Skip to content

Latest commit

 

History

History
72 lines (41 loc) · 5.03 KB

File metadata and controls

72 lines (41 loc) · 5.03 KB

Angular Components Library

A great set of reusable @angular components, directives and testing utilities.

One of our main goals is to allow easy integration with the great collection of components already provided by the @angular/material team and to build everything with the following in mind:

  • 🤘 a11y
  • 🚀 performance
  • 💖 look and feel

npm Build Status Test Coverage Maintainability License

Browser Support

IE / Edge
IE / Edge
Firefox
Firefox
Chrome
Chrome
Safari
Safari
IE11, Edge last 2 versions last 2 versions last 2 versions

Installation

npm i @uipath/angular

Documentation

Visit our docs page here. Thanks go to the @compodoc team! 👌

Git Conventions

Commit Messages

We initially started off by using a custom commit message convention, now that we've gone OS, we realize it will be much easier for everyone if we adhere to the conventional-commit standard.

The project comes preconfigured with commitlint and husky hooks in order to automate checks.

Another advantage of conventional-commits is that it will allow us to later on integrate semantic-release with ease.

Conflict Resolution

We all know that once you open a PR, it sometimes goes through a rigorous code review process. Sometimes, due to this delay, we end up with a PR that has conflicts.

In this scenarios, we're usually used to either merge or rebase.

We personally prefer that branches get rebased, and usually stay away from merge commits as they mostly end up polluting the history needlessly. At this point, most of you are probabily thinking: Yeah, but it's not a good practice to rebase if multiple people are working on the same branch!.

True, it's not recommended to rebase when multiple people are active on the same branch, but in this scenario git offers a very elegant solution: fixup commits. By using fixup commits, the branch can later go through an interactive rebase once the review process is done and the PR is approved.

For example: after you make your changes and stage them you can run a git commit --fixup HEAD~ to add the changes in the last but one commit. Before the merge, you can then simply git rebase -i <ref> --autosquash and every change will be squashed in the correct commit.

Why?

Q: Why all the hassle, does it really matter?

A: We've seen too many branches where merge commits get introduced without a good reason, too many branches that have almost the same number merge commits as commits so we want to enforce a git usage guidline that promotes responsibility. 🐱‍👤

Q: But I want to integrate a branch, not yet merged and implement a new feature.

A: In this scenario, you have the master branch, and the feature branch that you depend upon. What we do in this case, is create a local branch where we merge the required feature branch into master and start our branch from there. Once the feature branch we depend upon, gets merged into master, we rebase over master. Easy peasy! 🤓

Q: But what if I just want to add a feature?

A: We think that in software, it's not just about the feature, it's also about the fact that others will later on need to support that feature or modify it, having a good historical track helps everyone better understand the evolution of the codebase.

Coming Up (most likely v1)

  • Getting Started / Contribution Guidlines
  • Enhanced Component Demo's