Games [WIP] Virtual Warfighter

Just did a new devlog, and might as well create a thread here too since its a Unity game after all. We are working hard on first person hands right now

 

New update devlog out

Oh, and a new trailer

 

One of our playtesters doing the hated Martyrdom perk from Call of duty

 

looks cool!
following

 

Our game is one of the most physics driven VR games if not the most. Since we are a shooter latency free tracking of Rigidbodies was a must (Otherwise you cant aim if the rigidbody lags behind). This we solved early on at the cost of funky physics when colliding. We have now fixed it so it works much better when colliding

 

Released a new update yesterday, containing alot of improvements and stuff implemented after feedback from our discord community. Discord is such a great tool for a early access title such as ours. One of the improvements was better controls for Rift. A huge challange is supporting both Vive and Rift users and make the experience native like. Every person has its own taste when it comes to interacting with virtual objects. So what we did was make it completely configurable how you interact with different categories of virtual items and use different presets for Vive and Rift that the user can override.

 

The community has created a subreddit, nice of them https://www.reddit.com/r/virtualwarfighter/

In the meantime we work on bots, a little teaser showing the new bots patrolling. We need to work with the navmesh and how to ground the character correctly



My girlfriend has done a butiful job rigging the hands on our new Bullpup SMG. Its was hard rigging because its a tight fit and those GI Joe hands are pretty bulky

 

A very brief teaser showing the now patrolling AI

 

Today we only have bolt action type weapons both manual (pump action and sniper rifle) and automatic rifles etc. These are all very similar since they use the same base behavior. Now i'm working on refactoring our game logic to support revolvers which dont have a action in a similar manner since its action is the drum.

Hand rigging etc is just placeholder

It's very important to take your time when doing refactoring like these to keep the code clean and maintainable

 

Continue my WIP of the revolver

Finger IK when extracting casings are nice

Its also a nice to be able to choose between single action and double action (and offcourse you can decock the hammer). Also when doing double action you will get a small penelity in that the gun will tilt just slighly upwards because of the harder pressure required on the trigger

Still to fix is a Rigidbody enabled drum and a million state bugs regarding networking all the state

 

New update out, some cool features

 

Some time since I put up something here

My brother is working on the bots, currently he is working on a system for choke points so the AI can map a path for flanking, regrouping etc, this will make the bots feel alot smarter and the gameplay will feel more dynamic.

I'm working with polish at the same time, first out is a tutorial system, here is WIP video of that

Something that I wanted to fix for a long time is command mapping so the player can map what ever buttons he like for the different fire operations. Yesterday I finally took my time and implemented this, so now you can change controls for the basic operations of the firearms. It also comes with a validation system so mappings can't conflict



We have also fixed so the player can use both A and B buttons on the Touch controller

Also something I wanted to fix for a long time is double grip, didn't get to it because it needed a good system to change the main grip hand since double grip changes both hands, at least if you want it to look good.

Wip hand rigging, but still looks pretty awsome

 

Im improving our hand IK solution, before it was hardcoded to trigger finger and read it's data from the trigger on the controller. Now I have fixed a generic solution so any finger can use IK and you can use any feeder to drive the IK not just the twigger, here is a example were I use the ratio from the hammer to drive the thumb on the hammer

 

As you can see in above video the thumb is bent in a bit of unnatural angle. I fixed this with a new mechanic that let me specify a new hand rig when the hammer is in its cocked position. We carefully mapped so the fingers on the grip does not move, but they do rotate a bit to accomodate for the new lower hand position.

What you guys think, does the new way look better?

 

A new devlog and superfun it got alot of attention on reddit sofar

https://www.reddit.com/r/Vive/comments/9x1won/virtual_warfighter_and_knuckles_ev3/

 

I don't mean to be insulting because its obvious you've put a lot of hard work in to this buuuutt...

Ugh. So boring. Shooters are a dime a dozen and require as much creativity as hammering a nail. That's just my opinion, take it with a grain of salt. Other than that it looks great!

 

Most important is that our fan base has a fun time which they seam to have

 

Quality work here, keep it up!

 

Hi guys, we've produced a very short devlog about the pathfinding / environmental awareness part of the AI development process for Virtual Warfighter. It describes how we've solved the problem of allowing AI soldiers to make different pathfinding choices, more intelligent than just take the shortest path.

We will probably expand upon this in the future by describing more, such as how to make the AI avoid exposed routes in favor of more concealed ones, if possible.

All this is part of the work being done for our upcoming coop game mode.

 

Started to work on penetration, it will be a challenge to support all scenarios, but I think it can be achieved

 

Unit tests are a godsend when doing complex algorithms, use them folks!

 

Testing above algorithm works pretty descent

 

Working on syncing animation with A pathfinding projects and our waypoints. Its important the AI comes to a natural stop when he hits a waypoint

 

Cleaning up animations

Before:

After

Much more neutral and almost no need to randomize startpoint between actors

 

Some progress with AI detection logic

 

A question for you guys, usually in most games with stealth mechanics it's possible to "cheese" the bots by taking down one bot at a time and as long as the other bots don't see this they won't react. I'm considering hindering this by having some sort of 'call-in'-mechanic, whereby certain bots will be required to call in every x minutes or the alarm is raised. A group of bots could also become suspicious if you kill x of their members and they don't see them in x amount of minutes, giving you a bit of time pressure to take down them all.

What do you think, would it make it less fun or more fun with this type of "realism"?

 

responded:
I don't like if I can't control something in the game. I mean, the bots can check the other but then there is should be a possibility to prevent the alarm (or disable it)

 

Maybe if we make it so that Easy dont have it, normal has some and hard alot. or similar.

 

yes, this will be good. Ok, it is just me but most time I am playing the games on normal. And I like if the game is responsible. Do something and see what happens. And I would going mad, if my super hero has kicked some bots, then hiding like a ninja in the woods about 10 minutes and at this time one moron will have a small talk with his friends. But as said, all people play different. You should just find what the most of them want.

 

I think a compromise that only the group that the bot is part of is alerted most of the time could work pretty well... Hopefully a map could contain many groups and some lone patrolling AIs, which would react differently to suspicious events / alerts etc.

 

Small teaser of upcoming ballistics

 

Another teaser showing off our terminal ballistics

 

Audio testing

 

As part of the new ballistics we are reworking our lag comp and the desync is now down on sub centimeter even at 100ms ping

(target is moving parallel to the shooter at constant speed for testing)

 

A little video showing how we achieved one of the highest levels of terminal ballistics in the industry

 

All of our projectiles side by side. Shotgun looks a bit tired there on the right side

 

Started to configure the colliers out in the scene. Those that have accurate colliders just works because their thickness are real world. Others need to get their thickness adjusted by configure. For example this dumpster. The collider is about 50mm, I configured it so that its instead 2 layers of 2mm metal. 2 layers of 2mm metal is alot easier to penetrate than 1 layer of 4mm metal so its important to get this right

 

Have played around with decals. About 1ms to bake the mesh. Its too much so might have to scrap the idea. Since we are on forward mesh decals are basicly the only way

But it looks really sexy