunch of stuff that's already been built, and we're building out the backend database, what we call the entity graph, that is going to persist and remember the state of every single entity and object in the world. So if you stream out an area, you completely remember the state, where everything is, whether you dropped a Coke can in the middle of a forest on a planet. And that aspect of it opens up, obviously, the persistence we talked about, which is great, but is also a fundamental building block for the server meshing. So we've been really working hard on that. We're making really good progress on it. I don't promise any dates because it's by far the most technically challenging thing. I will say probably by the end of this year, we should hopefully have the proper persistence in there that we talk about, the proper global persistence, which is this streaming entity graph thing that I've just discussed, which is the new version of what was being called the iCache before, if people are getting confused about that. But yeah, we're working really hard on it. This year we're not doing a physical CitizenCon because it just felt a bit too early coming back from the pandemic, whether everyone could travel or not travel. I mean, even now, we're going to go to the UK next month and we're all vaccinated, but still it's going to be a pain and we've got to quarantine and do all this other stuff. So we'll probably do a physical CitizenCon the following year, but this year we're going to do a virtual one, which we didn't do last year just because we were just overwhelmed with everyone moving to work from home and everything. But this year we will do one, and one of the panels is going to be on server meshing, which will drill down much more into the technical setup and details. But I'm actually quite looking forward to Virtual CitizenCon, and I think everyone making the game loves to share what they're working on with everyone. So it's actually quite a morale boost for the teams. And it's nice when you do it in person because you get to meet people in person and you can see how much people really love and enjoy the work that everyone's working on. But the next best thing is to do a virtual one. So that will be later this year in October. And I'm looking forward to that and hopefully you guys will like it.

Well, I mean, outside of the meshing and the persistence, there's some really cool features that are coming online. Actually, the next patch, so 3.15, I believe it hasn't aired just yet. I don't know when we have this thing edited. So it may have come out in a Calling All Devs or not by the time you see this. But Richard Tyrer is going to go over the personal inventory and the medical gameplay, which is a big thing for 3.15. I mean, first of all, the personal inventory is another part of that proper persistence and the world behaving like it's meant to, where you don't have this infinite bag of holding where you have every single armor piece and weapon and cloth and you can access them all and equip all of them. That's not realistic if you're running around out in the Sahara Desert, you don't have access to all the stuff, you just have what you can carry on you. So personal inventory is going to be about what you can carry and how you can equip that, whether you have something in a backpack, whether you have something on your suit, and it'll be just limited to what you can use and carry. It's also coming with a whole new way to equip your character, because to be honest with you, the PMA is horrendous. It's the most counterintuitive...

PMA?

Yeah, it's a personal management app. It's like how you currently put clothes on, but anyway, it wasn't done properly all those years ago, and then we were always like, we're going to fix it up, and haven't got around to it, but we have a whole new way of equipping and doing it, which is much more a paper doll drag and drop setup, and it's going to be far, far more intuitive, far better, and will allow you to properly behave like, okay, I'm going to wear my civilian clothes now, okay, I'm going to take that off and now I'm going to put on my armor, but it will be limited to what you have around you, whether it's in your hab or in your backpack or you're right next to it. And it will make your behavior be much more like it would be in real life. So that's a big deal. And the other big deal is the medical gameplay, which is really the first stage of the death of a spaceman, which I talked about many, many years ago. And that involves allowing players the ability to fix up and heal people at varying levels of damage, more so than we can do right now with a medpen. Also revive people depending on the severity of injuries. And then it involves also the respawn mechanics. So if you die, you get brought back to life in a hospital bed. Currently, there's been a slight tweak to a bit of the lore, because I saw some people get quite upset about the fact that we mentioned that the respawn in a place away is now being covered by the fiction of you've recorded your DNA and you're cloning. But the bottom line is if you got atomized in an explosion on a planet somewhere, how are you finding that body to recreate the person? That just doesn't actually make much logical sense. So the kind of concept we would have is that you can only come back to life so many times, which is basically you've taken a little bit of your DNA and you can grow a clone, but you can only do that three or four times. It's like having three or four lives. After that, your DNA's degraded too much and you're basically permanently dead. Your next of kin will take over in the game, which is the death of a spaceman mechanic. So it's exactly the same as the death of a spaceman mechanic that we had before, which was you had only so many lives before you would have to pass on to your next of kin. But just scientifically working, I mean, it's all hand-wavy at this point, but it just logically makes a bit more sense and also allows players a choice if they don't want to wait around to see if someone can rescue them or find them. Because it may not be dead if you die on a planet, or you freeze to death on a planet and no one's around to come get you. That's not a fun game experience. You may want to just respawn back in New Babbage. But anyway, so it's the first stage of that. We delayed it a little bit. We could have pushed it to get it into 3.14. We wanted to hold it to 3.15, mostly because it's to do with the hospitals that you respawn in. If we had it in 3.14, we would only have two locations that would have hospitals, which would have been Orison and New Babbage. With 3.15, we also get all the space stations and Grim HEX. And the space stations are in orbit around every one of the planets. So even if we haven't got the Area 18 one or the Lorville one, Maria Pure of Heart, done just yet, they'll be in, I believe, the subsequent release, maybe the 3.16 one. I'm not sure, I'd have to take a look at the release schedule. But now we've got a place that no matter where you are, you would basically respawn where your home base is. So that was one of the biggest reasons. And just to give a little bit more time to polish things like the UI and the experience. And that's kind of one of the focuses that we'll have going forward. It's a bit of a shift, but to try to deliver more final or more polished features rather than a very basic version that you then leave for a while before you come back to it. So those are big things. The persistence, which I mentioned at the end of the year, the proper full persistence of being able to remember where all these items are, which opens up having your own hab that you can store things in, your own hangar. So we're going to have persistent hangars and persistent habs too. I don't think we've put those on the schedule just yet. They will be there shortly after persistence, but that's a big deal because then storage space and where you keep things and where you store things will be huge. And I think that will radically change the game experience that people will have. So those are all very cool stuff. We're also working on the proper iteration of the cargo stuff. So a whole revised cargo system that will be just much more physical and practical and ties into the object and item persistence that I've been talking about. So for the traders out there, it will be a lot better. And I swear to God, the AI guys are working on some really good, cool stuff. And I'm determined to make sure that in the live game server you actually see it, because really there's a lot of problems that come because of the low server tick rate. It's really quite frustrating because in reviews I see fun behaviors, a lot more behaviors than are actually in the PU right now. And it's a lot smarter, but you get into the live game like I did with my nine-year-old daughter last night, and there's just a lot of people standing on tables doing nothing.

You got reamed last night.

Yeah, I got told off, I was told I should be embarrassed to have all my AI standing on tables. But yeah, because I mean, the AI is actually very powerful and there's a lot of really interesting stuff that can be done. It's more a matter of the server being able to run fast enough to handle that. And then there's some synchronization things that we have to fix. The AI system, like the streaming in and out of stuff, the object container streaming, which is both on the server now and on the local, but the server one's the problematic one because that has to remember state when it comes in and out, and that's one of the reasons why the persistence that I talked about there, the entity graph, is really important, because you really need to remember all the state properly. And right now, we don't remember the state as well as we need to, and we do need it for server meshing, but that is one of the reasons why you see AI problems out there, is because they were streamed out in a state like sitting on a bench or doing something. And then when it gets streamed in, it hasn't remembered that state properly, and that's how they stand on a table or T-pose. Because we're always streaming in and out stuff as you're moving around the world. And the AI, if we were not streaming stuff in, you wouldn't see a lot of the problems you see. But unfortunately, we have to stream things in and out because we can't keep everything in memory all the time, so the global persistence will also help a lot with that. So that's a very short answer.

Well, I'm finishing up something that I started, I don't know, last year or something. I mean, I don't have that much time to really do all the coding, but I'm trying to finish up what we call physical skins. So all our characters in the game, the physics that determines whether you collide is defined on the skeleton. So the skeleton has a physical joint for like the arm and the hand and the head and the legs, but it's the same skeleton and the same physics geometry for every actor in the game, whether you've got heavy bulky armor on or you're just wearing thin clothes. And actually most games are like this. In fact, a lot of games just have a simple capsule that defines a character. But all our characters, when you see a helmet come on or they put the armor on... And that's what's called the skin, which is just geometry that gets mapped to the joints of the skeleton. And then as the joints move around, the geometry deforms with it. The change is to enable the skins to have physics information embedded with them. So when you bring in the armor torso piece of a heavy armor, it actually brings in physics geometry that basically represents its shape. So as you put different things on, your physics... And that's really important because that means, first of all, better collision and hit detection. So like, you could have a case right now where you have a big bulky helmet on, but the physics geometry is just the head size. And so a bullet could look like it's hitting the helmet, but it doesn't actually hit the physics geometry. Well, we can change that now. So the geometry is the same size as the helmet. And also it makes a big difference for things like physical interactions, like with cloth. So if I walk through a hanging dangly plastic, if I've got my proper shape, then it'll move out of the way where the visual render mesh is. And the same for if I have a cloak on, it will now drape over me properly. Because right now if we put a cloak on the back of a character with a normal skeleton and they had heavy armor, the cloak would look like it's going through part of the heavy armor, because the actual base skeleton physics geometry is smaller than the heavy armor.

So you're saying you're making it more specific?

Yes, it's essentially just making it more accurate. It's allowing for better simulation. It's necessary for cloth and hair simulation, so you can make sure that the hair is not going through what looks like the head because the physics geometry of the head is smaller than the actual render. So I've been working on that for quite a while. It's been something I've been doing in the background. And yeah, it's almost ready to go in, probably not this release or the next one, but the one after that. They don't put my code in right away. We need to make sure it gets tested so it doesn't blow everything up.

Yes, so when we get back to the office, which is probably going to be a little later on this year, we as a family are going to move over to the UK for some time to be with the Squadron team to get it finished. So instead of being here in sunny LA, we're going to be in...

Sunny Manchester.

Sunny Manchester, yeah. I mean, with global warming, everywhere's turning into California anyway. But yeah, so whether we're doing it at the end of this year or the beginning of next year, we're definitely over there. The UK office, we're going to be moving to new offices in the UK next year, which will be good because our current offices are not big enough for how many people we've got in Manchester now. We've grown up, but we've got some new offices that we'll be moving into next year in Germany because our lease was up there, and we've got a great new space. It's very close to where the old one was. And it's going to be a very cool place in Manchester, and the new offices will hopefully be the new flagship for the CIG offices, and we'll have about 400 people there. And anyway, that'll be fun. And I'll be there with everybody because that's where the core of the Squadron team is. And so I'm getting over there. I'm going to be there on a day-to-day basis with them getting Squadron finished out for you guys.