Integer coordinates

Written for blit386 1.8.0. The rest of blit386.dev documents the newest engine, so an API you find there may be missing from a game pinned to 1.8.0.

All screen positions and sizes must be whole numbers. The engine works in a pixel grid; fractions do not make sense and cause off-by-one rendering glitches.

Use Vector2i for points

import { Vector2i } from 'blit386';

const pos = new Vector2i(10, 20); // correct - integer point
const pos = { x: 10.5, y: 20.7 }; // wrong   - plain object with floats

Vector2i auto-floors its inputs, so new Vector2i(10.9, 20.1) becomes (10, 20).

Use Rect2i for rectangles

import { Rect2i } from 'blit386';

const r = new Rect2i(x, y, width, height); // correct

Convert floats before use

When you compute a position from a float (like a physics result or Math.sin), floor it before passing to the engine.

const px = Math.floor(someFloat);
BT.drawPixel(new Vector2i(px, py), COLOR);

BT getters already return integers

BT.displaySize, BT.pointerPos(0), and BT.camera all return Vector2i - no conversion needed. The grid helpers sheet.tileRect(...) and Rect2i.fromTile(col, row, tileSize) (engine 1.8.0+) return Rect2i source rectangles.

Sprite sizes are whole numbers too

SpriteDrawParams.scale takes a positive integer (or a Vector2i of them), tileSize a positive integer per axis, and a stretched sprite or nine-slice fills a Rect2i. There is no fractional scale; for an in-between size, pass a Rect2i destination.