Robert Hensing

They're [live](https://flake.parts/options/process-compose-flake#opt-perSystem.process-compose._name_.cli.options) again. > may have to remove it Of course you'll be free to do so, and conversely I may have to pin it again, but until either of...

The duration numbers are good, but the throughput on @RossComputerGuy's mac looks a little worrying; it appears to be half as efficient (`user`). This could be explained somewhat by the...

Note that Determinate Systems has already [blogged](https://determinate.systems/posts/parallel-nix-eval/) about this, which is fine (that's just part of content marketing, and it's more than ok to be excited), but let's recognize that...

> 20 │⠀⠀ It looks like this axis is meant to represent the relative cost of using extra threads, however > (.max * (.parameters.num_threads | tonumber) This overestimates the consumption....

That's not normal and likely caused by race conditions that this PR hasn't addressed yet.

I would expect the process to be sufficiently chaotic that the number doesn't really matter.

> * Infinite recursion detection through blackholing is currently disabled. Relevant paper with a design that implements this. - [Haskell on a Shared-Memory Multiprocessor](https://simonmar.github.io/bib/papers/multiproc.pdf) by Tim Harris and the Simons....

> > Blackholing > > easy to provoke, or a nice-to-have? Easy to provoke and very important for users.

A lock file shouldn't be necessary as long as the input is going to be committed to the same repository. It only becomes a problem when the reference points outside...

@zimbatm Disallowing `..` in relative paths forces the flakes to form a tree rather than a DAG. Suppose you have a repo with flakes with dependencies `A -> P ->...