Feature request

[ost-guy]

RFC: mintboot — A Compilable GRUB2-based Bootloader with Rust Tooling

1. Title
   mintboot — A custom GRUB2-based bootloader with Rust tooling

2. Status
   Draft / Proposal

3. Authors
   [Your Name / GitHub Handle]

4. Motivation
   The purpose of mintboot is to provide a custom bootloader with a Mint-themed user interface and multi-boot support. The project should remain fully compatible with existing Linux distributions and maintain the reliability of GRUB2.

A key requirement is that the project must be compilable, with a clear build process for both the bootloader core and the auxiliary tooling.

5. Scope
   The project scope includes:

• A GRUB2-based bootloader core (C)
• Custom configuration and installation tooling (Rust)
• A Mint-style theme and branding
• Multi-boot support using os-prober
• Clear and reproducible build instructions

6. Design is linux mint design

6.1 Bootloader Core (GRUB2 in C)
The bootloader core remains GRUB2. This ensures:

• BIOS and UEFI support
• Kernel loading
• os-prober integration
• Compatibility with existing systems

The GRUB2 fork may include minimal branding changes (name, theme defaults) while preserving core functionality.

6.2 Tooling (Rust)
Rust is used for the custom tooling, including:

• generating /boot/mintboot/grub.cfg
• installing mintboot to /boot/efi/gmint
• copying theme assets (fonts, backgrounds)
• automating os-prober execution and menu generation

The tooling must be designed for safety, maintainability, and ease of use.

7. Build Requirements
   The project must include a build system for both components:

7.1 GRUB2 Build

• Provide build instructions for compiling GRUB2 from source
• Include configuration options for the mintboot bootloader ID
• Output should be installed to /boot/efi/gmint/

7.2 Rust Tooling Build

• Provide Cargo-based build instructions
• Output binary should be named mintbootctl
• The tool should support installation and update commands

7.3 Reproducibility
The repository should include:

• a build script or Makefile
• optional Docker or GitHub Actions workflow for reproducible builds

8. Advantages

• Full compatibility with GRUB2 ecosystems
• Clear separation between bootloader core and custom tooling
• Safer and faster development of custom features using Rust
• Compilable and maintainable project structure

9. Conclusion
   mintboot aims to be a practical and compilable custom bootloader solution that leverages GRUB2’s proven capabilities while using Rust for modern tooling and automation. This hybrid approach ensures compatibility, maintainability, and ease of development.

[fosslover306]

What does this provide other then being a grub2 rust rewrite?

Asking any Linux distro to make a custom boot loader only for them selves and nobody else is a huge ask, and how much time do spend staring at your computers boot screen, between the nice animated spinner and the fact I only reboot when nessary usally every 3-9 months.

Lookup how Ubuntu's sudo rust rewrite went, now imagine Linux Mint doing it with 1 tenth the resources, do you think some magic is going to make everything go perfectly?

please not I do not care what programing language anything uses, computers are just a fun but critical tool for me.

[ost-guy]

I have mental insane ideas im autistic level 1

[fosslover306]

Insane is the correct description of what you want, the main problem is with these old code bases is that they started as a experimental idea, which then became the standard which usally results in poor documention, so overtime as stuff improves and the original designer either forgets, dies, or retires then basicly noboby knows how and why the code base works as a whole, and devs just fix what breaks and work with the quirks.

Unfortantly that little problem renders most rewrites challenging at best, almost impossible at worst because the replacement needs to behave the exact same way just to make an updated modern code base that can take it's place be viable, then get everyone to use it and stop melting down over stuff that either never changed, or matter. And after that pr disaster is over only then can improvments get made to it to make it better then the orriginal but not so fast any software get broken.

Navigate that minefield of problems, ego, lazyness, edge cases, and concerns of all kinds and the reward is a modern replacment written in rust with memmory safty, automatic documention, and easyier mantencince.... atleast 5 years later.

So like I said your idea is completly insane.

Just getting simple one line security and privacy improvments to be excepted is basicly immpossible, and that is with programs that already exist, and shell scripts I made that only needs to be run once for said program to do it's thing without any forseeable future maintince.

[ost-guy]

oh popular problem

[fosslover306]

Pretty much, for examples how well even simple improvements are considered just take a look at my open ideas in this project.