张桥雨

03 / 10

工厂建造 ROGUELITEFACTORY-BUILDING ROGUELITE

Fightory

把工厂建造和肉鸽结合在一起：玩家扫描随机分布的矿点，在敌人压力下搭建产线，再把产物转化为战斗资源。Factory building meets roguelite combat: scan randomly placed ore, build production lines under enemy pressure, and turn their output into combat resources.

个人项目｜2026.03–2026.05，暂时搁置。我负责玩法与系统设计和 Unity 玩法实现，项目改用 QFramework 时由朋友协助完成架构迁移。Personal project | March–May 2026, currently paused. I handled gameplay and systems design and Unity gameplay implementation; a friend helped migrate the architecture to QFramework.

演示视频待录制DEMO CAPTURE TODO↓
↓ 先看项目SHOW ME THE WORK

GAMEPLAY / DEMO VIDEO

Fightory

待录制CAPTURE TODO
01

语境与问题CONTEXT + PROBLEM

先定义玩家真正需要判断什么。Define what the player actually needs to judge.

2D 俯视角工厂建造 Roguelite。玩家在有限地图上侦测矿物、铺设生产设施，把资源送入对敌或对己的终端，并在战斗波次之间扩张。A top-down factory-building roguelite where the player scans for ore, lays out production, routes resources into offensive or defensive terminals, and expands between waves.

工厂布局一旦被解出就容易稳定下来。原型需要一种持续带回新信息的机制，让“重建”来自环境变化，而不是为了随机而随机。Once a factory layout is solved it tends to stabilize. The run needs a mechanism that continually reveals new information, making rebuilding a response to change rather than randomness for its own sake.

FIGHTORY / SYSTEM LOOP

主动扫描ACTIVE SCAN新信息NEW INFORMATION重组产线REBUILD LINE进入波次ENTER WAVE
02

关键决策KEY DECISIONS

规则不是功能清单，而是一组取舍。Rules are tradeoffs, not a feature list.

  1. 01

    把隐藏矿物交给主动扫描HIDDEN ORE REQUIRES ACTIVE SCANNING

    矿物不会一次性完全公开。扫描把信息获取变成玩家行动，并让地图上“哪里值得扩张”成为局内判断。Ore is not fully revealed up front. Scanning turns information acquisition into an action and makes expansion direction a run-level decision.

  2. 02

    让波次奖励改变可用地图WAVE REWARDS CHANGE THE USABLE MAP

    奖励扫描与地图扩张在波次之间揭示新资源节点，给玩家调整运输线的理由。新信息能否持续促使玩家改线，仍需要进一步观察。Reward scans and map expansion reveal new resource nodes between waves, giving the player a reason to reroute transport. Whether that information sustains rerouting decisions still needs observation.

  3. 03

    用“元素 × 目标”替代专用出口ELEMENT × TARGET INSTEAD OF DEDICATED EXITS

    我想让同一种资源能有更多用法，于是放宽了产线出口的资源限制：治疗资源也能给敌人回血，普通矿也能让玩家受伤。资源决定效果，出口决定作用对象。I wanted each resource to have more uses, so I relaxed the restrictions at production-line exits: healing resources can heal enemies, and ordinary ore can hurt the player. The resource defines the effect; the exit defines its target.

03

结果与反思OUTCOME + REFLECTION

把实现、验证和下一次判断连起来。Connect implementation, validation, and the next judgment.

当前原型已有开局扫描、随机矿点、基础产线搭建、运输加工、战斗波次与奖励扫描。配方仍是固定内容，完整随机配方循环和玩家侧火伤产线尚未完成；敌人干扰机器属于后续计划。The prototype includes opening scans, random ore, basic production, transport and processing, combat waves, and reward scans. Recipes are still fixed; the full random-recipe loop and player-side fire-damage line are unfinished. Enemy interference with machines remains a future plan.

放宽出口限制的代价是，接错线后产线不会马上堵住，玩家可能过一阵子才发现问题；好处是，同一种资源可以通过不同出口影响玩家或敌人。Relaxing exit restrictions has a cost: a misconnected line no longer jams immediately, so the player may discover the mistake later. In return, the same resource can affect the player or enemies through different exits.