Announcing novaWARS, a mobile game for gamers that brings console-quality gameplay to the touchscreen.
Since getting my first smart-phone, I imagined a future of awesome portable, console-like games. After all, these things have way more power than an Amiga or SNES or NeoGeo that enjoyed such greats as Star Control. Instead we got Flappy Bird and Cookie Clicker. I, like so many other gamers, was so disappointed at the lack of the types of games I enjoyed. The problem, as they say, is that control on touch screens isn’t accurate enough for arcade gameplay - it just can’t be done…
So I invented a solution and made novaWARS! Over six months has been spent on just the controls, refining and improving, with no release until they were ready. The ambition is to give gamers what they want, and inspire a new branch of mobile games that shun the current conventions of in-game consumable driven tappers to offer quality gaming experiences as we’ve enjoyed on our portable consoles.
It’s currently in alpha testing with a few Android and Windows players and they are reporting that it plays well. . If you want to join in, drop me a line. Open testing will be happening very soon!
Have a few refinements to do prior to public release. One of the concerns seems to be my advanced controls ‘moving’ throws of players and makes them uncomfortable. Also there are issues about tutorials. I go to the effort of creating tutorial videos but you can’t rely on people to watch them!
Couple of tests with non-gamers reminds in a big way how simple and easy your game has to be in the first play. They have no skills! Intro must be incredibly gentle and rewarding for even the most basic progress in learning the game.
Made the change to wraparound space. I always wanted this as it’s how Star Control was, but I thought the physics would be an epic fail due to the wrapping of cloned objects on the screen…until I realised that I only have one ship in the centre and not two ships on screen at once like SC! So just move the objects in LateUpdate() when they get too far from centre. Much nicer for new players unfamiliar with arcade action.
Have to create tiled nebulae now, so came up with this workflow using tiled textures from the 'net:
Quite simply, extract the noise from a texture as a greyscale mask and apply that to a coloured layer. Layer with ‘Screen’ blending into the final tile. There’s a wealth of interesting source graphics out there. Initially I used cloud texture, but then hit upon marble with its swirls, and noise textures. Rotating and changing the brightness/contrast tweaks creates multiple variations from each texture, so you don’t need many to be able to create a wide range of space nebulae.
Fixed a pathfinding fault quite nicely. Using PolyNav for navigation (nice asset!), it’s not really designed for dynamic maps which wraparound maps requires, repositioning the elements when they get too far from the ship. Thankfully the navigation geometry is so simple that it doesn’t impact performance, and I can call manual updates whenever the ship has travelled far enough to require one.
Daily blog - first impressions only go skin deep; need more pretties!
July 10, 2017
With the latest Early Access build rolling out featuring notable gameplay and control changes, I’ve realised that my game just isn’t pretty enough to garner attention. I grabbed Element4l on PlayStation Vita yesterday and on a mobile processor, it’s managing some pretty tasty post effects that give it a beautiful sense of light and shade.
Lighting is the primary language of visual art of these sorts, and my lighting is too plain and functional. I’ve experimented a little with pixel-based ‘volumetric’ lighting to cast shadows and think that’ll work well in my nebulae, giving a sense of dust scattering the light. Maybe. Thankfully the prototype on the Nexus 7 didn’t slow to crawl like the other post effects.
If I can’t get something that produces beautiful screenshots though, even if only the PC flavour, I’m going to have an uphill struggle with marketing. The unfortunate reality is, ‘skin deep’ is as far as you get with first impressions. It’s the pretty screenshots and videos that get the attention, and not the list of features or fabulous underlying technologies.
Created a 2D reactive lens flare, as Unity’s in built flares only respond to 3D colliders.
Creating the volumetric lighting is proving tricky to do efficiently. Getting alpha-channelled normal-mapped sprites to render onto a rendertexture isn’t happening. My reference asset, one of the numerous unsupported sprite-based lighting assets on the store, has separate sprites for shadow casters rather than drawing objects onto a rendertexture and using a shader to draw them as black. Maybe that’s because you can’t draw sprites onto RTs??
Development has been taking a hammering with car troubles. :( Plus finding info on some of the rarer systems of Unity like rendertextures and custom post effects is very hard, making the actual implementation of my volumetric lighting way trickier than I expected it to be! In the end I couldn’t get alpha blending to work on rendertextures in a shader, so I use a brightness-based mask and multiply blending.
I need to do some experiments to optimise the drawing of the shadows, to see if blitting with a material is faster than drawing with a camera and quads. By jittering the shadows and using multiple passes, I can get the drawing down from O(n) to O(log n) and improve quality. Plus need a simple blur filter and temporal accumulation to smooth things out. But I need to get gameplay tweaked for public Early Access release! Three weeks behind my thrice postponed release.
. If you want to join in, drop me a line. Open testing will be happening very soon!
Plus finding info on some of the rarer systems of Unity like rendertextures and custom post effects is very hard, making the actual implementation of my volumetric lighting way trickier than I expected it to be! In the end I couldn’t get alpha blending to work on rendertextures in a shader, so I use a brightness-based mask and multiply blending.
