llvm-mingw VS WSL

> Doing Python with a C plugin, or just compiling a command line C/C++ isn't really systems programming.

I care about a minimal set of tools in order to compile C/C++ programs. thats offered by:

https://github.com/mstorsjo/llvm-mingw/releases

and also MSYS2, and even the Zig C compiler. all less than 200 MB. meanwhile Visual Studio installing about 10 GB worth. If Microsoft can offer a similar experience then I am interested.

Pick a community-supported Clang-based Mingw-w64 distribution.

> AFAIK there is no native GCC compiler for Windows

might want to check your facts before spouting nonsense. there is, and has been for many, many years. more than one in fact:

https://github.com/mstorsjo/llvm-mingw

https://packages.msys2.org/base/mingw-w64-gcc

MinGW is notoriously slow to link compared to MSVC, unless using llvm-mingw with the link=lld SCons option. If using MSVC, make sure to use 2022 or at least 2019 if possible – recent linkers tend to be faster than older versions.

GCC doesn't support windows/arm64, but you should be able to do it with LLVM. I've never gotten it to work myself, but should be able to supply a cross toolchain: https://github.com/mstorsjo/llvm-mingw

Visual Studio is a bloated mess, and has been for many years. Its at least 10 times larger than other options, such as MinGW-LLVM:

https://github.com/mstorsjo/llvm-mingw

Sadly Qt ships MinGW 8.1 which is positively ancient (released in 2018). If you're starting a new project (which you likely are if you are installing an IDE aha) there's no reason not to go for more recent compilers - msys2 has GCC12 (https://packages.msys2.org/package/mingw-w64-x86_64-gcc) and Clang 14 (https://packages.msys2.org/package/mingw-w64-x86_64-clang) which just work better overall, have much more complete C++20 support, have less bugs, better compile times (especially clang with the various PCH options that appeared in the last few versions), better static analysis, etc.

Personally I use https://github.com/mstorsjo/llvm-mingw's releases directly which does not require MSYS but that's because I recompile all my libraries with specific options - if the MSYS libs as they are built are good for you there's no reason not to use them.

you can grab it here: https://github.com/mstorsjo/llvm-mingw/releases/tag/20211002

Clang (and lld) do support native TLS, and mingw-w64 does have the things that are needed. I think binutils also might have what's needed too, but AFAIK the thing that's missing is support for it in GCC.

Actually, (upstream) Clang defaults to native TLS instead of emulated TLS. In MSYS2, Clang is overridden to use emulated TLS by deafult to interoperate better with GCC built code and libstdc++ though.

The toolchain I maintain, https://github.com/mstorsjo/llvm-mingw, defaults to native TLS throughout.

It absolutely 100% can be true.

As an example: Windows Services for Linux 2 used a special init daemon to interact with the host OS.

That meant no systemd. That meant that the `systemctl` program wasn't there.

This baffled legions, armies, of wannabe sysadmins.

https://stackoverflow.com/questions/55579342/why-systemd-is-...

https://superuser.com/questions/1785697/systemd-in-wsl-on-wi...

https://github.com/microsoft/WSL/issues/9477

https://askubuntu.com/questions/1132230/unable-to-run-any-sy...

People on the whole have no idea how this stuff works, and they just copy magic incantations from StackOverflow to get stuff to happen. If that doesn't work, then this OS is broken. The end.

For these guys, WSL was broken.

Result:

MS hired Lennart Poettering.

https://www.theregister.com/2022/07/07/lennart_poettering_re...

He "fixed" it. Systemd now works in WSL2. All those guides for noobs now work. Everyone is happy.

In a world where tools like Flatpak and Snap are proliferating and it's driving deep divisions between Linux distros, if you think the average person struggling with Linux is going to use `ldd` to work out where the dependencies for something live, I'm afraid you are a deep guru who lives on a different plane of existence.

We now have widely-used packaging systems which simply embed an apps entire dependency tree into a package to avoid people having to work out the difference between `apt` and `rpm`. Thousands of terabytes of disk are being burned to make this stuff go away.

Yes, this is too hard. Way too hard.

Better source: https://github.com/Microsoft/WSL/issues/873#issuecomment-425...

One of the bugs where on the Docker side. As I have said, there have been several since release with a lot of impact period overlap. The latest and greatest is not resolved.

It's the issue of abysmal I/O performance in communication between the mounted WSL2 virtual hard disk and Windows mounts inside the WSL2 distro.

I found this github page and I was able to reproduce this from the answer

* Docker engine and not Docker Desktop in a VM. WSL2 works well after some configuration: https://github.com/microsoft/WSL/issues/6655#issuecomment-11...

Seems to more of a Defender issue than a WSL one, see https://github.com/microsoft/WSL/issues/8995

After adding exclusions for the fsnotifier-wsl process and and both variants of the WSL distro path my disk performance was improved.

Adding the idea64.exe process also helped since I was trying to run IntelliJ against projects inside WSL.