Skip to content

Latest commit

 

History

History
196 lines (141 loc) · 12.2 KB

CONTRIBUTING.md

File metadata and controls

196 lines (141 loc) · 12.2 KB

Introduction

What is Cube Cobra?

The main goal with Cube Cobra is to create the best cube management tool. I want to create a platform that is easy to use, that still has advanced features that allow users a high degree of freedom to organize and analyze their cube in a way that makes sense to them. I want to create the best possible platform for users to build, playtest, and share their cube.

Why contribute?

One of the best ways to create a sustainable software project is open source it. For cubecobra to become the best possible tool for cube managment it needs help from the community. By contributing to the project, you are giving back to the magic community. Maybe there's a lower priority feature you want to use or maybe there's just a bug that's been bothering you, either way contributing yourself is a great way to get it into the project. Open source software is often more reliable (more eyes on the code).

Don't feel nervous about making your first contribution, we accept developers with all levels of knowledge and will happily work with and help with whatever you need to make your first contribution. We are not code snobs, so don't be afraid to reach out with questions of any kind.

What kind of contributions do we need?

The issues tab in github is kept up to date and tagged, but any contribution is appriciated from code comments to experimental features. Just keep in mind completing a feature does not guarantee it will be merged. If you want to complete a feature that you don't see in the backlog, communicate with dekkerglen to make sure you don't waste any effort.

Ground Rules

Code of conduct.

Be a decent person. Copy pasted from the linux code of conduct: In the interest of fostering an open and welcoming environment, we as contributors and maintainers pledge to making participation in our project and our community a harassment-free experience for everyone, regardless of age, body size, disability, ethnicity, sex characteristics, gender identity and expression, level of experience, education, socio-economic status, nationality, personal appearance, race, religion, or sexual identity and orientation.

Characteristics of an ideal contributor

  • Creates issues for changes and enhancements they wish to make.
  • Discusses proposed changes transparently and listens to community feedback.
  • Keeps pull requests small.
  • Is welcoming to newcomers and encourages diverse new contributors from all backgrounds. See the Python Community Code of Conduct.

Contributing to Cube Cobra does not entitle any contributor to compensation of any kind. Contributions are made at will, with the goal of improving the tool for the entire community. Cube Cobra's hosting costs are paid through donations and affiliate links. These funds are managed solely by Gwen Dekker, who is solely in charge of hosting and managing the live site.

Your First Contribution

Issues are tagged with 'good first issue' if we think it's a good beginner task to tackle. Complete a few small changes to become familar with the codebase before diving into a huge feature.

How we collaborate - asynchronous communication

Our community of contributors is large, and growing fast. We don't have regularly scheduled working hours dedicated to Cube Cobra. Thus, we can't simply message each other on Discord to understand the current state of the project. Discord is a great tool for synchronous communication, but it falls short at allowing community members to discover conversations they didn't participate in. Our project also requires a focus on asynchronous methods of communication. The goal is for the current state of the project, including features being worked on, issues not yet resolved, and near-term roadmap plans, to be fully discoverable by a passerby without needing to chat in realtime with anyone else.

GitHub projects, issues, and pull requests serve this purpose beautifully. We treat the dekkerglen/CubeCobra repository as the source of truth about what work is complete, in progress, and not yet started. If you want to know whether someone is working on a feature, for example, or if anyone else has noticed the issue you're seeing, GitHub pull requests and issues are the first place you should look. We use GitHub this way because it has great support for discovering long-lived tickets even years after the fact. Put simply, it is a system of record. This practice, when adhered to vigilantly, leads to increased collective productivity and decreased blocking between community members and development efforts.

How we use GitHub issues

GitHub issues are the source of truth about known issues and planned features. If a planned feature doesn't have an open issue, it's not a planned feature. If a bug doesn't have an open issue, it's unknown and should have one.

When we file issues, we include links to related issues and pull requests in the issue description. GitHub makes linking to related issues and pull requests easy by automatically expanding the text #1234 into a link to issue 1234. These links provide a trail of breadcrumbs for community members to follow when learning about the group's thoughts on a given bug or feature request.

When we notice a bug, we first look for an existing open issue that references the bug. If we find one, we comment on the issue or use a GitHub reaction emoji to indicate that we have also noticed the bug. If we don't find one, we open a new issue about the bug. When we notice a change in a bug's behavior, or a new case in which the bug can be replicated, we leave comments to that effect on the issue.

When we begin actively working on a feature, we comment to that effect on the corresponding issue. This comment of "I'm working on this" indicates to other community members that this feature is covered, and to find something else to work on. Since GitHub does not allow community members to assign themselves to issues, we comment in this manner to make it clear to the community what work is currently in progress, and who owns that work.

How we use GitHub pull requests

Pull requests are the primary way that we share our work, both finished and in progress, with each other. When we open pull requests, we're communicating to the community that we are actively working on code changes, and we have some code to show for it. Though pull requests are not the source of truth about bug/feature ownership (issues get that title), they are a very useful tool for keeping community members abreast of one's progress without sending an @-everyone message in Discord.

When we're working on pull requests over the course of multiple days, sometimes we open [WIP] (work in progress) pull requests. We clearly indicate that these are WIP in the description to avoid premature code reviews from the community. When we do this, we're communicating that we'd like others to be aware of the details of our progress without requesting a detailed code review just yet.

When we feel that a feature we're working on is "code complete", we open pull requests. These pull requests have titles that reference the substantive changes they contain, descriptions that briefly summarize the changes, and sometimes include more detailed breakdowns of the architectural or coding approach. The goal of these descriptions is to aid in the code review process, making it easier for a community member with no prior knowledge of the pull request to perform a review. We also comment on our own pull requests asking for code review, sometimes even tagging specific individuals from whom we'd like a review.

Our pull requests have automated checks run against them, including code style linting and unit tests. When we request reviews on pull requests, we ensure that these checks are passing beforehand. When we add new functionality in a pull request, we also add new tests exercising that functionality in the same pull request. When we fix a bug in a pull request, we also add a unit test or adjust an existing one in the same pull request to prove that the bug has been fixed.

When we notice open pull requests with comments indicating they're ready for review, we review each other's code. The primary goal of our code reviews is to ensure that the code does not break existing functionality. Some secondary goals include ensuring that the code fully satisfies the related feature request or fixes the related bug, or that it conforms to best practices for code efficiency and style.

We review each other's code using the GitHub "review" workflow, especially its line-commenting feature. We leave comments on specific lines that we notice problems with. We also make liberal use of links - to lines of code, other issues and pull requests, or external documentation - to strengthen the points we bring up in our code reviews. We do this also to leave a trail of breadcrumbs explaining our thought process for future readers of the code review, including our future selves.

When a pull request is closed or merged, there are sometimes threads left hanging that require more work to be done. When this happens, we open issues tracking that remaining work.

How we use git branches

The master branch is the branch with changes that are pending for the next update. The release branches are forked from the master branch, and the latest release branch is the branch the server runs off of. First, assign yourself to an issue, and create a fork of the code. When you are finished with the feature, create a pull request back into the master branch. An admin will review your code and merge if it is accepted. Expect some comments and feedback.

How we use Discord

Our community collaborates in an asynchronous manner because it minimizes blocking and conflicts between various threads of work. We manage our known bugs, feature requests, and work in progress using GitHub's asychronous tools. Even so, there are some times when it's useful to have access to synchronous communication with other community members.

Our community's most important use for Discord is interaction with non-technical users of Cube Cobra. Discord is a great way to get to know the needs and wants of the user base, and can inform plans for features and bugfixes.

Another good use of Discord for contributors is staying abreast of the current state of the production server. If there are new bugs being reported frequently that require rapid response, a Discord "war room" can be the best way to handle these reports.

Our community does not treat Discord as a system of record. Though it's not private, we assume that everything we type on Discord will either be read within a few minutes or not at all. Thus, when we have anything to say that we think someone might care about later than a few minutes from now, we put it in a GitHub issue, comment, or pull request.

Code Style

Cube Cobra currently uses 4 languages: Javascript, Jade/PUG, CSS, and JSX. Before creating your pull request, you can fix any style errors by running npm run-script beautify.

Development Server

The development server can be accessed at: http://cubecobradev.com/ This server usually runs on a more up to date, experimental branch than the production server. The development server is a great way to test new features in an environment closer to the live server. Please report any bugs you see on the development server as an issue. The development server database is not wiped periodically, but it is subject to instability, so don't use it as a reliable way to store any data.

How to report a bug

If you find a security vulnerability, do NOT open an issue. Send a private message to DEKKARU#2784 on discord instead. You can also email support@cubecobra.com. Bugs not related to security can be reported through GitHub issues.

How to suggest a feature or enhancement

Join the discord disscussion.

https://discord.gg/Hn39bCU

Community

We have a fairly active discord and would love for you to join the discussion. https://discord.gg/Hn39bCU

Private Message 'DEKKARU', the admin on Discord, to gain the contributor role and unlock the contributor-only channels.

Community feedback / product ideas

Feedback from the community is collected and synthesized into roadmap ideas in a shared document here. To gain read/write access, PM 'DEKKARU' on Discord.