We provide these guidelines to help our contributors make additions or corrections to our documentation.
Welcome. We're glad you're here to help make our technical documentation even better. We develop content that leads our users to success using Grafana products. Technical accuracy is our primary consideration, and we value the use of inclusive language. We regard your feedback as a gift - thanks for reading through these guidelines.
We write these guidelines for contributors who are interested in improving our technical content.
All Grafana Enterprise and OSS documentation is located in the Grafana open source project GitHub repository: https://github.com/grafana/grafana/tree/main/docs/sources.
- The sources directory organizes content by topic areas, for example administration and alerting.
- Topic directories include an
_index.md
file, which provides an overview of the topic, and optionally includes subtopics that provide more detail.
The
_index.md
file is required.
We write technical documentation using Markdown. We've put together a short guide to help you how to structure and format your content.
To access the markdown guide, refer to Markdown style guide.
We're thrilled that you are considering contributing to the documentation. You can contribute content in the following ways:
Request a change when you want to make a suggestion about a topic, but don't want to provide an edit that generates a pull request. Requesting a change gives you the freedom to express your ideas without committing language. Your suggestion can reflect a small change to wording or can reflect larger, more substantive changes.
GitHub captures your request as an Issue logged against the repository.
Before you begin:
- Create a GitHub account.
To request a change, complete the following steps:
- While viewing the topic, click Request a change.
The Issue title auto-populates with the location of the file about which you are requesting a change.
-
Enter a change request description.
-
Add the type/docs label.
-
Click Submit new issue.
If you want to recommend a small change, such as suggesting a correction to a topic, you can edit the topic directly in GitHub. You are not required to fork and clone the repo to use this approach.
Other small changes might include:
- Adding steps to a task
- Adding clarifying language to a concept
- Providing an example
Before you begin:
- Create a GitHub account.
To edit a topic, complete the following steps:
- While viewing the topic you want to edit, click Edit this page.
-
Add your changes to the topic.
-
Scroll to the bottom of the page and enter a branch name.
For example, enter
clarified dashboard panel definition
. -
Click Commit.
GitHub prompts you to create a PR.
-
Complete the prompts provided in the body of the PR.
-
Click Create pull request.
At Grafana Labs, we use the principles of topic-based authoring when we write technical documentation. Topic-based authoring provides guidelines for writing three types of technical documentation: concept, task, and reference. Before you begin writing, establish the topic type you want to write.
Technical content is divided into three topic types: concept, task, and reference.
- Concept: A concept topic explains what a feature (or idea) is, and why it is important.
- Task: A task topic explains how to complete an end user procedure in the system. Task topics contain steps.
- Reference A reference topic contains lookup information that a user might consult when they complete a task. Documenting a list of values with descriptions is a common form of reference topic.
Example
Suppose you are writing content for a site called Doggie handbook. You might organize your topics like this:
Concepts
- What a dog is
- Brief history of dogs
- Why you might want a dog
- Tasks dogs can be trained to do
Tasks
- Feed the dog
- Groom the dog
- Train the dog
References
- List of dog equipment you will need
- Table of breeds that includes breed name, size range, short or long hair, and type of dog
Before you begin writing, we recommend that you fork and clone the Grafana repository so that you can use a text editor locally to create branches, commit your changes, and create a PR.
While this document doesn't include git commands or descriptions of Github operations, you might find these links useful.
- Install git: We store all source code, including documentation, in Git repositories.
- Fork a repo: Locate the repo you want to clone, and fork it.
- Clone a repo: Clone the repository to your local machine.
- Create a branch: Before you make change, create a branch. Do not push changes against the
main
branch. - Create a PR: After you add, commit, and push your changes, create a PR in Github.
We have provided documentation templates that align with each topic type:
Each template provides additional usage and formatting guidelines. We recommend that you make a copy of whichever template you are using, then add content.
Remove any unused content before you commit your changes.
Prior to pushing your changes, you can view a local build of the documentation so that you can review your work.
To view a local build:
-
Install Docker.
-
Run Docker.
-
Navigate to the docs root directory.
-
Run
make build
. -
Open
localhost:3002
to review your changes.
When you are ready for other people to review your work, perform the following tasks.
-
Add your changes, which prepares your content for the next commit.
-
Commit your changes.
-
Push your changes to Github.
-
Create a PR in Github.
The docs build system automatically conducts a series of tests to ensure that the content doesn't conflict with other content in the docs repository.
When a PR is added to the repo with a label associated with docs
, it will be reviewed by a member of the technical writing team. Depending on the size of the PR and the priority of other work, the PR will either be immediately reviewed and merged (minor fixes typically follow this pattern) or the PR will be triaged and placed in the backlog of work or moved into further development.
For general discussions on documentation, you’re welcome to join the #docs
channel on our public Grafana Slack team.
While we don't provide a comprehensive style guide in this document, here's a few writing tips to help the writing process.
Write for an audience that is computer literate and has general technical knowledge, but is not necessarily familiar with Grafana or the finer points of observability.
Pretend you are explaining your topic to a brand new Grafana user or developer.
Use the following sentence structure when you write: subject—verb—object. If you are telling a user to do something, write an imperative sentence. For example “Enter the refresh rate time interval and click Save.”
- You can also start a sentence with an if clause, which positions the condition before the action.
- Limit the number of words in a sentence to 20..
Active voice makes the performer of the action the subject of the sentence. When you write in passive voice, the recipient of the action (and not the performer) becomes the subject of the sentence.
Active-voice sentences are more direct and clearly identify who is doing what. Not all tasks are completed by a user; sometimes the system can also be a performer (and by extension, the subject of the sentence). When you write in active voice you clearly make that distinction, which results in more engaging and less wordy content.
Use (active) | Avoid (passive) |
---|---|
After you upgrade the software, restart the computer. | After the software has been upgraded, the computer can be restarted. |
Click OK to save the dashboard. | The dashboard is saved when the OK button is clicked. |
Create a dashboard. | A dashboard is created by you. |
Users might be unfamiliar with some non-English words and abbreviations such as per, vs, and via. Latin abbreviations in particular, like i.e., e.g., and etc., are vague.
Use | Don’t use |
---|---|
through | via |
that is | i.e. |
according to | per, as per |
and so on | etc. |
Thanks to search engines, every page in the documentation might be a reader's entry point. This means that each page needs to be self-contained and make sense on its own. The reader should not need to read other topics in order to perform the task or understand the concept.
However, try to be helpful and link to related information. Using the Doggie handbook example, the concept topic that explains what dogs can be trained to do might link to the Train the dog task.