Fabrizio Ferrai

> Whatever decision we make for Spago should be mirrored to Pulp Should it? If I understand correctly our aim in implementing the publishing flow now is to (1) validate...

> The latter transforms record updates to record literals, which is a JS performance gain, but a performance loss for another backend I was also wondering about this. It might...

> To me, this sort of feels like 'language extensions' but for backends Could you elaborate more on this? The main issue I see with language extensions (as they are...

We now consider https://github.com/cdepillabout/purescript2nix/issues/4 a priority, so I'll pivot this issue to "Spago can use the new package sets"

I'll move this issue to Spago's tracker

This issue is now about tracking the first experimental/draft implementation of https://github.com/purescript/spago/issues/842

Things ended up going in a totally different direction, and the most update prototype is: https://github.com/f-f/spaghetto

This said - I think the best course for integrating nix would be for Spago to create a lockfile, much like a lot of the existing `*2nix` solutions do. I...

In the issue mentioned just above this comment I had the thought to bundle zephyr into Spago, but it makes so much more sense to have it in the compiler...

Some more notes on your points: 1. The main thing we need to support for this is the notion of "targets", and indeed there's #681 open to track this change...