Ben Gamari

Yes, that makes sense and indeed it looks like that is what is happening here. > The solution would then be to make the rule directly build them. I'm not...

Whoops, indeed it look like we came to the same conclusion @snowleopard; your comment didn't show up until I had already submitted.

> The right solution seems to be to create a separate build rule for `integer-gmp/package-data.mk`. If there are no better ideas, I could implement this (in a few days, as...

> @bgamari One possible solution is to move libraries/integer-gmp/gmp/config.mk into _build directory, which is probably the Right Thing to do, but that will require a change to integer-gmp's configure. Perhaps,...

Andrey Mokhov writes: > @bgamari Thanks! I assume this is not a blocking issue for merging Hadrian into GHC, so I've assigned it to the `GHC Transcendence` project (feel free...

> Adding a memory resource is fairly straightforward. Would you like to try this? I'm happy to help. I don't expect any difficulties, but you have a better understanding of...

Andrey Mokhov writes: > > @bgamari @ndmitchell What do you think? This is not an automated > solution, but will probably allow Hadrian users to relatively easily > tailor GHC...

I started a very naive [approach](https://github.com/bgamari/openscad/tree/parallel) to introducing parallelism into `GeometryEvaluator`. Unfortunately it seems that CGAL isn't entirely [thread-safe](https://sympa.inria.fr/sympa/arc/cgal-discuss/2012-05/msg00001.html). Unfortunately I've run out of time for now.

@terrylm sadly they have they not. `Handle_for` remains thread unsafe.

This looks like a great start! Thanks!