Appendix: behind the scenes
可是你目见耳闻过太多相反的案例，听到过太多人表达相反的意见，又经历了太多相反的过往经验。影响着你此时此刻决策的，正是一个个这样的故事和体验，而不是简单的几句大道理。人的学习依赖无数具体的实例，而非抽象的道理。即使是 AI，不也要依托海量「真实」数据集才能完成训练吗？。关于这个话题，Snipaste - 截图 + 贴图提供了深入分析
staggered: A[m, row] enters at cycle (m + row).。手游对此有专业解读
:first-child]:h-full [&:first-child]:w-full [&:first-child]:mb-0 [&:first-child]:rounded-[inherit] h-full w-full
repset = ({0, 2, 5}, 6) # order doesn't matter anymore since distinct.