2ship2harkinian/CONTRIBUTING.md

Contributing to the Majora's Mask Decompilation Project

Thanks for helping us reverse engineer The Legend of Zelda: Majora's Mask for the N64! All contributions are welcome. This is a group effort, and even small contributions can make a difference. Some tasks also don't require much knowledge to get started.

This document is meant to be a set of tips and guidelines for contributing to the project. For general information about the project, see our readme.

Most discussions happen on our Discord Server where you are welcome to ask if you need help getting started, or if you have any questions regarding this project and other decompilation projects.

Useful Links

• Discord - Primary discussion platform.

• Google Sheets reservation board - We use this to track decompilation progress, not GitHub Issues.

• Installation guide - Instructions for getting this repository set up and built on your machine.

• Code Review Guidelines - These are the guidelines that reviewers will be using when reviewing your code. Good to be familiar with these before submitting your code.

• Zelda 64 Reverse Engineering Website - Our homepage, with links to other resources.

• Introduction to OOT decomp - A very detailed tutorial on how to get started with decomp. For OOT, but largely applicable to MM as well.

• The #resources channel on the Discord contains many more links

Getting Started

What should I know to take part?

Basic knowledge of C, particularly arrays and pointers, is extremely useful. Knowledge of MIPS is not required initially, you can pick it up as you work through some decompilation.

You should be familiar with using git and GitHub. There are a number of tutorials available online, such as this one which can help you get started.

The most useful knowledge to have is a general understanding of how the game works. An afternoon of constructive mucking about in the OOT Practice Rom (aka GZ) or the MM Practice Rom (aka KZ) will be very beneficial if you have not looked at either of the N64 Zelda's internals before. The OOT Decompilation Project is farther along than this project, so it can also be a great resource.

This project only uses publicly available code. Anyone who wishes to contribute to the OOT or MM projects must not have accessed leaked source code at any point in time for Nintendo 64 SDK, iQue player SDK, libultra, Ocarina of Time, Majora's Mask, Animal Crossing/Animal Forest, or any other game that shares the same game engine or significant portions of code to a Zelda 64 game or any other console similar to the Nintendo 64.

Environment Setup

Get started by following the installation guide in the readme. You should be able to build a matching ROM before you start making any changes.

First Contribution

Usually, the best place to get started is to decompile an actor overlay. An actor is any thing in the game that moves or performs actions or interactions. This includes things like Link, enemies, NPCs, doors, pots, etc.

You should join the Discord to say hello and get suggestions on where to start on the #mm-decomp channel.

We track who is working on what on the Google Sheets reservation board. After joining the Discord, open the Google Sheets reservation board so you can claim your code file and avoid duplicate work.

The work flow is: Reserve a file, decompile it, submit a PR, and then repeat while addressing review comments. The expectation is that one reservation goes to one file which ends up in a one file PR.

Please note that unless it is communicated beforehand you will be expected to fully complete the file if you reserve it on Google Sheets reservation board. Exceptions are always easy to approve if it's communicated to the team. Real life circumstances can prevent someone from finishing. In these cases they should link their repo/branch in the Google Sheets reservation board and unreserve immediately. Communicate any issues with your reservations as early as possible.

Style Guide & Conventions

Most of the C formatting style is enforced by the format.sh script which is based on clang-format. Running ./format.sh will apply our standard style to all .c files in the repository.

There are some conventions that cannot be automatically enforced.

Naming Scheme

The following overlays are good examples of our naming conventions:

• ovl_Dm_Ravine
• ovl_En_Jc_Mato

These files demonstrate the following:

• Word order in names are from least-to-most specific (DM_RAVINE_STATE_ACTIVE, not DM_RAVINE_ACTIVE_STATE)
• Functions, structs, unions, enums, and typedefs are TitleCase (DmRavine)
  • "Methods" for objects separate the object from the verb with an underscore (DmRavine_Init)
• Variable names are camelCase (actionFunc)
  • Global variables start with g (gSaveContext)
  • Static global variables start with s (sSphereInit)
• Macros and enum constants are SCREAMING_SNAKE_CASE (DM_RAVINE_STATE_ACTIVE)
• Public structs, enums, and constants should be defined in the header file in the following cases:
  • The primary Actor struct (DmRavine)
  • The actor's InitVars, as an extern symbol (Dm_Ravine_InitVars)
  • Types used in other public declarations, like actionFunc typedefs (DmRavineActionFunc)
  • Macros and enums for accessing & setting actor params
  • The struct/enum/define is needed by other files
• Trailing commas in array and struct definitions (see EnJcMatoDamageTable)
• Constants converted to whichever looks best in context: hexadecimal, decimal, or float
  • Rotation angles should always be in hexadecimal
• Define macros for bitwise access to actor->params

NON_MATCHING & NON_EQUIVALENT

If you make significant progress on decompiling a function, but can't get it to match perfectly, you can use a NON_MATCHING block to commit your code but have it excluded from the build, like so:

#ifdef NON_MATCHING
void CollisionCheck_SpawnWaterDroplets(GlobalContext* globalCtx, Vec3f* v) {
    // ... 
}
#else
#pragma GLOBAL_ASM("asm/non_matchings/code/z_collision_check/CollisionCheck_SpawnWaterDroplets.s")
#endif

Before using NON_MATCHING, first try to use the decomp-permuter tool to find a closer match.

NON_EQUIVALENT can be used with the same syntax as NON_MATCHING, but it is used to mark sections of code which do not match and do not have the same behavior as the original code.

Matching vs. Documenting

Usually, the first step of decompiling a section of code is to get it matching: to produce a C version of the code that can be compiled into an identical ROM.

However, the goal of this project is to produce a codebase that can be understood and modified. So, beyond producing matching C code, the next steps are documenting the code.

Documenting is more than just adding comments. Documenting also includes:

• Renaming functions, variables, and struct members
• Using (or adding) constants, enums, and macros when possible
• Explaining sections of code that are not straightforward

Overlays are not required to be documented at this time, but code/ and boot/ should be documented. When documentation on a file has been started it should be as complete as possible.

For right now, object segment symbols should not be documented/renamed. These will be given names when object reconstruction is ready. However, it can be useful to add comments with name suggestions.

Pull Requests (PRs)

Checklist

Before opening a PR, walk through the following steps to ensure that your code conforms to the style guide and conventions.

• ./format.sh was run to apply standard formatting.
• make successfully builds a matching ROM.
• No new compiler warnings were introduced during the build process.
  • Can be verified locally by running tools/warnings_count/check_new_warnings.sh
• New variables & functions should follow standard naming conventions.
• Comments and variables have correct spelling.

Feel free to reach out on the Discord if you have any questions about these steps or encounter any issues.

Pull Request Process

After opening a PR, the Jenkins server will verify your changes. If there is an error, double-check that you can successfully make clean && make locally. If the build is OK, the next thing to check is that all added/modified files were git add-ed to your commit. The final check before posting on Discord for help is that there are no new warnings added to the code causing Jenkins to fail. You can check this by running: tools/warnings_count/check_new_warnings.sh.

Each PR needs a review from one reviewer and the project lead.

The PR author marks the comments as resolved: the commenter may not have permissions to do so. Once all comments are addressed, it is courteous to ping the reviewer on either Discord or GitHub via re-requesting a review.

After all the comments are addressed and at least one contributor has approved the review, the project lead can then review and merge the code. The project lead is also responsible for ensuring that all of these procedures are followed.

Throughout the PR process, you (the author) should update the row on Google Sheets reservation board with the appropriate information as the decompilation process progresses.