Disclosure: I'm the 'Dan' in the article :-) There are two distinct (but related) use cases for our the MinGW work. First for developers who work on Linux on a day-to-day basis, but want to make sure their code still compiles against Windows. Second, for developers who actually want to ship Windows binaries built with MinGW. For the first case, the lack of the security features in VS is not a problem, since it is basically just being used as a developer testing service. For the second case, clearly there is a downside to not have these features at all, but GCC itself does have potential to include some of them & more could be added if deemed important enough. Fundamentally though, if you want to work with a 100% open source toolchain (I do) and still provide support for Windows you're more or less limited to MinGW or Cygwin as options, MicroSoft VS doesn't come into the equation no matter what it can potentially do. If a Windows user really does want features that VS provides (& has shelled out for its license), then they have the freedom to ignore our MinGW binaries and build from source themselves.