But isn't this part of the point the author of the article is trying to make? In case of Superfish it was United States Department of Homeland Security who advised uninstalling it and in case of Sony's rootkit, a class-action lawsuit. In both cases higher authorities hired by people to protect them, protected them. Not FSF, nor the fact that the code wasn't GPL.
Even if Superfish was GPL licensed, the issue would've persisted. As article mentions even laptops coming with Linux pre-installed hardly come with the Kernel and userspace tools source code and you need to download them separately. Should Lenovo have preloaded Superfish as Linux kernel module and shipped it to its consumer laptops, United States Department of Homeland Security would have still advised removing it.
A bit unrelated but I wrote https://github.com/amir/hepccn to list TCP connections by domains rather than just IPs. It filters connections to port 443 and 8443, downloads the peer certificate and then extracts the subject name (pretty sure I learned this trick from an article on Julia Evans' website but can't find it now). I wanted to add more functionalities to it but I stumbled upon an issue and that's for when SNI is enabled. I guess there's no way for me to recover the hostname the client has sent at the start of the handshaking process without introducing MITM, right?
> Editor tooling is sub-par (especially compared to language communities like Java and C#) and finicky - we often end up just compiling in a separate terminal.
I'm a subscriber but I don't look at it as a subscription to a weekly magazine but an investment. Enabling knowledgeable people to write in-depth analysis of various patch sets or subsystems of Linux kernel—which rarely change dramatically—ensures that, later on, there'll always be an article in Google search results that'll explain the reason behind certain behaviours in a more approachable manner than reading the source code or doc.txts—they're invaluable but a bit intimidating.
Policeman was the closest thing —that I know of— to uMatrix for Firefox users, but —at least for me— Firefox is always complaining that Policeman is slowing down the browser. And also, it's nice that you can easily import your Chrome uMatrix rules to Firefox.
I just thought you might want to know that web applications hosted on GAE are not accessible to Iranians, and presumably to people of other countries in OFAC sanction list.
They open sourced some of their stuff though, e.g., the Ohcount project which uses Ragel FSM for counting LOC in a variety of languages (what generates most of statistics in Ohloh).
At the first I thought I'd better clean-up the code, fix some bugs, add features requested here, and add some comments before releasing the code, but I then thought it'd better to release the code as it's in the released build.
It doesn't require to read your phone state, I forgot add a minimum SDK requirement, and Android adds that automatically for some compatibility reason.