Street House Builder
台灣街道常見違建、增築、鐵窗與招牌，這些元素來自不同使用者在不同時間對空間的擴張，彼此疊加後，形成看似混亂，卻具有自身秩序與強烈生命力的建築表情
Taiwan’s streets are shaped by informal additions, extensions, window grilles, and signs. Added by different occupants at different times to meet changing spatial needs, these elements accumulate into an architectural expression that appears chaotic yet carries its own order and a distinct sense of vitality.
在 Blender Geometry Nodes 有很多自動生成建築作品的案例，這邊則嘗試以 Grasshopper 重現這類日常街景，究以建築立面為主要操作對象，先用點位建立水平與垂直矩陣，再將立面拆分為樓梯間、一樓店面、住宅樓層、屋頂加蓋與廣告招牌等單元
Inspired by the many procedural building studies created with Blender Geometry Nodes, this experiment uses Grasshopper to reconstruct the character of these everyday streetscapes. Taking the façade as its primary field of operation, it begins with a horizontal and vertical point grid, then divides the elevation into components such as stairwells, ground-floor shops, residential floors, rooftop additions, and advertising signs.
每類單元建立數種 Block，透過規則配置到對應位置，使建築能隨面寬與樓層增減，同時維持基本的構成關係
Several block variations are created for each component and assigned to corresponding positions through a set of rules. This allows the building to expand or contract in width and height while retaining a basic compositional logic.
參數化系統可以快速產生具有特定街屋語彙的立面變化，但真實街景並不只是幾何元素的隨機排列，而是時間、需求與個別行為共同累積的結果，有限的模組與規則終究難以涵蓋現實世界中看似隨機、實則合理的複雜性
A parametric system can quickly generate façade variations with a recognizable street-house vocabulary. Yet a real streetscape is more than a random arrangement of geometry: it is the cumulative result of time, necessity, and individual action. A finite set of modules and rules can evoke its appearance, but cannot fully capture the seemingly accidental yet deeply grounded complexity of the built environment.