Christoph Buchner

> MSYS + MSVC (recommended by VS-Code) A pure-msys toolchain (i.e. without MSVC) seems useful (and from the PR so far seems to be planned 👍 ) Also, what about...

> Until there's movement on [conda-forge/conda-forge.github.io#1044](https://github.com/conda-forge/conda-forge.github.io/issues/1044), I think we'll be stuck with gcc 5.3.0, though (which is ~ the only reason why I'm using MSYS...) It seems that by now...

> > The runtime libs are due for a relatively [large](https://discourse.llvm.org/t/rfc-support-cmake-option-to-control-link-type-built-for-flang-runtime-libraries/71602) [overhaul](https://reviews.llvm.org/D154869) > > This landed now as [llvm/llvm-project@6403287](https://github.com/llvm/llvm-project/commit/6403287eff71a3d6f6c862346d6ed3f0f000eb70), but I realised one thing: we probably don't need to add...

I checked, no, this is something for @cjermain.

Cross-ref where lack of this is relevant when using FMPy: https://github.com/CATIA-Systems/FMPy/issues/303#issuecomment-891043884

> Hi, is this feature only supports Co-Simulation? I add `-d=fmuExperimental` in OMEdit > Simulation Setup > Translation Flags > Additional Translation Flags, but the exported FMU show that only...

Is it documented/planned somewhere when the `canGetAndSetFMUstate` capability could transition out of the "fmuExperimental" flag into support by default, or what limitation is blocking that transition?

> strongly depends from the numerical noise when both the inputs are closed to the same number (zero in this case) and the result may change by changing the tool....

> @ time 1.0999000000099904e-9 `plrToPul.tim.y` comes back to zero but the threshold stays fixed to TRUE Just to point out a detail that I'm not sure you're implying: according to...

Do you think it makes sense to plot/examine the outputs of **all** blocks involved in the loop here (for one of the failing examples), and check if any others disagree...