工具与资源gwern.net
Gwern 的互联网搜索技巧指南
Internet Search Tips
2018 — 2021，它一次次回到首页
黑点是上首页（≥ 50 分）的那次投稿，空心点是没冲上去的投稿，点越大分数越高。
导读
这是 Gwern 撰写的互联网搜索技巧指南，系统整理了高级搜索运算符、替代搜索引擎（如 Yandex）、学术论文与电子书的获取渠道、浏览器快捷键以及如何组合关键词等实用方法。文章基于作者长期检索冷门资料的经验，覆盖面广且细节丰富。
为什么经久不衰：搜索能力是长期有用的元技能，而文中许多技巧（尤其是替代引擎、冷门资料获取和 Google 的限制）至今仍不广为人知，值得反复查阅。
HN 评论精选
Yandex 在非政治搜索和图像搜索上可能比 Google 更少受限，但代价是零隐私。
这里有些不错的东西。不过文中只是非常简短地提到（简短到我一开始都没注意到）：在很多场景下用 Yandex 替代 Google 效果很好。作为俄罗斯公司，Yandex 无疑受到严格审查，但只针对俄罗斯政治相关的重要事项。讽刺的是，这意味着对非俄罗斯用户来说，它比 Google 的审查少得多——Google 近年来以政治、“政治”、DMCA 疯狂、“被遗忘权”等名义严重削弱了搜索。图像搜索尤其令人印象深刻。还记得 Google 图片以前能帮你找到冷门图片的源头吗？Yandex 仍然可以，而且还能做很多其他事，比如对文字较多的图片自动尝试转写文字。我的直觉是 Google 图片也具备很多这种能力，但要么对用户隐藏，要么为了阻止那些“邪恶的内容盗版者”而故意削弱。当然毫无隐私可言：假设你敲入又一个无聊搜索时，俄罗斯政府正在实时监视。但对某些用例来说这没问题。
ergot_vacation查看英文原文
Some neat stuff here. Only VERY briefly mentioned (so briefly I missed it at first) however: Substituting Yandex for Google is great for many use cases. Being Russian, Yandex is no doubt heavily censored, but only for things important to Russian politics. Ironically, this means that for non-Russian users, it's considerably LESS censored than Google, which has SEVERELY crippled its search in recent years in the name of politics, "politics," DMCA madness, "right to be forgotten" etc. The image search is especially impressive. Remember when Google Images used to give you actual results when trying to find the source of an obscure image? Yandex still does, and it does a bunch of other neat things too, like automatically trying to transcribe text from an image if it's text-heavy. My instinct is that a lot of this capacity exists in Google Images, but is either mostly hidden from the user or deliberately hobbled to stop the oh so evil content pirates. Zero privacy of course. Assume the Russian government is watching in realtime as you hammer in another inane search. But for some use cases that's fine.
Google 的高级搜索运算符可能触发封禁，且对冷门内容的检索能力较十年前下降。
不幸的是，“过度”使用高级搜索运算符可能导致你被 Google 封禁几个小时，你会遇到无限循环的 CAPTCHA。什么算“过度”似乎差异很大，但我曾因为用 site: 搜索某个冷门短语，只一次查询就触发了。对于冷门内容，Google 肯定比几年前或十年前差多了。我大约在 2010 年开始注意到这一点。
userbinator查看英文原文
Unfortunately, using the advanced search operators "too much" can get you banned from Google for a few hours, where you get an infinite series of CAPTCHAs. What counts as too much seems to vary widely, but I've triggered it with as few as one query for some obscure phrase using site: . Google is definitely far worse for obscure things than it was a few years or a decade ago. 2010 is roughly when I started noticing it.
补充了荷兰 narcis.nl 作为可免费获取大量旧学位论文的渠道。
> 硕士/博士论文：抱歉，如果是 2000 年以前的，大概没戏。如果你有大学代理，也许能从 ProQuest 拿到副本。否则你需要完整的大学馆际互借服务，即使那样也可能不够（令人惊讶的是，很多大学似乎只允许本校学生/教职员工访问，而大多数学位论文存储在缩微胶片上，这更增加了复杂性）。至少荷兰有 [1]，它索引了许多论文/学位论文，包括非常古老的，而且很多可以免费获取。[1] https://www.narcis.nl/
guidovranken查看英文原文
> master/PhD theses: sorry. It’s probably hopeless if it’s pre-2000. If you have a university proxy, you may be able to get a copy off ProQuest. Otherwise, you need full university ILL services, and even that might not be enough (a surprising number of universities appear to restrict access only to the university students/faculty, with the complicating factor of most theses being stored on microfilm). At least in the Netherlands there is [1] which indexes many papers/theses including very old ones, and many are freely accessible. [1] https://www.narcis.nl/
搜索词的选择最关键，可用宽泛结果发现更具体的词，并收藏高质量专业网站。
根据我的经验，最重要的是你选择的搜索词。有时想一想大多数人会如何表述一个问题会有帮助。结果往往太宽泛……所以用它们来选择更具体的词加入查询，以“聚焦”到你需要的结 果。（例如，加入年份甚至具体日期可以增加结果的聚焦度和质量。）保持一个不断增长的书签列表，收藏那些内容丰富的优质“专业”网站会很有帮助。
8bitsrule查看英文原文
Most important in my experience are the terms that you choose for a search. Sometimes it helps to think about how most people would phrase a question. Often results can be too broad ... so use them to choose more specific terms to add to your query to 'focus in' on the result you need. (E.g. adding years (even specific dates) can add focus -and quality- to your results.) It helps to keep a growing list of bookmarks for high-quality 'specialist' sites with -a lot- of content.
历次讨论
同类还值得读
直接在浏览器中启动 Linux 内核
Boot a linux kernel right inside your browser.
Fabrice Bellard 的 JSLinux 页面，用 JavaScript 在浏览器里启动一个 Linux 内核，并提供一个可交互的终端环境。
工具与资源上首页 5 次 · 2011–2025bellard.orgHacker News 未公开功能与行为清单（2018）
A List of Hacker News's Undocumented Features and Behaviors (2018)
数据科学家 Max Woolf 整理的 Hacker News 未公开功能与行为清单，涵盖投票机制、排名算法、账号惩罚等社区运作细节。
工具与资源上首页 5 次 · 2019–2025github.comMapSCII：在终端中用盲文和 ASCII 渲染地图
MapSCII – Braille and ASCII map renderer for the console
MapSCII 是一个在命令行终端里渲染地图的开源工具，用盲文字符和 ASCII 字符把地理数据画成可交互浏览的字符地图。
工具与资源上首页 5 次 · 2017–2025github.com顶尖大学的免费在线课程
Free Online Courses from Top Universities
Open Culture 整理的一份免费在线课程合集，汇集了全球顶尖大学（如 MIT、斯坦福、耶鲁等）向公众开放的网络课程资源，涵盖音频、视频等多种形式。
工具与资源上首页 5 次 · 2010–2022openculture.com淘气字符串：容易引发用户输入问题的字符串清单
Naughty Strings: A list of strings likely to cause issues as user-input data
一个 GitHub 仓库，收录了大量容易让程序出问题的用户输入字符串，如特殊字符、emoji、SQL 注入片段、系统命令等，供开发者用于测试输入处理逻辑。
工具与资源上首页 4 次 · 2015–2020github.com