Tile World War development journal
GIT-IN
I started making this after watching streamer Hasangbyeong play similar games and thinking, “Could I make something like that?” It was supposed to be small — then it grew. This is the record of how that happened.
2026. 08. 18
On Railway and Vercel, more players meant more server bills. I started building revenue — Google AdSense first.
Auto ads would wreck the UI, so I placed slots by hand only where they fit. I try not to cover the play screen.
If you’re willing, an occasional ad click really helps keep the servers running :)
Maybe someday I’ll add donations — even coffee money would be amazing haha
2026. 08. 17
Matches used to auto-start when a cap filled or a timer ran out. Friends often missed the queue: “I didn’t get in — next round?”
I added Create War for private groups. The host opens a room, shares the link, players pick nickname and color, and the host hits start.
Like FIFA kit clashes, some villains pick nearly the same color. If your HEX is too close to someone else’s, we warn you to pick again.
2026. 08. 16
Then it hit me: I fell in love watching Hasangbyeong on my phone in bed, and the game in those videos was mobile first. I had only built a PC web UI. “That’s wrong!” — straight into mobile optimization.
PC tooltips covered the small screen, so I cut them all. Tap source → tap target: launch, move, or cancel in two touches.
I dropped heavy resource info boxes and left only troop count and region level on each tile. The game got leaner and faster.
Similar colors on a phone were hard to read. I cranked contrast and stamped each player’s flag on their tiles for clarity.
2026. 08. 14
On a wide map where everyone is an enemy, one snowballing player leaves the rest helpless. I added alliances as a political tool for ganging up.
Send SOS when you’re in trouble and get reinforcements. But there are no permanent friends — slip a missile onto ally land or push troops in, and you’re branded a traitor in brutal diplomacy.
More alliance perks will come in future updates.
Alliance design
2026. 08. 13
When a match ends and the screen just goes black, it feels empty. I built a slick result report so you can brag to friends and spread the word.
At the top, a timelapse replay of troop movement and conquest history. Below, a card with rank, peak tiles, favorite weapon, and more — copy a link or screenshot straight to the group chat.
I’ll make a TILE TAKE Twitter account later — please post there too!
2026. 08. 11
At first you moved troops and chipped 1 off at a time to capture. Cell Expansion Wars — my homage — ended fast on a small map. TILE TAKE uses world and continental regional grids, so the board was huge.
Pushing one tile at a time made matches long and loose. I added buildings and weapons to grow strong regions or shake distant land at once, so the front line wouldn’t freeze into troop streams only.
Strategy flavor: buildings & weapons
Buildings (fortify your land)
| Building | Cost | Build | Effect |
|---|---|---|---|
| City | 500 | 20s | 2× gold income speed on this tile |
| Barracks | 550 | 22s | 2× troop production speed on this tile |
| Port | 580 | 24s | Coast only. 2× sea speed; 0.15s departures from port |
| Rail | 600 | 28s | 0.1s departures when both ends have rail |
| Fort | 650 | 25s | 50% less damage when this tile is attacked |
| Factory | 700 | 25s | Build and store weapons (one per factory) |
| THAAD | 900 | 30s | 90% missile intercept before hit (not smoke/supply) |
Strategic weapons (one shot to flip the board)
| Weapon | Cost | Build | Effect |
|---|---|---|---|
| Precision | 2,000 | 30s | One enemy/neutral tile. −150 troops, destroys building; instant capture if troops hit 0 |
| EMP | 3,000 | 30s | Center + neighbors disabled 30s. Breaks alliance if ally in blast |
| Bio | 1,500 | 20s | 20s halt troop growth + 5/sec loss; instant capture if troops hit 0 |
| Nuke | 10,000 | 60s | Center + neighbors: wipe troops and buildings |
| Smoke | 1,000 | 30s | 50% less damage taken on this tile for 30s |
| Supply | 1,000 | 20s | +150 troops on friendly or allied tiles |
2026. 08. 10
I sketched the world map as a rough tile grid first. Coastlines looked jagged and blocky — “Is this even a map?”
So I rebuilt with GeoJSON, real terrain data. Instead of square tiles, the grid follows actual borders as closely as I could.
Real map data made expansion easy: not just the world, but Asia, Europe, the Americas — continental battlefields on demand.
If everyone always picked the same juicy spawn, strategy would get stale. Each match now rolls random cities worldwide as spawn candidates.
Random spawns are spicy. Two players can land on one tiny island and brawl instantly; with luck, you quietly snowball from a corner with no neighbors.
That day I deployed the game server on Railway and the web on Vercel. It finally left my laptop — anyone with a link could play.
2026. 08. 09
YouTube’s algorithm dropped a Hasangbyeong video in my feed. I watched, zoned out, and thought: if it’s this fun to watch, playing it must be a dopamine bomb.
So I started planning something in the same genre, mixing in rules from Cell Expansion Wars — an old favorite of mine.
Nations clash on a map; troops march from your land and swallow the next tile in your color. No download, no sign-up — just a nickname and you’re in the war.
Early design