Daniel Probst

Yes, It is a suboptimal solution. I got the current version of TMAP running on *some* Windows installs but not on others. I'm currently thinking about writing a small library...

Hi @liudongliangHI, thanks for the info. I bumped `faerun-python` to version 0.4.2. This bug is now fixed. I introduced the bug because I'm testing a pip installable version of `tmap`...

Hey @meddwl, it's a bit tricky to say--it looks like the package didn't install correctly, could you share more details on how you install it? Did you use: ```bash conda...

Dear @kr-hansen and others, I will have time to work on the project towards the end of May for a couple of weeks. Getting it to be redistributable using pip...

I managed to get both ogdf and tmap into a single wheel (see the development branch action artifacts). The only issue I have to work out is a SIGILL erro...

Update: Had to build OGDF with `-DOGDF_MEMORY_MANAGER=MALLOC_TS` to get it working (@N-Coder, see pyproject.toml and wheels.yml in actions to see how to get packages for manylinux, win and macos). `pip...

I did some more debugging on the `SIGILL` issue and it seems like the missing instruction set extensions are the AVX-512 ones (VL and/or F, as `SIGILL` is thrown on...

Quick update: Adding `-mno-avx512f` to the [build flags](https://github.com/ogdf/ogdf/blob/7596a518d2b8a964523c2bbd86af360aa405c6cf/cmake/compiler-specifics.cmake#L18) fixes the `SIGILL` issue on the cloud services. I will check the performance impacts on TMAP next.

Yes, especially with M1 et al becoming more common. I didn't have time yet to look at the performance--what do you reckon the impact is?

Thanks for the info, I have been a bit busy preparing papers for conferences, so I'll take a look once I can get back to maintaining TMAP. Hatch looks really...