Author: parabellumdev

• 

On the weeked I bashed together a kind-of “mood” concept for the game. It’s a mix of painted and photograph, so I’m hesitant to call it anything but a mood piece. I wanted to capture the kind of atmosphere I ultimately want to evoke.

I also put the new tank art into the game, and ironed out all the headaches that come along with that. This is first implementation; the track rig is buggy, and one set of wheels are flipped, but it works.

I restored the cannon’s recoil animation, so when a vehicle fires there’s a visible jolt which is lovely. I also restored the impulse on the vehicle when it fires, so the vehicle rocks back when it’s fired.

I also added in some basic texture panning, so when the vehicle is moving, the tracks look like they’re rolling. It’s a bit jank but, it’s an incidental detail, so I think it’ll do for now.

I’ve decided to down tools on the Sabre for now, and try to get it in-game. I’m not entirely happy with it, but I’ve learned a fair bit about the process so I’ll take that as a win for now.

Firstly, I’ve realised just how rusty I am. It’s been a while since I’ve made a vehicle and it shows. However, I feel like I’ve really quickly gotten back to grips with some key things, and there are plenty of things I’ll do differently on the next vehicle.

Here’s two renders of what I’m settling for.

• 

• 

This weekend, I’m aiming to replace the current old placeholder tank with this new model. I want to get this playable, and see how it holds up in the first person perspective. I think it’s important to test it before I put any more time into it. I also want to hook this up with SFX and VFX, so I can drive it around and get a rough idea of how it should feel in the final game to drive it.

I plan to move on to the enemy vehicle next, using lessons learned from this vehicle. I might come back afterwards and iterate on this guy again after that’s done, but I’m trying not to dwell too much on it. I can come back later.

– Rob

I just wanted to post a quick update, as it’s been almost a month since my last. I’m still working on the model for the two main vehicles, it’s just I’ve been lacking motivation and trying to work out how I’m going to do things.

Mainly, I’m trying to get the interior of the cupola right, since it’ll be in front of the camera a lot. It’s an ongoing process still, a lot of experimentation and reworking. I did do a version this week, but it just looked poor in first person, so I’m not using it.

It’s a bit of a learning process at the moment. Also I’ve been playing some games to relax after work (my job) some nights instead of working on this, which I just need to do to stay sane y’know.

Apologies for the lack of updates, I’ve been modelling the “Sabre”, the main player vehicle. It’s been a difficult process, mainly getting the silhouette and shape right, and a lot of learning as to what’s important and needed on a tank.

Working on this model has raised a bunch of questions for me. It’s a very time consuming, and honestly tedious, way of making the assets. More time consuming that I thought, my estimations there were way off. I’m also not having a ton of fun with it, partly due to the time it’s taking weighing quite heavily on my mind.

I’ve had some thoughts as how to rememdy this. Work smart, not hard, after all.

I’m thinking I should make use of kit bashing far more. By that I mean, only model the general silhouette fresh each time, and use a suite of finished trims and parts to finish off the model.

It’s something I’m going to research. It’s 10.30PM (ish) here on a Sunday, no more UV unwrapping for me. My hand is cramping.

I want to put some serious thought into how I can get a vehicle out maybe each week, in this current part-time commitment. I think keeping a pace with content is more important to the project than creating the most flash graphics.

I can always come back and improve the assets once there’s an actual game to share.

Apologies for the absence, I’ve had a couple of really stressful and busy weeks, really eating up any time and energy I had for this project.

I’ve been trying to settle on a basic friendly/enemy tank design to begin with. I know these will likely change so I’m trying not to get too hung up on the details, just going for a general aesthetic and pallete.

These were the initial three sketches I did. The Sickle felt too modern, like an Abrams or Leopard 2. The Jackal looked a bit exotic, but it was interesting. The Sabre was heavily based on the Centurion and Chieftain tanks.

I expanded these out into two faction sets. This is the friendly/ally/player set, which features a few vehicles.

I wanted each vehicle to potentially have a different flavour of gameplay. The Recon car would be nimble, but lightly armed, serving more as a command car. The APC has a powerful auto-cannon for lighter enemies, but would have to manually guide it’s ATGM to the target, making for a tense pause in movement. The Tank Destroyer would be a glass cannon, making use of it’s agility to avoid shots. The MBT will be a balance of speed, firepower and armour.

I also did a set for the enemy faction. The inspiration for the enemy faction was a mixture of WWI, WWII Germany, and some of the mid-century Middle Eastern conflicts.

In Operation Desert Storm, the Coalition would encounter somewhat outdated Soviet designed tanks. Some would be upgraded with kits, others would not. It was a one sided battle; the outdated T series vehicles were no match for the Abrams and Challenger tanks they faced.

There were M4 Sherman tanks, from WWII, still serving in armies into the 1950’s, often up-gunned. Some of the last battles fought by WWII era vehicles were in the Middle East, long after that war had ended, as the vehicles were bought or gifted to developing armies.

I knew that, in gameplay, the player would be outnumbered but would overcome the enemy. I wanted the narrative to fit, so I took a bit of inspiration from Desert Storm and decided to make the enemy vehicles a little older than the friendly ones.

Also, often, an army will have such a surplus or stockpile of equipment, it will use it up before upgrading or selling it on. There was a M48/M60 Patton in Iraq, refitted to be a recovery vehicle. A Cold War era Patton, in the Gulf!

The enemy vehicles somewhat mirror the friendly ones, but with a few different approaches.

They have an artillery vehicle, firstly. Their medium vehicle, the AIFV, sports a recoilless rifle, instead of the APC’s auto-cannon and ATGM combo, making it more of a purpose built vehicle than a general support one.

The “Assault Gun” is an ambush vehicle, as opposed to the “Tank Destroyer” which is supposed to “shoot and scoot”. The enemy is on the defensive so it made a bit more sense to me.

The “Heavy Tank” is a bit of a misleading name, it is a heavy tank, but it’s an outdated vehicle, fielded only because it’s too costly to replace the vast stores they have. The friendly MBT is more than capable of destroying it, because of more modern ammunition.

This tank is actually somewhat of a homage to the classic “Tiger” or “IS” tanks of WWII. In that conflict, you had light medium and heavy tanks, with the heavier tanks literally shrugging off round after round. As we progressed into the Cold War, and tank weaponry far outpaced tank armour, the “heavy” tank became obsolete. A modern tank can usually kill another in a single shot, so it becomes more about who fires first, than heavy armour.

I dunno, I just thought it was a fun nod to history, to have the enemy using up it’s stockpile of heavy tanks in a conflict where the very concept is outdated.

Anyway, tomorrow I’ll be actually modelling these things and putting them into game. Exciting!

I’ve been trying to establish a good solid foundation for the main biome in Parabellum, which is based on kind of moorlands found in England and Scotland. As the past few posts have been very word heavy, this one will be picture heavy instead!

I put some simple audio against a short video of the biome in its current state, to mock up how the ambience should be:

The biome makes use of several ground textures, placed procedurally based on rules such as terrain slope and height. It allows for quick, precise editing and iteration, freeing me from having to paint it all by hand each time I want to redesign an area.

Same goes for grass, and ground cover. This is an early shot to try and capture mood, using a little bit of good old fog.

I’ve put a few trees together to try and populate it out, though the rendering on them is still a bit iffy, I want to go back and make this look much better.

Today I made a few large rocks to use as cover for vehicles in this wide open areas:

Here’s a shot with them present:

I think I’m going to strip back a lot of the tree cover, and have a much sparser biome for this first environment. I like the second image below the video above, the open terrain more closely resembles the environment I’m trying to imitate, and will be better suited for tank combat.

It’ll also be better performance wise, too. I think I’ve gone too heavy on quantity as I’ve developed this, and the lighting/post processing has become too heavy and off target. I’ll make adjustments to it while in this early stage, hopefully bringing it back in line with the original vision of the ‘moors.

I’m going to take a brief break from this now to focus on bringing new GUI in, and then on to a vehicle for each side. I don’t want to get locked into the environment 100% while the vehicle is still a placeholder leftover model.

Once the GUI has had a pass, and there’s a new vehicle for each side, I’ll come back to polish this, make it look much nicer, and compose it into the first map.

This weekend I got started on a basic set of assets for the “Moorland” biome, which is based on the English and Scottish highlands. The idea is to go for a wet, windy, foggy atmosphere; overcast days with gusting winds, light spatterings of rain and dew, all in a peaceful setting of rolling craggy hills.

The progress so far is a combination of procedural generation (of a sort) and hand made assets by myself. I picked up a copy of a useful tool called “World Creator” in order to build the actual terrain quickly, and using semi-accurate natural rules.

World Creator allows you to restrict ground materials using custom rules, for example, only putting rock where there’s a slope, and in high ground. Only putting mud in the lowest points, and in flat ground. Same for grass, and trees, which helps generate a true “biome” that can be shared between multiple maps.

My current priority is to improve the rendering of the trees, and add more tree varieties into the scene. Bushes and other ground cover will help break up the terrain, too. After that, I’ll need to spend time making rocks and cliffs, to insert into slopes and cliff faces to help them feel less like smooth “game” terrain and more like actual cliffs.

I’m happy with this initial step, though, it’s heading in the right direction.

No screenshot or video this time, unfortunately. I’m tired of uploading the same variation of whitebox world, so I’m going to hold off posting any media until it looks a little more like a game!

I laid the groundwork out for both the Assault and Defence mission types.

In Assault, you need to capture a strategic location, and begin with a number of tanks under your control. Once you’ve cleared the enemy out of the location, you can capture it. Once captured, enemy reinforcements will appear, and counter-attack. Destroy the reinforcements, and victory is yours. A simple spin on a simple battle, that sees you fighting emplacements and infantry at first, but then having to hold out against tanks in the second half.

In Defence, you must hold a strategic location against a number of waves of enemies. The time before the intial attack is measured in “time units” — this is your currency for placing down defences. Each defences costs a little time to prepare, so spend your time wisely. Once you’re out of time, there’s a short grace period, then a wave of enemies appears. Once defeated, you’re afforded a bit more time, and the cycle repeats. Once you’ve defeated the last wave, you win!

They’re by no means innovative or creative game modes, but they’re hopefully enough to vary up gameplay in the free play mode and create some interesting battles. They’re working, but they’re rough.

I’m calling v0.02 and v0.03 done now. I’ve adjusted my milestone tasks to make v0.04 all about taking what’s there, and solidifying it. Make it robust, make it work, make it fun. Add in HUD to visually communicate gameplay information, rather than relying on the debug log like I currently do.

v0.04 will also include an art pass, so I’ve made it a long one. I’ve given it a month. It should bring everything currently in the game up to a proper “game” standard, and pull this out of the whitebox phase. Very exciting! 🙂

I’ve been laying out the groundwork for v0.02, working on the “Conquest” (Assault) objective type. AI Emplacements were already in and working, so I was just testing out a quick AI battle to make sure nothing had broken since I last touched it all.

Emplaced anti-tank guns seemed to be invincible — no matter how many rounded slammed into them, they never seemed to die. Upon closer inspection, I noticed that their components were missing, indicating that a round had hit and destroyed that component. This should trigger an explosion, it’s how damage and destruction is handled.

Scratching my head, I did a whole lot of debugging, using breakpoints, to try and figure out what wasn’t firing. The anti-tank guns share the same “brain” as the tanks — they share a script, and use the parts that are appropriate (i.e ignore movement). The tanks were exploding fine, so the code was good.

Then I noticed down in the console, there was an error. The destruction logic was attempting to place and aligh the wreckage (blown up vehicle object) at the navigation/movement component, which the turret did not have. An oversight for sure, that was easily fixed, but I was confused at why it was skipping the rest of the code.

From what I understand, an error like that basically prematurely ends the code block. This.. makes some previous debugging headaches suddenly make a lot more sense.

Moral of the story? If you hit an error, you need to fix it, otherwise you cannot debug.

I hit a brick wall with the project. Simply put, it all became a bit overwhelming to think about, I lost track of where I was and what needed to be done, and I kept putting off working on it. It wasn’t fun, it was stressful. I constantly remind myself this is going to be a small project – not a full game, but more a proof-of-concept demo – but the reality is that doesn’t make it any easier. Game development, in general, is gruelling and difficult.

I think taking a break has really helped. I’ve returned feeling excited again, and with a clearer picture of what needs to be done.

I took out some paper, and broke everything down into milestones. I decided the first version of the game – v0.1 – should focus on a simple “free play” mode, and not attempt any kind of small campaign. I broke it down into 8 sub milestones, each covering a small chunk of what’s needed, in order. When v0.1 is complete, the idea is there should be a playable build, that I can get feedback on.

For v0.01, the goal is to have three maps, and three game modes. All three maps will be in the same biome, and heavily share assets. The three game modes I have in mind are called; Skirmish, Conquest and Invasion. Essentially, three flashy names for; deathmatch, assault, and defend.

Today I finished up the last requirement for the first sub-milestone, so “Skirmish” is working. The AI vehicles will move in, attack each other, and can be destroyed. The game recognises when the mission is over and issues victory or defeat.

A little aside; before I took a break, I had been building an objective system that I intended to be robust and friendly to use, dragging-and-dropping objectives in and linking them. I initially tried to expand on it, before I realised I was vastly over designing what should be a relatively simple system. Instead, I made a simple “Skirmish” mission script, that had enough options to allow customisation, but wasn’t so overdesigned it needed a whole bunch of nested objects inter-linked to work.

I was trying to design a perfect system, when a simple script was “good enough”. At the end of the day, it achieves the same thing, but it is vastly simpler and less maintenance.

The next sub-milestone is about getting the “Conquest” game type working. As I can re-use a lot of the Skirmish mission script for it, the focus is instead on making sure emplacements and infantry work properly, and that the friendly AI will advance on the village.