Spatial VLM: Endowing Vision-Language Models with Spatial Reasoning Capabilities | Hacker News Reader