Qwen-VLA: Vision-Language-Action Modeling Across Tasks, Environments, and Robots | Hacker News Reader