Vapourware Codexian Game Development Thread

[Polanski]

How should I get my game ready for Beta testing on Google Play?

The Game
Auto Armies is a premium singleplayer auto-chess roguelite for mobile devices. I have been working on it as a solo developer in between paid projects for the last year.

The Plan
My current plan is to get a beta ready for May, where I will make calls for beta testers on relevant subreddits, forums etc., where the interested will get access access the beta programme for free. My hope is that this can both help me get the game ready for release at a later point and also help build a community.

Questions to my fellow game devs
- How do you like my plan?

- Does anyone have experience with open beta testing through Google Play Store? What problems did you encounter?

- What ways do you recommend for getting in touch with the testers before and during testing? (SoMe platforms, data collection, other?)

- Do you think potential beta testers will be interested this kind of game in its current state?

Any feedback on the game or its presentation is greatly appreciated.

 

[Reinhardt]

Just downloaded bunch of rpg maker games to see what you can actually make with it. Maybe steal something for myself.

 

[TheDiceMustRoll]

Just downloaded bunch of rpg maker games to see what you can actually make with it. Maybe steal something for myself.

You can make this: https://www.youtube.com/watch?v=TgNUgeEryCM

 

[just]

anyone tried 2d tiles in godot? i'm hearing it's shit

 

[Talby]

My only problem with Godot is the load times. I just can't stand waiting for Godot.

 

[Ysaye]

anyone tried 2d tiles in godot? i'm hearing it's shit

What specifically are you hearing the issue is with tiling? I haven't used the tile map side of it much but my assessment of it is basic (the options are not like what you get for dedicated tiling software) but workable.

 

[just]

anyone tried 2d tiles in godot? i'm hearing it's shit

What specifically are you hearing the issue is with tiling? I haven't used the tile map side of it much but my assessment of it is basic (the options are not like what you get for dedicated tiling software) but workable.

standard comments like missing features, unintuitive workflow, shit that, shit this, could be better,... probably people just complaining that it doesnt have what they want
i've tried it, feels alright to me
guess more problems arise if you want something more advanced

 

[just]

it's actually really easy if you worked with godot before i dont know whats all the fuss about
set which tiles belong together and start painting
https://files.catbox.moe/kfeoo6.webm
just need a title and can release it on steam early access

 

[Reinhardt]

Looks like remake of that japanese game about horse girl racing.

 

[TheDiceMustRoll]

Anyone got a good guide for doing grid based movement in Unity3d?

 

[Angelo85]

Anyone got a good guide for doing grid based movement in Unity3d?

I don't know if these tutorials still work for the current version of Unity if you simply copy & paste the code blindly, but they should at the very least give you the right pointers. They are for turn-based/AP style systems but you should be able adapt them easily to real-time grid based movement as well with minimal programming knowledge:





/e: forgot the most important tip lol, the A* Pathfinding Project from Granberg is quite simple to integrate:
https://arongranberg.com/astar/

 

[TheDiceMustRoll]

Angelo85

Thanks!

My game is gonna be made in Unity since I've already invested a couple months into learning how to do unity (learning some more coding fundamentals atm) and I wanted to do something like this:

I really like the idea of zooming in and out in real time instead of just having the standard two options for a roguelike, and the camera being rotateable is really appealing to me. He used 3D for this, right? The creatures and player are 2D. But I like the way it looks.

 

[Nathaniel3W]

Thinking about what I'm going to do next and learning to use UE4. Years ago I toyed with the idea of a strategy game on a globe. I spent this week messing with UE4 and getting it to generate some geodesic spheres.

It was good practice and a good introduction to UE4, C++, and quaternions.

 

[cw8]

I hate quaternions to the core. If you turn them into eulerangles they become another nightmare because of gimbal lock. Still trying to get used to them. Generally, I prefer to just get a general direction to rotate to than specify a specific degree to rotate.

 

[Nathaniel3W]

Before now, I had never used quaternions. Now, I still don't really understand the math behind them, but I can use them. I first started working on a project like this years ago and I kind of figured out how to make it work with Euler angles and a lot of trigonometry. This time I forced myself to figure it out with quaternions and it's a lot more elegant: Start with these these two vectors, find the closest spherical distance between them, and place new hexagons at regular intervals as you slerp along that quaternion.

And before when I tried adjusting the rotation, it was a mess. I still think in Euler angle rotations, but I know how to apply them to quaternions. So say if you have a pentagon and it's flat against the sphere, but facing the wrong direction. If you just add 180 degrees to its yaw, there's really no telling what it will actually do. But if you take

(Pentagon->GetRotation().Quaternion() * FRotator(0, 180, 0).Quaternion()).Rotator()

That gives you the desired rotation. If I knew quaternions better, I might be able to apply rotations without the conversion, but this works pretty well for me. I just come up with the desired local rotation, turn it into a quaternion, apply it, and turn it back into a rotator.

 

[TheDiceMustRoll]

Thinking about what I'm going to do next and learning to use UE4. Years ago I toyed with the idea of a strategy game on a globe. I spent this week messing with UE4 and getting it to generate some geodesic spheres.

It was good practice and a good introduction to UE4, C++, and quaternions.

Real time strategy on a globe? Mario Galaxy meets Command and Conquer? That sounds...

 

[RobotSquirrel]

Years ago I toyed with the idea of a strategy game on a globe. I spent this week messing with UE4 and getting it to generate some geodesic spheres.

myself I abandoned this approach years ago in favor of just doing a "Cube" sphere and render texturing the map onto its UV shells. Geodesic spheres are cool but they are also really difficult to get working as most game engines are built to see the game world as flat - this is very true for nav meshes for example as they project straight down. That's how I worked around it but sadly it means the globe itself is more a cosmetic illusion than an actual game map. It looks cool but its about as useful as Civ4's Globe.

Just be aware you've got 2 different tile types, 1 for the hexagons and 1 for the pentagons at the poles. It means you'll be doubling up on assets which is why I'm not so big on Geodesic spheres. Good work though.

 

[AdolfSatan]

Real time strategy on a globe? Mario Galaxy meets Command and Conquer? That sounds...

 

[MF]

Years ago I toyed with the idea of a strategy game on a globe.

Get used to dealing with Haversine formulas in quaternion terms or come up with a way to project the sphere on something you can work with. Titan Outpost's world map is a true mathematical sphere and it's a decision I regret. It's cool, but has too many drawbacks and was probably unnecessarily difficult to program.

Pop3

Populous 3 was a jaw-dropping game, but it's not a tessellated sphere. It's just a 2D grid wrapped around a torus and rendered with spherical distortion. You can only see a tiny portion at once so it looks convincing. You can navigate using non-Euclidean geometry so you know it's fake.

There's a game called Planetary Annihilation that does a better job of getting this effect, but it has its own problems.

The biggest issue, however, is that it's just not very intuitive for players to have to deal with a realistic model a sphere. Navigation isn't trivial in real life for a reason.

 

[Hag]

Years ago I toyed with the idea of a strategy game on a globe.

Get used to dealing with Haversine formulas in quaternion terms or come up with a way to project the sphere on something you can work with. Titan Outpost's world map is a true mathematical sphere and it's a decision I regret. It's cool, but has too many drawbacks and was probably unnecessarily difficult to program.

In this case, can't you use spherical coordinate system ? You only get two coordinates, rho and phi, each with constant increment between tiles. You can easily get neighbors tiles from that.

 

[zwanzig_zwoelf]

Decided to launch the Steam page for my new game early to slowly drum up interest while working on a trailer / demo / finishing the game / etc. Especially since those in the know are well-aware of what they're getting into.

 

[Nathaniel3W]

Years ago I toyed with the idea of a strategy game on a globe.

Get used to dealing with Haversine formulas in quaternion terms or come up with a way to project the sphere on something you can work with. Titan Outpost's world map is a true mathematical sphere and it's a decision I regret. It's cool, but has too many drawbacks and was probably unnecessarily difficult to program.

In this case, can't you use spherical coordinate system ? You only get two coordinates, rho and phi, each with constant increment between tiles. You can easily get neighbors tiles from that.

I'm not sure if Unreal Engine has support for spherical coordinates. I'm not super concerned about it though now. My biggest concern for the future of a game on a sphere is handling gravity.

Years ago I toyed with the idea of a strategy game on a globe.

Get used to dealing with Haversine formulas in quaternion terms or come up with a way to project the sphere on something you can work with. Titan Outpost's world map is a true mathematical sphere and it's a decision I regret. It's cool, but has too many drawbacks and was probably unnecessarily difficult to program.

I think you probably made the right decision--anything that involves exploring a planet or moon really should use a real sphere. I don't know how Civilization has gotten away with cylinder worlds for this long.

I think my game is going to stick with A* pathfinding from one hexagon (or pentagon) to another. Instead of using distance-squared as my heuristic for deciding which way to go, I'll use angular distance through the center of the sphere. [Edit: I guess that's what Haversine is, now that I'm actually reading about it... But I think everything I need for that is already built into the quaternion functions.]

I was thinking a bit yesterday about fractal world generation, and that it's probably easier in 2 dimensions. But I think I could just generate a fractal and then look up lat/long instead of XY. The fractal will get smaller toward the poles, but those are just going to be covered by ice caps anyway. But I'm probably just getting ahead of myself now.

 

[Wysardry]

I haven't used it, but the Unreal Marketplace has a Procedural Terrain Generator that can create spheres. It's also on sale for the next few hours.

Also, there's a Hexagon Planet Generator which is also on sale, but more expensive.

 

[rusty_shackleford]

Video games are illusions, don't get hung up in the implementation part. Very few people cared that civ couldn't generate true spherical maps.

And if you want to argue about getting hung up on the details, you should know that planets aren't spheres. Earth's geometry is approximated using an ellipsoid -- more specifically, an oblate spheroid.

 

[zwanzig_zwoelf]

The Dungeon Crawler Jam is fun fun fun.