{ "schema_version": "v1", "id": "selfteaching-craft-and-smart-questions-20260816", "date": "2026-08-16", "title": "《自学是门手艺》与《How To Ask Questions The Smart Way》：自学、作品与提问", "visibility": "public", "classification": { "category": "humanities-thought", "tags": [ "自学", "阅读方法", "记忆与整理", "刻意练习", "问题拆解", "作品与反馈", "技术沟通", "聪明提问", "知识库方法" ] }, "reading_source": { "label": "用户于 2026-08-16 指定的两份 GitHub Markdown 资料及其补充读书笔记", "provenance": "raw_notes 保留用户本次提交的原文。研究补全以用户指定仓库中的原始 Markdown 为主，采用概括和转述，不把长段落转换成未经核验的引号引文；编程章节另列为具体技能课程地图。" }, "raw_notes": [ "今天还学了的内容，请补充：\n李笑来的自学是一门手艺，几乎全部关于自学的内容，把其中的关键词和关键论点记录下来 https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/00.cover.md\n以及 https://github.com/selfteaching/How-To-Ask-Questions-The-Smart-Way\n\n08-15 记录了什么要学习的词吗？", "补充关键词：\n前置引用 过早引用", "磨练 “只字不差” 的能力\n通过阅读习得新技能（尤其是 “尽量只通过阅读习得新技能”），肯定与 “通过阅读获得心灵愉悦” 很不相同。\n\n读个段子、读个小说，读个当前热搜文章，通常情况下不需要 “精读” —— 草草浏览已经足够，顶多对自己特别感兴趣的地方，慢下来仔细看看……\n\n但是，若是为了习得新技能去阅读，就要施展 “只字不差地阅读” 这项专门的技能。\n\n对，“只字不差地阅读” 是所有自学能力强的人都会且都经常使用的技能。尤其是当你在阅读一个重要概念的定义之时，你就是这么干的：定义中的每个字都是有用的，每个词的内涵外延都是需要进行推敲的，它是什么，它不是什么，它的内涵外延都是什么，因此，在使用的时候需要注意什么……\n\n很有趣的一个现象是，绝大多数自学能力差的人，都是把一切都当作小说去看，随便看看，粗略看看……\n\n你有没有注意到一个现象，人们在看电影的时候，绝大多数人会错过绝大多数细节；但这好像并不会削减他们的观影体验；并且，他们有能力使用错过了无数细节之后剩下的那些碎片拼接出一个 “完整的故事” —— 当然，通常干脆是 “另一个貌似完整的故事”。于是，当你在跟他们讨论同一个电影的时候，常常像是你们没坐在同一个电影院，看的不是同一个电影似的……\n\n所谓的自学能力差，很可能最重要的坑就在这里：\n\n每一次学习新技能的时候，很多人只不过是因为做不到只字不差地阅读，于是总是会错过很多细节；于是，最终就好像 “看了另外一个山寨版电影一样”，实际上 “习得了另外一个山寨版技能”……", "好的记忆力很重要\n“就算读不懂也要读完” 的更高境界，是 “就算不明白也要先记住”。\n\n人们普遍讨厌 “死记硬背”…… 不过，说实话，这很肤浅。虽然确实也有 “擅长死记硬背却就是什么都不会的人”，但是，其实有更多记忆力强的人，实际上更可能是 “博闻强识”。\n\n面对 “过早引用” 常见的知识领域，好记忆力是超强加分项。记不清、记不住、甚至干脆忘了 —— 这是自学过程中最耽误事的缺点。尤其在有 “过早引用知识点” 存在的时候，更是如此。\n\n然而，很多人并没有意识到的是，记忆力也是 “一门手艺” 而已。并且，事实上，它是任何时候都可以通过刻意练习加强的 “手艺”。\n\n更为重要的是，记忆力这个东西，有一百种方法去弥补 —— 比如，最明显、最简单的办法就是 “好记性不如烂笔头”……\n\n所以，在绝大多数正常情况下，所谓的 “记不清、记不住、甚至干脆忘了”，都只不过是懒的结果 —— 若是一个人懒，且不肯承认自己懒，又因为不肯承认而已就不去纠正，那…… 那就算了，那就那么活下去罢。\n\n然而，提高对有效知识的记忆力还有另外一个简单实用的方法 —— 而市面上有各种 “快速记忆法”，通常相对于这个方法来看用处并不大。\n\n这个方法就是以下要讲到的 “整理归纳总结” —— 反复做整理归纳总结，记不住才怪呢！" ], "sources": [ { "id": "craft-cover", "title": "《自学是门手艺》：目录与总览", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/00.cover.md", "date": "2026-08-16", "evidence_role": "用户指定的原始目录与作品入口；用于确定全书范围" }, { "id": "craft-reading", "title": "《自学是门手艺》：只通过阅读习得新技能", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.1.C.must.learn.sth.only.by.reading.md", "date": "2026-08-16", "evidence_role": "原始章节；用于阅读、记忆和技能学习方法" }, { "id": "craft-forward", "title": "《自学是门手艺》：处理前置引用", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.1.F.deal-with-forward-references.md", "date": "2026-08-16", "evidence_role": "原始章节；用于 forward reference、先记住后理解和依赖关系" }, { "id": "craft-practice", "title": "《自学是门手艺》：刻意练习", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.2.B.deliberate-practicing.md", "date": "2026-08-16", "evidence_role": "原始章节；用于学—练—用—造、刻意练习和预算" }, { "id": "craft-thinking", "title": "《自学是门手艺》：刻意思考", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.2.E.deliberate-thinking.md", "date": "2026-08-16", "evidence_role": "原始章节；用于迁移、使用和思考训练" }, { "id": "craft-difficulty", "title": "《自学是门手艺》：征服困难", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.3.A.conquering-difficulties.md", "date": "2026-08-16", "evidence_role": "原始章节；用于困难幻觉、横向/纵向拆解和难点处理" }, { "id": "craft-thorough", "title": "《自学是门手艺》：全面与彻底", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.3.E.to-be-thorough.md", "date": "2026-08-16", "evidence_role": "原始章节；用于全面学习、完整作品和教与学" }, { "id": "craft-social", "title": "《自学是门手艺》：社会化自学", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.3.F.social-selfteaching.md", "date": "2026-08-16", "evidence_role": "原始章节；用于榜样、反馈和作品社交" }, { "id": "craft-google", "title": "《自学是门手艺》：Google 与自学黄金时代", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.3.G.the-golden-age-and-google.md", "date": "2026-08-16", "evidence_role": "原始章节；用于搜索、官方文档和互联网学习" }, { "id": "craft-focus", "title": "《自学是门手艺》：防止注意力漂移", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/the-craft-of-selfteaching/blob/master/markdown/Part.3.H.prevent-focus-drifting.md", "date": "2026-08-16", "evidence_role": "原始章节；用于目标、优先级、完成定义和注意力管理" }, { "id": "archive-20260815-video", "title": "2026-08-15 视频学习条目的 Git 历史归档", "kind": "historical_local_archive", "date": "2026-08-15", "evidence_role": "用于回答用户关于 08-15 显式单词记录的回看问题；不替代当前本地 canonical 文件", "availability": "本地来源；原始路径未发布，公开归档保留来源标题和证据角色" }, { "id": "smart-questions", "title": "How To Ask Questions The Smart Way", "kind": "github_primary_markdown", "url": "https://github.com/selfteaching/How-To-Ask-Questions-The-Smart-Way/blob/master/How-To-Ask-Questions-The-Smart-Way.md", "date": "2014-05-21", "evidence_role": "用户指定的原始文章；用于提问、搜索、最小复现和反馈礼仪" } ], "research_completions": [ { "question": "《自学是门手艺》主要在教什么？", "answer": "它把自学当作可以拆解、练习、复盘和反复使用的一门手艺，而不是神秘天赋。全书以编程学习为主要示例，但核心对象是可迁移的学习能力：阅读、记忆、练习、思考、作品、反馈和继续学习。", "status": "paraphrase", "learning_value": "把“我有没有天赋”改写成“我能否建立一套可重复的学习流程”，这是全书最重要的视角变化。", "source_ids": [ "craft-cover", "craft-reading", "craft-forward", "craft-practice", "craft-thinking", "craft-difficulty", "craft-thorough", "craft-social", "craft-google", "craft-focus" ] }, { "question": "为什么作者从编程开始讲自学？", "answer": "编程是一个适合展示自学过程的例子：资料和官方文档相对可获得，知识会出现前置引用，成果可以运行、测试和展示，难点也能被拆成更小的可操作任务。编程是示范场，不等于全书只适用于程序员。", "status": "paraphrase", "learning_value": "迁移到其他技能时，寻找同样的可验证产物和反馈回路，而不是只模仿编程术语。", "source_ids": [ "craft-cover", "craft-reading" ] }, { "question": "“尽量只通过阅读习得新技能”是什么意思？", "answer": "重点不是拒绝实践，而是训练从文字、定义、目录、文档、示例和错误信息中建立模型的能力；实践用于检验和修正阅读所得的模型。娱乐性浏览可以粗读，为习得技能而读则要主动理解结构、边界、例外和使用条件。", "status": "paraphrase", "learning_value": "阅读不再只是接收信息，而是把文字转成可操作模型，再用任务或作品验证。", "source_ids": [ "craft-reading" ] }, { "question": "为什么第一次阅读可以囫囵吞枣？", "answer": "第一次通读的目标是画出地图，知道章节怎样连接、哪些概念反复出现、哪里需要回头；之后通过多次阅读逐层加深。粗读是建立导航，不是把粗读误当成掌握。", "status": "paraphrase", "learning_value": "给每次阅读设定目标：先要地图，再要定义、边界、例外和可用性，避免一开始因求全而卡住。", "source_ids": [ "craft-reading", "craft-cover" ] }, { "question": "“只字不差地阅读”具体要求什么？", "answer": "面对重要定义、规则、接口和关键论证，要逐字检查每个词的含义、内涵、外延、反例和使用条件，区分“它是什么”与“它不是什么”。自学能力差的常见坑，是把技能材料一律当小说浏览，漏掉细节后拼出一个貌似完整但实际变形的技能。", "status": "paraphrase", "learning_value": "在学习日志里记录定义、边界、例外、例子和未解决词，而不是只记一段模糊感想。", "source_ids": [ "craft-reading" ] }, { "question": "为什么“就算不明白也要先记住”？", "answer": "遇到前置引用时，当前概念可能依赖后文或另一门知识；先把术语、关系和原句位置记住，继续读完并重复回看，之后再补理解。记住但暂时不理解是一个过渡状态，不是把死记硬背当成最终目标。", "status": "paraphrase", "learning_value": "对暂时不懂的词建立待回收清单，记录它出现的位置和依赖关系，降低重新定位的成本。", "source_ids": [ "craft-forward", "craft-reading" ] }, { "question": "“前置引用”和“过早引用”在本条笔记里怎样理解？", "answer": "用户把两者并列记录；本书较稳定的英文术语是 forward reference，指后文或后续学习才解释的概念被提前使用。中文“过早引用”可以作为记忆提示，但不把它扩展成所有形式的引用错误；关键是识别依赖、先记录、继续推进、回头补全。", "status": "paraphrase", "learning_value": "遇到看似突然出现的术语，先判断它是前置引用、定义缺失、翻译问题还是自己漏读，不要立刻断言作者写错。", "source_ids": [ "craft-forward" ] }, { "question": "好的记忆力为什么是自学加分项？", "answer": "记忆力能让人保留术语、定义、关系和过去的推理，从而减少反复寻找和断裂。作者同时强调记忆力可以训练，也可以用笔记、整理、归纳和总结来补偿；重点不是崇拜天生记忆，而是建立可靠的外部记忆和重复回收机制。", "status": "paraphrase", "learning_value": "把“记不住”拆成记忆、检索、记录和整理四个问题，先用低成本的记录与复盘补上，而不是直接归因于能力不足。", "source_ids": [ "craft-forward", "craft-reading" ] }, { "question": "“整理归纳总结”为什么能帮助记忆？", "answer": "整理会迫使学习者比较、分类、命名、压缩和重新表达；归纳让分散的例子形成结构，总结让结构变成可调用的短版本。反复做这些动作会增加主动提取和关系重建，而不只是重复看原文。", "status": "paraphrase", "learning_value": "每次读完一个小节，产出关键词、结构图、自己的解释和一个使用例子，形成可复习的记忆单元。", "source_ids": [ "craft-forward", "craft-thorough" ] }, { "question": "全书的“学—练—用—造”怎样连接？", "answer": "学习获得材料和模型，练习把局部动作做稳定，使用把技能放进真实任务，创造或完成作品则要求把多个能力整合起来并接受外部检验。作者强调以用带练、找活干，说明真实任务会暴露真正的薄弱点。", "status": "paraphrase", "learning_value": "不要把“看完教程”当成终点；尽早寻找一个规模可控、能交付、能验证的使用场景。", "source_ids": [ "craft-practice", "craft-cover" ] }, { "question": "刻意练习和刻意思考有什么区别？", "answer": "刻意练习针对具体弱点设计重复、反馈和难度；刻意思考则主动追问一个知识能做什么、还能做什么、限制在哪里，以及如何把大问题拆开。前者改善执行，后者改善选择练什么和如何迁移。", "status": "paraphrase", "learning_value": "每次练习同时写下“我正在修哪个弱点”和“这个能力还能迁移到什么任务”，避免机械刷题。", "source_ids": [ "craft-practice", "craft-thinking" ] }, { "question": "学习预算为什么重要？", "answer": "预算不仅是钱，更是时间、注意力、精力、重复次数和反馈机会。低估预算会造成焦虑、半途而废，或把尚未投入足够时间造成的结果误判成自己没有天赋；预算应随任务难度和现实约束调整。", "status": "paraphrase", "learning_value": "开始前写出可投入的时间、每周重复次数、可接受的失败次数和验收作品，学习计划才可检验。", "source_ids": [ "craft-practice" ] }, { "question": "困难幻觉和拆解方法是什么？", "answer": "困难常常来自任务规模、层级和依赖没有分开。横向拆解把同一层的工作拆成并列小任务；纵向拆解把目标、抽象、实现、工具和细节分层，再逐层验证。拆解不是逃避难点，而是把难点变成可处理的序列。", "status": "paraphrase", "learning_value": "遇到“太难”时先画输入、输出、依赖、层级和最小验证，不要用继续搜集资料替代真正拆解。", "source_ids": [ "craft-difficulty" ] }, { "question": "为什么不能只学简单部分？", "answer": "如果一个技能的关键约束、边界或核心难点始终没有学会，表面会做几个例子并不等于拥有完整技能。作者鼓励正面处理难点，同时用预算、拆解、重复和作品把难点分摊到可执行阶段。", "status": "paraphrase", "learning_value": "在学习地图中标出“必须会而尚未会”的节点，并为每个节点安排一个小实验或解释任务。", "source_ids": [ "craft-difficulty", "craft-practice" ] }, { "question": "“刚需幻觉”和时间幻觉提醒什么？", "answer": "人在尚未掌握技能之前，往往感受不到它的真实用途，于是误以为“不需要”；而对时间的估计也常常过于乐观。需要感和熟练后的收益可能是学习的结果，而不是学习的前提。", "status": "paraphrase", "learning_value": "选择技能时估计机会成本，但不要把当前没有强烈需求当作唯一否决理由；用小作品验证真实需求。", "source_ids": [ "craft-difficulty", "craft-focus" ] }, { "question": "为什么“全面”与“完整作品”都重要？", "answer": "一份材料、一个角度或几个局部技巧不能代表完整理解；全面意味着补齐相关视角、文档、例子、实践和反馈。小作品不必宏大，但必须从输入到输出闭环，完整性比规模更能暴露依赖和证明能力。", "status": "paraphrase", "learning_value": "用“最小完整作品”作为阶段验收，而不是用读过多少页或收藏多少资料作为完成标准。", "source_ids": [ "craft-thorough", "craft-social" ] }, { "question": "自学是否意味着独自学习、拒绝社交？", "answer": "不是。自学是对学习责任和路径的掌握，不是拒绝老师、同伴、社区或反馈。有效的社会化自学包括寻找榜样、阅读他人作品、请人指出问题、公开作品并参与有内容的讨论；泛泛社交本身不等于学习。", "status": "paraphrase", "learning_value": "优先建立作品—反馈—修正的关系，寻找能提供具体意见的人，而不是只积累联系人。", "source_ids": [ "craft-social" ] }, { "question": "作品为什么是自学者的证据和社交货币？", "answer": "作品把抽象的“我学过”变成可观察、可运行、可阅读或可使用的结果，同时暴露跨领域缺口：表达、测试、设计、交付和协作。作品越小越容易完成，但需要真实闭环和可复盘记录。", "status": "paraphrase", "learning_value": "每个学习阶段都保留一个可展示的最小成果，并记录它暴露出的下一个问题。", "source_ids": [ "craft-social", "craft-thorough" ] }, { "question": "互联网、搜索和官方文档在自学中扮演什么角色？", "answer": "互联网降低了获得教材、源码、官方文档、实例和他人经验的成本；搜索不是最后一步，而是自学者的基本动作。遇到问题先查原始文档、错误信息和已有讨论，再带着具体证据提问。", "status": "paraphrase", "learning_value": "建立“官方文档—源码/示例—搜索结果—自己的实验—精确提问”的证据链。", "source_ids": [ "craft-google", "smart-questions" ] }, { "question": "如何防止注意力漂移？", "answer": "先定义本轮学习的完成条件、优先级和不做清单；把与当前目标无关但有价值的发现放入停车场。否则学习者会不断打磨边缘细节、换工具或扩展范围，却没有完成最初承诺的作品。", "status": "paraphrase", "learning_value": "开始前写一句“完成意味着什么”，结束时按这句话验收，不用新兴趣重写原目标。", "source_ids": [ "craft-focus" ] }, { "question": "聪明提问前应该做哪些准备？", "answer": "先搜索论坛和网页、阅读手册与 FAQ、检查源码或最小实验，确认问题属于合适的社区，并说明已经查过什么、试过什么。准备不是为了证明自己聪明，而是减少重复劳动，使回答者能直接定位缺口。", "status": "paraphrase", "learning_value": "提问前保留搜索词、文档链接、命令、结果和仍然不明白的那一小段。", "source_ids": [ "smart-questions" ] }, { "question": "一个高质量技术问题应该包含什么？", "answer": "清晰主题、目标、环境、版本、现象、复现步骤、已做诊断、最近变化、最小可复现例子和明确请求。描述症状而不是先写未经验证的理论；描述目标而不是只罗列自己走过的步骤。", "status": "paraphrase", "learning_value": "使用“目标—环境—症状—已查—已试—最小复现—明确请求”的固定模板，再按社区格式调整。", "source_ids": [ "smart-questions" ] }, { "question": "为什么要区分症状、猜测和目标？", "answer": "症状是可以被他人观察或复现的事实，猜测是对原因的解释，目标是希望最终达到的状态。把猜测写成事实会把回答者带进错误方向；只写步骤又可能让人不知道你真正想实现什么。", "status": "paraphrase", "learning_value": "先给可验证事实，再单独列假设和目标；让回答者可以纠正假设，而不是只能顺着假设找补丁。", "source_ids": [ "smart-questions" ] }, { "question": "最小可复现例子为什么重要？", "answer": "它把复杂环境压缩为能稳定展示问题的最小案例，让别人可以运行、比较和定位；也迫使提问者确认问题究竟在哪里。代码审查请求与 bug 复现请求要分开说明。", "status": "paraphrase", "learning_value": "先删掉无关代码、依赖和背景，保留能失败的最短测试，再附完整环境信息作为补充。", "source_ids": [ "smart-questions" ] }, { "question": "RTFM 和 STFW 应该怎样理解？", "answer": "它们分别是“先读手册”和“先搜索网页”的粗略提醒，不是对提问者人格的判决。若手册或搜索无法解决问题，应具体说明读了哪一节、试了哪些关键词、哪一步仍不清楚；回答者也应避免只丢缩写。", "status": "paraphrase", "learning_value": "把“我查过了”变成可核验的查找路径，问题会从泛问变成可回答的局部问题。", "source_ids": [ "smart-questions" ] }, { "question": "怎样回答别人才能真正帮助学习？", "answer": "先回答实际问题，再解释判断依据；可以通过追问暴露缺失信息，也可以指出更好的搜索和调试方法。高质量回答应减少羞辱和模糊权威感，并在必要时改进文档、FAQ 或示例，让下一个人不必重复提问。", "status": "paraphrase", "learning_value": "把问答当作共同改善知识环境的过程，而不是展示谁更懂行。", "source_ids": [ "smart-questions" ] }, { "question": "书中编程章节应该怎样放进今天的笔记？", "answer": "Part 1 依次覆盖布尔值与运算、控制流、函数、字符串、容器和文件；Part 2/3 继续到函数细节、模块、文档字符串、测试驱动开发、类、装饰器、迭代器、生成器、正则表达式和语法表示。它们是“通过阅读学习一门具体技能”的课程地图，不应被误写成普适自学原则。", "status": "paraphrase", "learning_value": "将元学习方法和编程知识分成两个层级：一层记录怎么学，另一层记录正在学什么，避免知识库结构混乱。", "source_ids": [ "craft-cover", "craft-reading" ] }, { "question": "08-15 记录了什么要学习的词？", "answer": "08-15 的视频学习条目有 14 条显式 vocabulary 记录：Change your personality、intelligentsia、Rhythm、Melody、Form、Chord、Accompaniment、Verse、Chorus、Style、Articulate why、Western culture、Repetition、beat。其中 intelligentsia 已核验为“知识分子阶层/知识界”，其余多数保留为 provisional 词条；同日《江湖丛谈》条目没有单独的 vocabulary 行，只有关键词。", "status": "verified", "learning_value": "回顾时把 08-15 的词汇库与《江湖丛谈》的概念关键词分开，不把关键词数量误当成显式单词记录数量。", "source_ids": [ "archive-20260815-video" ] } ], "keywords": [ { "term": "自学是一门手艺", "definition": "自学不是神秘天赋，而是可以通过时间、重复、记录、反馈和作品不断改进的能力系统。", "category": "自学总论", "usefulness": "把天赋焦虑转成可练习的流程设计。", "source_ids": [ "craft-cover", "craft-reading", "craft-forward", "craft-practice", "craft-thinking", "craft-difficulty", "craft-thorough", "craft-social", "craft-google", "craft-focus" ] }, { "term": "自学能力", "definition": "从材料中建立模型、安排练习、处理困难、验证成果并继续迭代的综合能力。", "category": "自学总论", "usefulness": "作为所有具体技能之上的可迁移元技能。", "source_ids": [ "craft-cover", "craft-reading", "craft-forward", "craft-practice", "craft-thinking", "craft-difficulty", "craft-thorough", "craft-social", "craft-google", "craft-focus" ] }, { "term": "学—练—用—造", "definition": "从获得知识，到稳定动作，再到真实使用，最后整合成可交付作品的四层路径。", "category": "学习路径", "usefulness": "用阶段性产物判断是否真正前进。", "source_ids": [ "craft-practice" ] }, { "term": "只靠阅读习得新技能", "definition": "把阅读训练成理解定义、依赖、例外和用法的主动技能，并用实践验证模型。", "category": "阅读方法", "usefulness": "适合建立文档学习和自学日志。", "source_ids": [ "craft-reading" ] }, { "term": "囫囵吞枣", "definition": "第一次通读时先获取地图和结构，不要求立即掌握所有细节；它不能替代后续精读。", "category": "阅读方法", "usefulness": "减少初读卡顿，同时防止把通读错当掌握。", "source_ids": [ "craft-reading" ] }, { "term": "只字不差地阅读", "definition": "对重要定义、规则和接口逐词检查含义、边界、例外和使用条件。", "category": "阅读方法", "usefulness": "避免遗漏细节后拼出一个山寨版技能。", "source_ids": [ "craft-reading" ] }, { "term": "前置引用", "definition": "后文或后续学习才会解释的概念被提前使用，英文常对应 forward reference。", "category": "依赖与理解", "usefulness": "遇到不懂术语时先定位依赖，不急于断言错误。", "source_ids": [ "craft-forward" ] }, { "term": "过早引用", "definition": "用户对前置引用的中文记忆提示；在本条中不把它扩大为所有引用时机不当的问题。", "category": "依赖与理解", "usefulness": "保留用户原词，同时区分本书术语和个人解释。", "source_ids": [ "craft-forward" ] }, { "term": "记住但暂时不理解", "definition": "面对前置引用时，先保留术语、关系和位置，继续阅读并在后续回收理解。", "category": "记忆与复习", "usefulness": "让不理解成为可管理的中间状态。", "source_ids": [ "craft-forward" ] }, { "term": "好的记忆力", "definition": "保留定义、关系、推理和过去经验的能力；可以通过记录、整理和重复来训练或补偿。", "category": "记忆与复习", "usefulness": "把记忆问题转成可操作的外部记忆系统。", "source_ids": [ "craft-forward", "craft-reading" ] }, { "term": "整理归纳总结", "definition": "对材料进行分类、压缩、重述、比较和结构化，以增加主动提取和关系重建。", "category": "记忆与复习", "usefulness": "把阅读转成可回忆的知识节点。", "source_ids": [ "craft-forward", "craft-thorough" ] }, { "term": "刻意练习", "definition": "针对明确弱点设计带反馈、可重复、略有挑战的练习，而不是无目标地重复熟悉动作。", "category": "练习设计", "usefulness": "提高练习的反馈密度和针对性。", "source_ids": [ "craft-practice" ] }, { "term": "刻意思考", "definition": "主动追问知识能做什么、还能做什么、边界在哪里以及如何迁移或拆解。", "category": "思考与迁移", "usefulness": "避免只会复述，不会选择和使用。", "source_ids": [ "craft-thinking" ] }, { "term": "预算", "definition": "学习中的时间、精力、注意力、重复次数、失败次数和反馈机会的总约束。", "category": "计划与约束", "usefulness": "防止低估投入后把结果误判成天赋问题。", "source_ids": [ "craft-practice" ] }, { "term": "找活干 / 以用带练", "definition": "尽快寻找能逼出真实问题的使用场景，用任务带动练习，而不是无限准备。", "category": "学习路径", "usefulness": "让练习暴露真正的薄弱环节。", "source_ids": [ "craft-practice" ] }, { "term": "困难幻觉", "definition": "复杂感可能来自任务规模、层级、依赖和抽象没有拆开，而不等于目标不可学。", "category": "困难处理", "usefulness": "先拆解和验证，再判断是否真的困难。", "source_ids": [ "craft-difficulty" ] }, { "term": "横向拆解", "definition": "在同一抽象层把大任务分成并列的小任务，并分别定义输入和输出。", "category": "困难处理", "usefulness": "降低单次操作的复杂度。", "source_ids": [ "craft-difficulty" ] }, { "term": "纵向拆解", "definition": "把目标、抽象、实现、工具和细节分层，逐层验证各自的问题。", "category": "困难处理", "usefulness": "避免把不同层次的困惑混成一个大问题。", "source_ids": [ "craft-difficulty" ] }, { "term": "刚需幻觉", "definition": "尚未掌握技能时感受不到它的用途，于是把“当前感觉不需要”误当成“长期没有价值”。", "category": "动机与选择", "usefulness": "用小作品验证需求，不被即时感觉完全支配。", "source_ids": [ "craft-difficulty" ] }, { "term": "全面", "definition": "不把单一教材、单一视角或局部例子当成完整理解，主动补充文档、实践、反馈和不同解释。", "category": "完整学习", "usefulness": "降低单一来源造成的盲区。", "source_ids": [ "craft-thorough" ] }, { "term": "完整作品", "definition": "规模可以很小，但必须从输入到输出闭环，能运行、展示、使用或被他人检查。", "category": "作品与验证", "usefulness": "用闭环成果替代“看过很多”的完成感。", "source_ids": [ "craft-thorough" ] }, { "term": "作品社交", "definition": "通过作品获得具体反馈、合作机会、信誉和进一步学习关系，而不是靠泛泛社交证明自己。", "category": "社会化自学", "usefulness": "把社交连接到可观察的价值交付。", "source_ids": [ "craft-social" ] }, { "term": "官方文档优先", "definition": "遇到技术问题先查一手说明、定义、参数、限制和示例，再参考二手解释。", "category": "搜索与证据", "usefulness": "减少把过时或错误转述当成规范。", "source_ids": [ "craft-google", "smart-questions" ] }, { "term": "注意力漂移", "definition": "学习过程中不断扩大范围、打磨边缘细节或更换工具，最后偏离原始完成条件。", "category": "注意力管理", "usefulness": "用优先级、完成定义和停车场保护交付。", "source_ids": [ "craft-focus" ] }, { "term": "聪明提问", "definition": "在搜索、阅读手册、实验和整理信息之后，向合适的人提出具体、可复现、容易回答的问题。", "category": "沟通与提问", "usefulness": "把提问变成自学流程的下一步，而不是搜索替代品。", "source_ids": [ "smart-questions" ] }, { "term": "症状而非猜测", "definition": "先描述可观察、可复现的现象，再把原因解释作为待验证假设单独列出。", "category": "技术沟通", "usefulness": "避免错误理论把排查方向带偏。", "source_ids": [ "smart-questions" ] }, { "term": "目标而非步骤", "definition": "说明想达到的状态，而不是只罗列已经走过的操作步骤。", "category": "技术沟通", "usefulness": "让回答者知道真正要解决的任务。", "source_ids": [ "smart-questions" ] }, { "term": "最小可复现例子", "definition": "能稳定展示问题且尽可能删去无关背景、代码和依赖的最小案例。", "category": "技术沟通", "usefulness": "提高复现、定位和回答效率。", "source_ids": [ "smart-questions" ] }, { "term": "RTFM / STFW", "definition": "分别提醒先读手册、先搜索网页；若仍未解决，应展示具体查找路径和剩余缺口。", "category": "搜索与提问", "usefulness": "把“查过了”变成可核验的证据链。", "source_ids": [ "smart-questions" ] }, { "term": "内容第一", "definition": "沟通先交付真实内容和解决问题的价值，再谈姿态、权威和表达风格。", "category": "沟通与写作", "usefulness": "减少空泛礼貌和权威表演。", "source_ids": [ "smart-questions" ] }, { "term": "输送价值观", "definition": "交流不仅传递答案，也传递如何搜索、验证、提问、协作和对待他人的方法。", "category": "沟通与写作", "usefulness": "把一次问答转成可复用的学习环境改进。", "source_ids": [ "smart-questions" ] }, { "term": "编程课程地图", "definition": "从基础值、控制流、函数、字符串、容器、文件，到模块、测试、类、迭代器、生成器、装饰器、正则和语法表示的学习路线。", "category": "具体技能示例", "usefulness": "将“怎么学”与“学什么”分层管理。", "source_ids": [ "craft-cover" ] } ], "vocabulary": [ { "id": "self-teaching", "term": "self-teaching", "translation": "自学；自主学习", "language": "en", "definition": "Learning a skill by taking responsibility for the path, materials, practice, and verification.", "category": "学习方法", "levels": [ "CEFR-B2", "Academic/PhD" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-cover", "craft-reading" ] }, { "id": "proof-of-work", "term": "proof of work", "translation": "工作证明；作品证据", "language": "en", "definition": "A concrete artifact or result that makes learning or capability observable.", "category": "学习与作品", "levels": [ "CEFR-C1", "Academic/PhD" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-cover", "craft-social" ] }, { "id": "forward-reference", "term": "forward reference", "translation": "前置引用；前向引用", "language": "en", "definition": "A concept or dependency used before the explanation or definition appears later in the material.", "category": "阅读与依赖", "levels": [ "CEFR-C1", "Academic/PhD" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-forward" ] }, { "id": "deliberate-practice", "term": "deliberate practice", "translation": "刻意练习", "language": "en", "definition": "Practice designed around a specific weakness, repetition, feedback, and an appropriate challenge.", "category": "练习设计", "levels": [ "CEFR-C1", "Academic/PhD" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-practice" ] }, { "id": "deliberate-thinking", "term": "deliberate thinking", "translation": "刻意思考", "language": "en", "definition": "Intentional reasoning about uses, alternatives, limits, decomposition, and transfer.", "category": "思考与迁移", "levels": [ "CEFR-C1", "Academic/PhD" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-thinking" ] }, { "id": "learning-curve", "term": "learning curve", "translation": "学习曲线", "language": "en", "definition": "The changing relationship between practice or time and performance in learning a skill.", "category": "学习方法", "levels": [ "CEFR-B2" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-practice" ] }, { "id": "horizontal-decomposition", "term": "horizontal decomposition", "translation": "横向拆解", "language": "en", "definition": "Breaking work into parallel tasks at the same level of abstraction.", "category": "困难处理", "levels": [ "CEFR-C1", "Academic/PhD" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-difficulty" ] }, { "id": "vertical-decomposition", "term": "vertical decomposition", "translation": "纵向拆解", "language": "en", "definition": "Separating a problem into layers such as goal, abstraction, implementation, tools, and details.", "category": "困难处理", "levels": [ "CEFR-C1", "Academic/PhD" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-difficulty" ] }, { "id": "attention-drift", "term": "attention drift", "translation": "注意力漂移", "language": "en", "definition": "A gradual movement away from the original priority or completion condition during work.", "category": "注意力管理", "levels": [ "CEFR-C1" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-focus" ] }, { "id": "official-documentation", "term": "official documentation", "translation": "官方文档", "language": "en", "definition": "Documentation maintained by the authority or project responsible for the relevant system.", "category": "搜索与证据", "levels": [ "CEFR-B2" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-google", "smart-questions" ] }, { "id": "table-of-contents", "term": "table of contents", "translation": "目录", "language": "en", "definition": "A structured list of sections that helps a reader build a map before detailed reading.", "category": "阅读方法", "levels": [ "CEFR-B1" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-reading" ] }, { "id": "glossary", "term": "glossary", "translation": "术语表；词汇表", "language": "en", "definition": "An alphabetical or organized list explaining specialized terms in a work.", "category": "阅读方法", "levels": [ "CEFR-B2" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "craft-reading" ] }, { "id": "minimal-reproducible-example", "term": "minimal reproducible example", "translation": "最小可复现例子", "language": "en", "definition": "The smallest case that reliably demonstrates a problem and allows another person to reproduce it.", "category": "技术沟通", "levels": [ "CEFR-C1", "Academic/PhD" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "smart-questions" ] }, { "id": "object-deviation", "term": "object - deviation", "translation": "对象—偏差；主题—异常", "language": "en", "definition": "A suggested question-subject pattern that names the object and the observed deviation.", "category": "技术沟通", "levels": [ "CEFR-C1" ], "level_basis": "expert_estimate", "level_confidence": "low", "status": "provisional", "source_ids": [ "smart-questions" ] }, { "id": "symptoms-not-guesses", "term": "symptoms, not guesses", "translation": "描述症状，不先下猜测", "language": "en", "definition": "A communication principle: state observable behavior before asserting an unverified cause.", "category": "技术沟通", "levels": [ "CEFR-C1" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "smart-questions" ] }, { "id": "goal-not-steps", "term": "goal, not just steps", "translation": "说明目标，不只罗列步骤", "language": "en", "definition": "A communication principle: tell helpers the desired outcome, not only the path already attempted.", "category": "技术沟通", "levels": [ "CEFR-C1" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "smart-questions" ] }, { "id": "read-the-manual", "term": "RTFM", "translation": "先读手册", "language": "en", "definition": "A blunt abbreviation for reading the relevant manual or documentation before asking a basic question.", "category": "搜索与提问", "levels": [ "CEFR-C1" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "notes": "原文解释了这个缩写；语气较尖锐，不应替代真正有帮助的回答。", "source_ids": [ "smart-questions" ] }, { "id": "search-the-web", "term": "STFW", "translation": "先搜索网页", "language": "en", "definition": "A blunt abbreviation for searching the web before asking a question that may already be answered.", "category": "搜索与提问", "levels": [ "CEFR-C1" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "notes": "原文解释了这个缩写；提问时应记录实际搜索路径。", "source_ids": [ "smart-questions" ] }, { "id": "cross-posting", "term": "cross-posting", "translation": "跨社区重复发帖", "language": "en", "definition": "Posting the same question to multiple forums or groups instead of choosing a suitable venue.", "category": "技术沟通", "levels": [ "CEFR-C1" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "smart-questions" ] }, { "id": "shotgun-question", "term": "shotgun question", "translation": "散射式提问；一口气乱问很多问题", "language": "en", "definition": "A vague or broad question that makes it difficult for a responder to identify the actual issue.", "category": "技术沟通", "levels": [ "CEFR-C1" ], "level_basis": "expert_estimate", "level_confidence": "medium", "status": "verified", "source_ids": [ "smart-questions" ] } ], "reading_contexts": [ { "id": "selflearning-loop", "summary": "这本书的自学主线可以压缩为：先用目录和通读建立地图，再对定义和关键依赖进行只字不差的阅读；把暂时不懂的内容记住并通过整理归纳总结回收；用刻意练习和刻意思考处理弱点；通过真实任务和小而完整的作品验证；再借助搜索、提问、反馈和社会化学习继续迭代。", "author": "李笑来", "work": "《自学是门手艺》", "writing_style": "以编程学习为具体示例，把阅读、记忆、练习、思考、作品和社交连接成一套可操作的学习论。", "writing_use": "可用于写学习方法、知识管理、技能迁移和作品导向的文章。", "register": "面向普通自学者的实践性论说", "collocations": [ "read through", "read carefully", "deliberate practice", "deliberate thinking", "proof of work" ], "source_context": "用户指定的 GitHub Markdown 版本及相关章节。", "source_locator": "00.cover.md；Part.1.C；Part.1.F；Part.2.B；Part.2.E；Part.3.A；Part.3.E—H。", "vocabulary_ids": [ "self-teaching", "proof-of-work", "forward-reference", "deliberate-practice", "deliberate-thinking", "learning-curve", "horizontal-decomposition", "vertical-decomposition", "attention-drift", "official-documentation", "table-of-contents", "glossary" ], "source_ids": [ "craft-cover", "craft-reading", "craft-forward", "craft-practice", "craft-thinking", "craft-difficulty", "craft-thorough", "craft-social", "craft-google", "craft-focus" ] }, { "id": "smart-question-loop", "summary": "聪明提问不是把自学交给别人，而是把自己的搜索、阅读和实验推进到一个精确缺口，再把这个缺口以可复现、容易回答的形式交给合适的社区。提问和回答都应改善问题解决能力与公共文档，而不是制造羞辱或权威表演。", "author": "Eric Steven Raymond；用户指定仓库为 selfteaching 的 Markdown 版本", "work": "How To Ask Questions The Smart Way", "writing_style": "以直接的社区规范和反例说明怎样准备、提问、回答与跟进。", "writing_use": "可用于技术沟通、团队协作、知识库提问模板和写作编辑。", "register": "技术社区行为指南", "collocations": [ "minimal reproducible example", "official documentation", "symptoms", "goal", "follow-up" ], "source_context": "用户指定 GitHub 仓库中的 Markdown 原文；该页面记录的版本修订信息属于历史材料，不等同于今天所有平台的规则。", "source_locator": "How-To-Ask-Questions-The-Smart-Way.md；Revision 3.10, 21 May 2014。", "vocabulary_ids": [ "minimal-reproducible-example", "object-deviation", "symptoms-not-guesses", "goal-not-steps", "read-the-manual", "search-the-web", "cross-posting", "shotgun-question" ], "source_ids": [ "smart-questions" ] } ], "quotes": [], "reading_comprehension": { "summary": "两份材料可以连成一个自学闭环：先用目录和通读建立地图，再在关键定义、依赖和边界处只字不差地阅读；对前置引用先记住并继续推进，通过整理归纳总结把信息变成可回收的结构；用刻意练习修复弱点、用刻意思考寻找迁移和拆解；用真实任务和小而完整的作品验证；最后通过搜索、聪明提问、反馈和社会化学习继续迭代。自学因此不是闭门造车，也不是把问题随手丢给别人，而是对自己的证据链和下一步负责。", "implications_label": "学习与知识管理含义", "concept_connections": [ "前置引用 → 记住但暂时不理解 → 继续阅读 → 回头补全依赖", "只字不差地阅读 → 定义、边界、例外 → 避免习得山寨版技能", "整理归纳总结 → 主动回忆 → 记忆与知识结构稳定化", "学—练—用—造 → 真实任务 → 小而完整的作品", "刻意练习 + 刻意思考 → 针对弱点 → 迁移与拆解", "困难幻觉 + 学习预算 → 可执行的分层任务 → 及时反馈", "官方文档/搜索 → 最小实验 → 聪明提问 → 社区反馈", "作品社交 → 具体反馈和信誉 → 社会化自学的下一轮" ], "market_implications": [ "本条不提供市场判断；它只更新个人学习、知识整理和技术沟通方法，不应直接转成投资观点或交易动作。" ] }, "market_update": { "section_title": "今日市场更新（无）", "as_of": "2026-08-16", "summary": "本次阅读主题是自学方法与技术提问，没有新增当前市场数据、价格、行业判断或投资建议。", "points": [] }, "knowledge_updates": [ { "original_idea": "只要看过材料，就大致算学会了。", "verdict": "需要修正", "updated_view": "技能学习需要区分地图式通读、只字不差的关键阅读、练习、使用和完整作品；浏览量不是掌握度。", "practical_change": "每次学习记录阅读层次、一个可验证任务和一个尚未解决的边界问题。" }, { "original_idea": "不理解就说明当前不能继续。", "verdict": "需要修正", "updated_view": "前置引用允许先记住术语和关系，读完、重复、整理后再补理解；但不能把暂不理解永久合理化。", "practical_change": "维护“待回收概念”列表，并为每项记录原文位置、依赖和回收日期。" }, { "original_idea": "自学就是不找人、不提问。", "verdict": "需要修正", "updated_view": "自学强调责任和证据链；搜索、官方文档、实验、作品、同伴反馈和具体提问都是自学环节。", "practical_change": "提问前完成搜索和最小实验，提问时提供目标、症状、环境和最小复现。" }, { "original_idea": "做大项目才足以证明学会。", "verdict": "需要修正", "updated_view": "小而完整的作品更适合作为阶段验收，因为它能暴露依赖并形成闭环反馈；规模可以逐步扩大。", "practical_change": "给每个学习主题定义一个最小输入—处理—输出作品，并保存版本和复盘。" }, { "original_idea": "记不住主要是记忆天赋不足。", "verdict": "需要修正", "updated_view": "记忆力可训练，也可用笔记、整理、归纳、总结和主动回忆补偿；应区分存储、检索与组织问题。", "practical_change": "读完小节后不看原文写关键词、结构和例子，再回看纠错。" } ], "next_actions": [ { "action": "从今天的一个具体技能中选一篇主文档，先看目录、术语表和整体结构，再做一次只字不差的定义阅读。", "why": "把“先建地图、后精读”的方法变成可观察行为。", "evidence_required": "保存目录、关键定义、边界、例外和自己的重述。", "guardrail": "不要把首次通读或摘抄数量当作完成；必须附一个使用或验证任务。" }, { "action": "建立“前置引用/待回收概念”清单。", "why": "减少遇到不懂术语时的卡顿和重复搜索。", "evidence_required": "每项包含原文位置、依赖猜测、暂时记忆版本和后续理解。", "guardrail": "把猜测标为猜测，不把个人补全写成原文事实。" }, { "action": "为当前学习主题做一个小而完整的作品。", "why": "让学—练—用—造进入真实闭环，并暴露下一步薄弱点。", "evidence_required": "明确输入、输出、验收条件、版本和一次外部反馈。", "guardrail": "限制范围和预算，防止注意力漂移成无限扩展。" }, { "action": "把技术问题按“目标—环境—症状—已查—已试—最小复现—明确请求”写成草稿。", "why": "将聪明提问变成自学流程的固定接口。", "evidence_required": "附搜索词、文档链接、实验结果和仍然不清楚的具体句子。", "guardrail": "不暴露密钥、隐私或无关生产数据；不要未经验证就把猜测写成根因。" }, { "action": "用主动回忆复述今天的两条主线：只字不差地阅读，以及如何提出可回答的问题。", "why": "测试是否从“看过”进入了能解释、能使用的状态。", "evidence_required": "脱离原文写一段总结，并回答 review_cards。", "guardrail": "答案不清楚时回到原始章节核对，保留不确定性。" } ], "risk_guardrails": [ "书中的统计、年代判断、作者经验和带有绝对语气的断言属于作者论述或历史文本；不能自动当作当前心理学、教育学或技术社区的普遍事实。", "“只字不差地阅读”“先记住再理解”和“整理归纳总结”是学习策略，不是保证任何人必然学会的神奇方法；效果需要通过自己的时间、任务和复盘记录检验。", "“刚需幻觉”是作者的解释框架，不是对所有动机和职业选择的心理学诊断；选择技能仍要考虑机会成本、现实需求和可用预算。", "How To Ask Questions The Smart Way 的语气和社区规范具有历史语境；今天提问还要遵守具体平台、团队、隐私、安全和行为规则。", "最小可复现例子必须去除密钥、个人信息、生产数据和未授权内容；不要为了复现问题公开敏感材料。", "本条不是医疗、法律、投资或职业保证，也不提供自动交易、收益、仓位或市场预测。" ], "review_cards": [ { "id": "selfteaching-four-stages", "kind": "reading", "prompt": "请说出“学—练—用—造”四个阶段，以及每个阶段的验收方式。", "answer": "学是获得模型和材料；练是针对弱点稳定动作；用是进入真实任务；造是整合成可交付、可验证的完整作品。", "source_ids": [ "craft-practice" ] }, { "id": "selfteaching-forward-reference", "kind": "reading", "prompt": "遇到前置引用时，为什么可以先记住再理解？", "answer": "因为当前概念可能依赖后文或另一知识；先保留术语、关系和位置，继续推进，再通过重复和整理回收理解。", "source_ids": [ "craft-forward" ] }, { "id": "selfteaching-literal-reading", "kind": "reading", "prompt": "什么时候需要“只字不差地阅读”？", "answer": "面对重要定义、规则、接口、边界和例外时；普通娱乐性阅读可以粗读，但技能学习不能把所有材料都当小说浏览。", "source_ids": [ "craft-reading" ] }, { "id": "selfteaching-memory", "kind": "reading", "prompt": "记忆力不足时，除了死记硬背还能做什么？", "answer": "建立可靠笔记，并反复整理、归纳、总结和主动回忆；把记忆、检索、记录和组织分别处理。", "source_ids": [ "craft-forward" ] }, { "id": "selfteaching-practice-thinking", "kind": "reading", "prompt": "刻意练习和刻意思考分别解决什么问题？", "answer": "刻意练习修复具体执行弱点；刻意思考决定练什么、能怎样使用、怎样迁移和拆解。", "source_ids": [ "craft-practice", "craft-thinking" ] }, { "id": "selfteaching-budget", "kind": "reading", "prompt": "学习预算包含哪些东西？", "answer": "时间、精力、注意力、重复次数、失败次数和反馈机会；低估预算会把投入不足误判成天赋不足。", "source_ids": [ "craft-practice" ] }, { "id": "selfteaching-decomposition", "kind": "reading", "prompt": "横向拆解和纵向拆解有什么区别？", "answer": "横向拆同一层的并列任务；纵向拆目标、抽象、实现、工具和细节等层级。", "source_ids": [ "craft-difficulty" ] }, { "id": "selfteaching-complete-work", "kind": "reading", "prompt": "为什么小而完整比大而半成品更适合作为阶段成果？", "answer": "完整闭环能暴露真实依赖、产生反馈并证明交付能力；规模可以随后扩大。", "source_ids": [ "craft-thorough", "craft-social" ] }, { "id": "smart-question-template", "kind": "reading", "prompt": "请背出一个可用的技术提问模板。", "answer": "目标—环境—症状—已查—已试—最小可复现例子—明确请求；根据社区要求补版本、日志和最近变化。", "source_ids": [ "smart-questions" ] }, { "id": "smart-question-symptoms", "kind": "reading", "prompt": "为什么要先写症状而不是原因猜测？", "answer": "症状是可观察事实，猜测只是待验证解释；先写事实能让别人纠正假设并复现问题。", "source_ids": [ "smart-questions" ] }, { "id": "smart-question-rtfm-stfw", "kind": "reading", "prompt": "RTFM/STFW 的正确用法是什么？", "answer": "把它们当作先读手册、先搜索的提醒；若仍不懂，要指出具体看过的材料和剩余缺口，不能只用缩写打发人。", "source_ids": [ "smart-questions" ] } ] }