I dug into this and it turns out there's an active, multi year effort underway to solve exactly this, along the exact ways you´d expect, with references to guix and bootstrappable.org etc, making steady progress: https://discourse.haskell.org/t/what-s-needed-to-bootstrap-g...

Reading through that thread gives me a very different idea of the state of haskell than I got from reading your comments.

▲

lrvick 4 days ago | parent [-]

Efforts pop up every once in a while, usually with no results. As I said elsewhere, if they actually pull it off this time then my tone changes to one of willingness from my team and I to put in the work to use these efforts to support deterministic multi-signed builds in stagex.

It would make GHC (and pandoc) something we could have in tree which would be awesome. I am just annoyed it seems like such a low priority.

	▲	nothrabannosir 4 days ago \| parent [-]
		This entire thread, including your original comment, isn’t really about your efforts though. It was about Haskell. As I said : I read your comments, formed an idea based on that, dug deeper to see for myself, and got a completely different picture. Whether that is because of how you write or because of how I read I’ll leave for you to decide.