… a.k.a. Gimbal Cop v0.3!
Today (just in time for GDC), the game formerly known as “Gimbal Cop (And The Case of the Objectionable Hypercorrection)” finally loses its sticky working title, and becomes … drumroll…
Along with this spiffy new title, we’re very excited to announce the beginning of the CosmoKnots alpha on mobile!! The game is currently available on Android through Humble, and iOS users can email us to get on the TestFlight (first-come-first-serve!).
Big news covered, let’s talk a bit of backstory. Fair warning: this post became a bit of a novel (and some of it is re-telling), but this was a good opportunity to tell the whole story and explain ourselves.
There’s been a long road to bring us to this point, and a significant chunk of road ahead, but let’s talk about this new name, and what else is in store for this update.
The name CosmoKnots came from our enthusiastic and quick-witted summer intern, David “Danish” Nishball back in the summer of 2012 (my, if you put it that way it sounds like an age ago). Back then we were calling the game Gimbal Cop, but even when we chose the name, none of us were very happy about it and all intended to change it “eventually”. The original name for the project was TedBrush. Let me take you through a very brief history about what brought us through these different names, and how we landed upon what you see today.
CosmoKnots (yeah, that’s right, it’s CosmoKnots now… get used to it) can trace its origin back to our inception project, Platformer. We make games, not names. I started doing some experiments to improve the world builder tech under the hood. The problem I was trying to tackle was that our system involved taking what essentially amounted to 3D tiles and piecing together the level. Our final approach to reduce the number of draw calls would be to essentially just use the CombineMesh function to put all of those tiles into a single object, leaving behind duplicate vertices and perhaps many different material references. In fact we never even got to this point, so the number of draw calls was way too high (techie talk for “slows the game down”).
This problem led me to delve into the Unity Mesh class and read up on how exactly 3D meshes work, to try and create the mesh entirely by code, or at least be able to manipulate vertices and “stitch” things together. That approach ultimately failed but on a whim I decided to try to create what amounted to a “real-time-extruder” to just pull a shape out of the air. I thought that if we could manipulate this shape as we extruded it, that that could somehow turn into a level. I think you can already see where this is going…
That little demo soon turned into a prototype where the extruder is attached to an object flying through space, letting you trace whatever kind of shape you wanted. This was already a lot of fun, but certainly was no game. Since the experiment was originally conceived to create terrain, the name in my head was Terrain Brush. That felt a little boring, so I thought “TBrush” along the lines of “ZBrush.” Jono was quick to point out that certain users call it ZedBrush because they’re british or whatever, so the obvious conclusion, therefore, was that it was called TedBrush. We happened to be living with our friend Ted at the time (Ted still wants the game to be called TedBrush).
After a lot of blue-sky thinking I came up with the idea of a mobile game where your job is to essentially match your tilt with a tilt-trail that the other person created. Basically just wiggle the phone in weird ways, and the other person has to match as closely as possible. The multiplayer-interaction was a bit far off, but I quickly came up with a way to fake that with some randomness on a rotating controller. Getting the physics to work turned out to be a challenge and is still very sketchy in the existing versions of that game, but I was able to make the character fall on the track, added some cubes to jump over and doge, health and a score, and it was officially a game. This was in August, 2011,just a few months before the Endless Running genre took off. Oh well.
Shortly after passing this prototype around, our 3D artist, John Jankowski greeted me one morning with a grin and told me to check out an art asset that he’d made. It was a model of the then-very-popular Nyan Cat, and his intention was to put him at the front of a rainbow track. We were VERY close to releasing the game in this version at that point. I suppose it would have been called NyanBrush. For various reasons, nobody was certain that this should be our debut title, so we tried to delve into something more grand (classic noob indie mistake).
At a certain point down the “more grand” trajectory, while my mind was deep within the problem of creating a “sticky”, responsive fully-3D tilt control system, and dealing with problems of gimbal lock, I happened to have a brain wave while I read the term Gimbal Lock as an article title. At the time, our concept artist Jon Elliott was creating a Cop-and-robber theme for the game. The concept at this point was that there is a race between The Architect and The Runner (a.k.a. The Cop) and they mess with each other as The Runner attempts to run up the track to bonk The Architect. I thought that Gimbal Cop could be a cute play on Gimbal Lock and also refer to the cop character.
The sheen faded both on the name and the concept as we struggled to create fun and balanced gameplay. It was fun as the runner to dodge obstacles and catch up to the architect, but frustrating if the architect’s only power was to fly around and try to make the runner fall off the track. It was fun as the architect to grief the runner, but then he pretty much always ended up losing. We tried a lot of different variations. While Jono was working on the game code, I ended up taking a couple of deep dives into the track tech, struggling to keep mobile frame rates at a reasonable level while creating a procedural track that was detailed and interesting enough to see close-up (and had an LOD version when you saw a piece from far away). The game was too free-form. The art and gameplay lacked focus. Long story short, Jono decided to take advantage of the track tech from a completely different angle: 3D Snake.
I’ll admit that I was at first resistant to this shift. For one thing the original prototype that Jono had made was a start-from-scratch project, and I think that that gave me the impression that we were just abandoning Gimbal Cop altogether. Besides, the game was SO CLOSE to being good in its current form. For a while we flirted with the idea of a dual-mode game. At the time of Boston FIG 2013, we had a very respectable version of the game that had both snake and tunnel mode (an attempt at focusing the Archi-Runner experience) running on desktop and mobile devices. Literally while demo-ing the game to prospective fans, we realized that we essentially had two games: a multiplayer, mobile experience with an architect and a runner, and a single-player, desktop (VR-capbable, oooh) experience flying the architect through rings. We and the fans had also come to another conclusion: Snake Mode was better.
We saw that Snake Mode was already a more solid game, but Tunnel mode still needed more iterations in the lab which had no clear end in site. Eager to finally officially release a game — we decided to scope down and cut Tunnel mode for the time being and just focus on Snake for the initial alpha. We still plan to incorporate Runner gameplay, probably first as a single or multiplayer racer experience racing on tracks you (or others) have previously made in Snake mode, and after that maybe we’ll finally nail our grand vision of having multiple Runners and one or more Architects (people playing Snake mode) competing simultaneously. That endless runner game still has hopes of surfacing, but we feel that that space is already almost tapped out.
That pretty much brings us to the here and now. The most notable change, aside from swapping the title graphics, is that newfangled mobile version.
The game is running well on modern Android and iOS devices (some not so modern ones, too!), and as of today you should be able to get an APK from Humble Bundle if you buy the game now for $5+ or if you have already purchased the game. Getting your hands on the iOS version is tricky. We can’t technically charge you for the alpha, but if you get in touch with us, we can invite you to the TestFlight project. Access will be first-come-first-served. There is a limit to the number of devices we can register, and we would also like to keep some of our slots open for our own use, too, so please don’t be offended if we say “there’s no more room.” With any luck, the game will be on the App Store soon enough.
So what about that, then? When will CosmoKnots debut on Steam, Google Play, and App Store? When it’s done! But, we should be publicising our road-map soon, which should give you an idea of what’s left and when we hope to finish (sorry, no release dates). To list the goals quickly, and in no particular order, we would like mobile support (it’s currently very rough), tighter controls, more levels, multiplayer snake action (still in the design phases there), and in-game leaderboards, as well as a user-friendly web interface for scores, replays, etc. Once we’ve gotten the game to that point, we’ll be submitting it to marketplaces, contacting media outlets, and doing a for-serious promotional campaign. We would love the support of our fans in this! As always, there will be more news to come. Happy GDC everybody!