Chrome Abyss: Glint

Glint is a single-player, push-your-luck puzzle game for mobile with a community leaderboard, daily challenges and competitions. The goal is to collect alien gems by banking combinations on a board.

Comment

Inspiration

Close to a year ago I started designing Chrome Abyss: Hybrid Uprising — a sci-fi tabletop RPG built on deck-building and Farkle-inspired dice mechanics mixed with D&D dice, where seven Archetypes battle to conquer the galaxy. What began as a card game opened a can of worms I hadn't planned for: a full worldbuilding exercise. Out Chrome Abyss universe needed to feel real. It needed a story. It needed a history. I embraced it and got happily lost. Seven species, seven worlds, factions, lore, and the connective tissue between them all. Including centuries of history and a plot for something that might some day turn into a graphical novel or book.

To playtest Hybrid Uprising I built a digital simulator, which spun off a simpler card game, Chrome Abyss: Squads, with a digital playtest version of its own. Somewhere in that process two things happened at once: I realised I could actually build digital games much more easily and effectively than I imagined, and our worldbuilding had already sketched a pub game played across the galaxy.

At the centre of our universe sits the Neon Bazaar on Fringe Market, the one place where every species meets in peace. They trade, but not in money; money is worthless in the Chrome Abyss. Value lives in tech, augments, chips, and the minerals that make them. And in every corner of the Bazaar, they gamble those minerals on the house game: Glint.

I had half an idea living in the Chrome Abyss lore and a new itch to build something digital-first. Putting them together was inevitable. Chrome Abyss: Glint is officially the third game in our universe, yet it is the first one to go live.

What it does

Glint is a push-your-luck combo puzzle. You place alien gems on a hexagonal board, activate combos (matched sets and straights), and decide every single turn whether to bank your points or push for a bigger multiplier. Overreach and you bust. As the board empties, the Abyss collapses it, twice, into GLINT RUSH, a frantic final round where any combo banks instantly and you can cash out at any moment. The goal is simple: clear the board and bank as many points as you can.

Every bank resolves in one breath:

$$\text{bank} = \left(\sum \text{combos}\right) \times \text{multiplier} + \text{chain bonus}$$

In the local playtest version, I hadn't yet built a community features. Then this competition came along and I realised the capabilities Reddit opens up from a community perspective could be further explored and integrated now. On Reddit, Glint becomes a community sport. The Community Daily hands the whole subreddit the same board every day (one shared seed, midnight UTC rollover), ranked on a rotating challenge — Highest score anchors Mon/Wed/Fri; Highest single bank, Most Nebulite refined, Most Nebulite banked, Most banks in one game and Most chains banked cycle through the rest. One board, six ways to play it. Your best run lands on the community leaderboard under your Reddit name. No signups and no downloads required. Around that core: a 101-level campaign, daily challenge objectives, milestones, achievements that unlock hidden ability gems, and a collection economy fed by the Nebulite you bank.

How we built it

The game is React + TypeScript + Vite. The heart of it is a pure, deterministic, seeded engine — a framework-free module that never touches the UI, validated by a test suite that includes multi-hundred-game seeded sweeps. That determinism turned out to be the key that unlocked community play: same seed, same board, for everyone.

For Reddit we used Devvit Web: the game runs in the post's web view, a lightweight splash card expands into the full experience, and a serverless backend (Hono) hands out the daily seed and keeps per-day and all-time leaderboards in Redis — best score per redditor, identity riding entirely on Reddit's own accounts.

Challenges we ran into

  • The port diet. Our web version had grown an experimental 3D layer with 150+ MB of scenery. The Reddit build needed to be lean, so we stripped it to the classic 2D presentation: a 17 MB bundle with under 1 MB of game code.
  • Simple gets complex fast. Our concept was rooted in simple play. No steep learning curve. But even in simplicity sits complexity that you will only really surface through playtesting and seeing other people play the game. Striking the right balance was a real journey. The game may fool the greedy player who just wants to bank points fast, but that player will find they run out of gems fast. Strategy in Glint is essential and is something players will learn as they keep trying to clear a board.
  • Real playtesting finds real bugs. Testing surfaced a lot of bugs and improvements, but that didn't stop when we ported our game to Reddit. Our first daily leaderboard run exposed a beauty: pressing Play again after a daily attempt silently dealt a random board that no longer counted. Dailies now survive restarts, zero-scores never submit, and runs that cross the midnight rollover can't pollute the next day's board.
  • Designing fair daily metrics. Every challenge had to be a per-run "higher is better" number — never first-to-X, which rewards timezones instead of play.

Accomplishments that we're proud of

A complete, polished game with a campaign, economy, dailies, achievements, that doesn't live on our own site but now also lives natively inside a Reddit post; a shared-board daily that turns a leaderboard into a conversation; and a deterministic engine we trust enough to let a whole community race on the same seed. This port to Reddit is going to open the game up to a great audience and we hope to learn so much more from the people who will play it and hopefully share their thoughts.

What we learnt

Worldbuilding pays unexpected dividends. Glint's theme, currency and tension were sitting in our lore waiting to be shipped. Determinism is a superpower: build your engine seeded and pure, and multiplayer-adjacent features come almost for free. And nothing replaces playing your own game on a phone, in the real feed, the way a player would.

What's next for Chrome Abyss: Glint

  1. Its tabletop siblings — Hybrid Uprising and Squads — continue development, sharing one universe. We hope to ship these next year.
  2. Glint itself goes app-ready (our full build already includes the 3D world of the Ascent and the Sticker Book), carrying the Community Daily and leaderboards with it, and further community enhancements such as adding friends, to ultimately be able to...
  3. ...play with a friend. The next big step: a 2-player mode, because in the Neon Bazaar, Glint was never meant to be played alone. On a shared board, the game becomes even more challenging and interesting. We might even consider creating a tabletop version of it as well.

Built With

Share this project:

Updates

Submission history