安全与隐私robertheaton.com
Wacom 数位板会记录你打开的每个应用
Wacom tablets track every app you open
2020 — 2025，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
作者 Robert Heaton 逆向分析了 Wacom 数位板驱动，发现它会把用户打开过的每一个应用程序名称记录下来并发送到 Wacom 的 Google Analytics 账户，而这一行为藏在驱动设置深处、默认开启，且几乎没有向用户披露。文章既展示了具体的技术排查过程，也讨论了厂商为何想要这类数据、以及为什么偷偷摸摸地做才是问题所在。
为什么经久不衰：它把"厂商遥测"从抽象担忧变成了可复现的具体证据，加上那句"原则性的不"，多年来一直是讨论隐私与暗黑模式时的经典案例。
HN 评论精选
Linux 有社区开源驱动可规避追踪，但 Windows 缺少替代方案，值得追问原因。
在 Linux 上你用的是社区维护的驱动，所以我猜不会有这个问题（如果我错了请指正）。但据我所知，Windows 上没有对应的开源 Wacom 驱动。比我更懂 Windows 的人怎么看？是因为用 Windows 的人大概不在乎开源驱动、所以没有需求？还是 Windows 本身让替换驱动更难？Wacom 自己不提供 Linux 驱动，但看看 GPU、打印机驱动的现状，我隐约觉得就算他们提供了，Linux 社区也会有人做开源替代品。我在想有哪些 Windows 兼容硬件有第三方驱动可选——也许某些打印机？
danShumway查看英文原文
On Linux you'd be using the community-maintained drivers[0], so I assume this wouldn't be a problem (correct me if I'm wrong). But as far as I can tell, there's not equivalent Open Wacom drivers for Windows. People with more Windows knowledge than me: any thoughts on why? Is it just that someone using Windows probably doesn't care about Open drivers, so the demand isn't there? Or is there something about Windows that makes substituting drivers harder? Wacom doesn't provide their own Linux drivers, but looking at the state of drivers around GPUs, printers, I vaguely suspect that somebody in Linux would be working on Open alternatives even if they did. I'm trying to think off the top of my head what Windows-compatible hardware has 3rd-party driver options. Maybe some printers? [0]: https://linuxwacom.github.io/
遥测本身对开发者有价值，滥用它的厂商反而毁掉了正当收集的空间。
可悲的是，（合理的）遥测数据对软件工程真的非常有用。程序一大，bug 必然多到修不完；崩溃追踪和使用分析能让"修什么、忽略什么"变成数据驱动的决策，这对软件质量的提升是巨大的。我做过有遥测和没遥测的项目，没有遥测时感觉很荒谬——就像不知道车到底哪儿坏了，却随机去修后视镜。而那些过度收集、不作披露的厂商，恰恰毁掉了这个信息源，让想合规收集合理数据的开发者更难做。
Nitramp查看英文原文
The sad bit is that (some reasonable) telemetric data is really, really useful for software engineering. If you have a large enough program, of course it'll have way more bugs than you could ever fix. Crash tracking and usage analytics is how you make a data driven decision on what to fix, and what to ignore. This enables a data driven approach to software quality that's a huge improvement. Having worked on projects that did and did not have telemetrics, working without them feels absurd - it seems like you're just randomly fixing the side mirror on a car without any idea what's actually broken on it (independently of your overall testing posture). Vendors tracking excessive information without proper disclosure destroy this information source for those developers that try to collect reasonable information (with consent, disclosure, in context, etc).
退出选项藏在多层菜单里且默认开启，是典型的暗黑模式。
非常好的文章。但我好奇作者为何没提到驱动的隐私设置：打开 Wacom Desktop Center，右上角"Login"旁有"More"，点进去有"Data Privacy Settings"，里面"Participate In Wacom Experience Program"可以开关。我的设置竟然是"开"——而我发誓凡是程序、网站、安装器问我，我一律拒绝。所以这一定是恶意默认值加暗黑 UI，才让这个超隐蔽的选项对我开着。Wacom 真丢人！
DerWOK查看英文原文
Really great article. But I wonder, why the author did not cover if/what the driver publishes, If I do: - open "Wacom Desktop Center" - Top right (next to "Login") is "More" (click!) - "Data Privacy Settings" (click!) - "Participate In Wacom Experience Program" => on => off! My setting was "On" - and I swear: whenever a program/website/installer asks I go "No thankx". So it must be dark UI patterns with evil defaults that this super-hidden thing was "on" for me. Shame on you, Wacom!
收集数据有正当理由，但隐蔽的做法才是信任崩塌的根源，透明征询完全可行。
可悲之处在于，Wacom 想了解某些信息其实完全合理，比如客户把数位板接在什么硬件、用什么操作系统和应用，这有助于他们决定资源投向、测试哪些软件。但正因为做得如此鬼祟，才显得特别可疑。为什么不光明正大地问用户，并展示将要发送的内容？（我几乎可以肯定 Wacom 只是众多这样做的公司之一。）我记得几年前用过某个程序的崩溃报告功能：a) 明确需要选择加入，b) 在问我要不要看之后，逐字显示将要发送的数据（包括调用栈），然后再问一次是否发送。所以我知道不同的处理方式不仅可行，而且确实有人这么做。
krylon查看英文原文
The sad thing about this is that there are perfectly valid reasons for Wacom to want certain information, like what hardware do their customers connect the drawing tablets to, what operating systems and applications do they use. This is useful information to have, in the sense that Wacom can choose where to put their resources, what applications to test for, etc. But the very fact they are so damned sneaky about it makes it look really shady. Why not openly ask the users those questions and show them what information would be sent to the vendor? (And I am pretty sure Wacom is but one of many companies behaving this way.) I vaguely recall using some applications built-in crash report a couple of years ago, and it was a) explicitly opt-in, and b) showed me (after asking me if I wanted to see it), verbatim, the data it was going to send to the vendor, including stack trace and stuff like that, and then asked me again if I wanted to send that crash report. (Unfortunately, I do not recall what software that was, though. sad emoji) So I know it is not only possible to handle these things differently, but some people/companies actually do that.
历次讨论
同类还值得读
非法素数
Illegal prime
维基百科条目，介绍“非法素数”：某些素数之所以被称为“非法”，是因为它们经特定编码后，恰好对应一段被法律禁止传播的数据——最著名的是 DVD 解密程序 DeCSS 的代码，有人把…
安全与隐私上首页 5 次 · 2010–2021en.wikipedia.org编辑精选关于 /dev/urandom 的常见误区
Myths about /dev/urandom
（Thomas Hühn 撰写，2014 年首发并持续更新）逐条反驳关于 /dev/urandom 的常见误解：它与 /dev/random 使用的是同一个 CSPRNG，一旦完…
安全与隐私上首页 5 次 · 2014–20202uo.de编辑精选AES 高级加密标准火柴人图解
A Stick Figure Guide to the Advanced Encryption Standard (AES)
一篇用火柴人漫画形式讲解 AES（高级加密标准）算法的经典科普文章。
安全与隐私上首页 5 次 · 2009–2022moserware.com编辑精选指纹是用户名，不是密码
Fingerprints are Usernames, not Passwords
作者（eCryptfs 维护者、Ubuntu 安全工程师）认为指纹等生物特征是公开、不可更改的标识，应视作用户名而非密码，不能当作秘密凭据用于认证。
安全与隐私上首页 4 次 · 2013–2017blog.dustinkirkland.comHTTPS 连接的最初几毫秒（2009）
The First Few Milliseconds of an HTTPS Connection (2009)
逐毫秒拆解一次 HTTPS 连接建立时发生了什么，重点讲 TLS 握手、证书验证、RSA 签名与密钥协商等机制，面向不熟悉该领域的读者。
安全与隐私上首页 4 次 · 2009–2020moserware.com编辑精选