Feature Branching & Pull Requests (Git Workflow)

adminDevOps0 Comments

Feature Branches & Pull Requests

Problem

When you're working on a project, there will be a bunch of different features or ideas in progress at any given time, not all of which are ready for prime time. Further more, as business priorities change, you might need to prioritize certain features and put others on the back burner. 

At the same time, business requirements mandate that you have a stable version that can be deployed at any given time. We know code can never be entirely bug free. Further more, once deployed there can be unintended consequences. Other times, managers simply change their mind and decide that a certain feature was premature or unnecessary. To mitigate the impact of these events, we need the ability to rollback a feature or cut-bait.

TL;DR: If everyone is working on the same branch such as master, it pollutes the commit history making it all but impossible to figure out which commits relate to specific features making rollbacks impossible. 

Solution

To solve this problem, the standard workflow called branching should be used religiously. Any time a new feature is developed it must be worked on in a separate branch. When you create a branch, you're creating an environment where you can freely test out new ideas without impacting others because changes made on a branch don't affect the master branch (or any other one). 

Further more, no two developers should ever commit to, or work on the same branch at the same time (unless they have permission with the branch stakeholder). Instead, they should create a new branch. 

The next thing that needs to happen is that the master branch is treated as the Holy Grail. Every effort is made to ensure it's stable and can be deployed to production at any time.

Once a feature is considered ready, the developer submits a Pull Request (or PR) and assigns it to a Subject Matter Expert (SME) or peer for review.

On the surface, this is what a well-formatted Pull Request looks like:

A Pull Request allows many things to happen:

  • Title: A “human readable” title that represents the feature

  • Description: A long description that details What was changed, Why it was deemed necessary, Who should review it, and any other References that might be useful (E.g. Jira ticket)
  • Comments: let anyone provide arbitrary feedback viewable by everyone.  
  • Diffs: show what changed between this feature and the current master branch
  • Formal Code Review Process:  let multiple people contribute to the code review process by submitting comments on a line-by-line basis. Having these code reviews formally documented serves as an excellent teaching tool. Over time, the reviews become faster and faster as developers learn what is expected.

  • Merging: Once the PR is approved, the developer can squash and merge their code into the master branch. Squashing allows the master branch to have a very clean commit history where every commit corresponds to a PR. 

  • Clean Commit History: means that every change to the master branch is documented and justified. No one is sneaking in changes.

  • History of Features and when they were added

  • Reverting: If a feature needs to be removed, with the click of a single button it can be removed from the master branch

Technical Details

Create a Branch

Whenever you begin work on a new feature or bugfix, it's important that you create a new branch. Not only is it proper git workflow, but it also keeps your changes organized and separated from the master branch so that you can easily submit and manage multiple pull requests for every task you complete.

To create a new branch and start working on it:

# Checkout the master branch - you want your new branch to come from master
git checkout master

# Pull down the latest changes
git pull origin master

# Create a new branch, with a descriptive name  (e.g. implement-xyz-widget)
git checkout -b newfeature

Now, go to town hacking away. When you're ready, push the changes up to the origin.

git push origin newfeature

Submitting a Pull Request

Prior to submitting your pull request, you might want to do a few things to clean up your branch and make it as simple as possible for the original repo's maintainer to test, accept, and merge your work.

If any commits have been made to the upstream master branch, you should rebase your development branch so that merging it will be a simple fast-forward that won't require any conflict resolution work.

# Fetch upstream master and merge with your repo's master branch
git pull origin master --rebase

Follow the prompts to correct any code conflicts. Any file that is conflicted needs to be manually reviewed. After you fix the problems run:

 git add filename
 
git rebase --continue

Once that is happy, push the rebased changes back to the origin.

git push origin newfeature -f

Then follow these instructions once you're ready:

https://help.github.com/articles/creating-a-pull-request/

Pull Request Template

Use the following markdown template to describe the Pull Request.

## what
* ...high-level explanation of what this PR accomplishes...

## why
* ...business justifications for making the changes...

## references
* ...related pull requests, issues, documents, or research...

Prepared By
Contact 

Erik Osterman
Cloud Posse, LLC

hello@cloudposse.com 
https://cloudposse.com/
310 496 6556

312 Arizona Ave
Santa Monica, CA



How to Easily Pair Program on the Console

adminDevOps0 Comments


Peer Console Development (tmate)

Tmate is a way to share a console window or set of console windows. It’s basically a fork of tmux that makes sessions network enabled. It doesn’t matter where the participants are located - inbound firewall rules don’t interfere since it’s all outbound connections. 

Install tmate

brew tap nviennot/tmate
brew install tmate

Start the SSH Server

tmate -S /tmp/tmate.sock new-session -d

Wait for a connection

tmate -S /tmp/tmate.sock wait tmate-ready 

Get the SSH Command

tmate -S /tmp/tmate.sock display -p '#{tmate_ssh}'

Share the output of the above command with someone you want to share the terminal window with. You should both run the command. The person who runs the SSH Server is the host of the session.

# Example session: (now expired)
ssh q9YsisZyUNtI3rvJcDbG1pAHu@sf1.tmate.io

Interacting with Tmate

^b + c create a new window
^b + n move to next window
^b + p move to previous window

Markdown Formatting “Best Practices”

adminBest Practices0 Comments

Markdown Formatting Best Practices

Using Markdown is essential for clear communication on mediums such as GitHub, Slack or just plan text. Here are some of our recommendations on when to use certain conventions.

Code Blocks

Use code blocks for anything more than 1 line. Use `code` for inline code, filenames, commands, etc.

```
# This is a code block
```

Tables

Use tables to communicate lists of options. 

:--------: should be used for “Default” and “Required” values

:--------- should be used for all other columns

Use  ``  for empty defaults

Use `value`for all values

Here's an example:

|  Name       |  Default |  Description                                  | Required |
|:------------|:--------:|:----------------------------------------------|:--------:|
| namespace   | ``       | Namespace (e.g. `cp` or `cloudposse`)         | Yes      |
| stage       | ``       | Stage (e.g. `prod`, `dev`, `staging`)         | Yes      |
| name        | ``       | Name  (e.g. `bastion` or `db`)                | Yes      |
| attributes  | []       | Additional attributes (e.g. `policy`)         | No       |
| tags        | {}       | Additional tags  (e.g. `map("Foo","XYZ")`)    | No       |

Which will render to something like this:

Features 

Use this format describe the features & benefits.

```
1. **Feature 1** - Explanation of benefits
2. **Feature 2** - Explanation of benefits
```

Quotes

Reference copyrighted text, quotes, and other unoriginal copy using `> `

```
> Amazon Simple Storage Service (Amazon S3) makes it simple and practical to collect, store, and analyze data - regardless of format – all at massive scale.
```

GitHub Repository “Best Practices”

adminBest Practices0 Comments

GitHub Repository Best Practices

Create a New Repository

1. Always initialize repository with a README.md and .gitignore

2. If it's an Open Source project, initialize it with the APACHE2 license

3. Update LICENSE file (if applicable)


Copyright {yyyy} {name of copyright owner}
should be
Copyright 2018 Cloud Posse, LLC

4. Set a meaningful Description that describes the purpose repository

4. Add some useful tags to help discovery of the project

5. Submit all new work as PR.

6. Create new label wip for “Work in Progress”

Repository Settings

  • Disable “Merge Commit” and disable “Rebase Merging”
  • Enable Branch Protections