The big thing is that I don't need byte strings in the result to be contagious in memory, so I don't want to require the input to be contagious in memory either.
Ok so I looked into it and from what I've gathered it looks doable. I think you could just install nix on the roboRIO like any other linux machine, package your robot program like any other nix package, and deploy with `nix copy`. You might need to package some of NI's libraries, but I don't think it would be that hard to write a derivation that downloads the rio image and extracts them.
Haskell stuff aside, this might actually be less work than whatever cursed docker thing the wpilib people use to build the cross toolchain. It would also make it (in theory) really easy to include any foss library that has a nix package in your robot program.
It would make setting up cross-compilers on other people's computers a lot easier. The alternative is using something like docker I guess.
the gist of it is that the `dunai` package is haskell is really cool, and is really well-suited for robotics programs, and I wrote some supporting code for frc stuff and figured out most of cross-compilation but got busy and stopped.
I think I'm suffering from "priority inversion". the "cool project" thread is scheduled, but it's spinning waiting for me to finish my "important work"
(Practical as in, the user interface is nice. For any large projects I imagine you couldn't actually run the compiler on the PineNote)
I wonder how practical it would be to use the PineNote as a laptop for programming/editing text, especially if you had an editor that did syntax highlighting in a more sophisticated way than just "shade the text a different color". I guess it might look more like when code snippets get typeset using LaTeX or something.
specifically thinking of this page: https://ncatlab.org/nlab/show/red+herring+principle
What's green and really far away?
The lime at infinity.
oh nevermind I forgot for algebras Hom(A,B) is just a module, not an algebra. crisis averted.
still pretty weird though
ahh I need to actually feed myself instead of trying to read nLab
that sounds like it would mean R-alg is "cocartesian closed", but according to https://ncatlab.org/nlab/show/cocartesian+closed+category that can't possibly be right...
dark mathemagician | 23 | they/them
Working on a math PhD, mainly algebraic geometry. I also do things with computers sometimes.