Anonymous Maarten

I ran some experiments last year, but it only works in C++20 mode and is theoretically IFNDR https://github.com/madebr/cmake-symbol-test It's worth to explore Icculus' suggestion about parsing link errors.

I tested that approach earlier, but discarded it because it also detects symbols that are available in the libc library but not available in the headers, potentially causing compile errors...

> Would it be straightforward to also disable the colorization [if environment variable `NO_COLOR` has a non-empty value](https://no-color.org/)? SDL3 is also "batteries included" :D So that's no issue at all...

> Would it be straightforward to also disable the colorization [if environment variable `NO_COLOR` has a non-empty value](https://no-color.org/)? The tests now support this (as long as they parse the arguments...

The environment variables are the values of the hint macros, so you should set `SDL_WINDOWS_DISABLE_THREAD_NAMING` instead.

> Try this out. Also please don't run binaries, randomly posted on forums. **edit** Looks like GitHub flagged and removed this post themselves.

> [@madebr](https://github.com/madebr) > Did that fix your issue? > [@madebr](https://github.com/madebr) His personal page is 404 now. Yup, we scared him away 🤣

What about this: ``` libc: glibc: ["2.18", "2.19", "2.20", ..., "2.31"] musl: ["0.9", "1.0", "1.1", "1.2"] compiler: gcc: libc: &libc (None for undefined or default) llvm: libc: &libc Visual Studio:...

Yes. Is that expected?

Thanks, so we'd need to inject a header at the top replacing the direct call with dynamically loaded ones.