The Netdata style guide establishes editorial guidelines for any writing produced by the Netdata team or the Netdata community, including documentation, articles, in-product UX copy, and more. Both internal Netdata teams and external contributors to any of Netdata's open-source projects should reference and adhere to this style guide as much as possible.
Netdata's writing should empower and educate. You want to help people understand Netdata's value, encourage them to learn more, and ultimately use Netdata's products to democratize monitoring in their organizations. To achieve these goals, your writing should be:
- Clear. Use simple words and sentences. Use strong, direct, and active language that encourages readers to action.
- Concise. Provide solutions and answers as quickly as possible. Give users the information they need right now, along with opportunities to learn more.
- Universal. Think of yourself as a guide giving a tour of Netdata's products, features, and capabilities to a diverse group of users. Write to reach the widest possible audience.
You can achieve these goals by reading and adhering to the principles outlined below.
One way we write empowering, educational content is by using a consistent voice and an appropriate tone.
Voice is like your personality, which doesn't really change day to day.
Tone is how you express your personality. Your expression changes based on your attitude or mood, or based on who you're around. In writing, your reflect tone in your word choice, punctuation, sentence structure, or even the use of emoji.
The same idea about voice and tone applies to organizations, too. Our voice shouldn't change much between two pieces of content, no matter who wrote each, but the tone might be quite different based on who we think is reading.
For example, a blog post and a press release should have a similar voice, despite most often being written by different people. However, blog posts are relaxed and witty, while press releases are focused and academic. You won't see any emoji in a press release.
Netdata's voice is authentic, passionate, playful, and respectful.
- Authentic writing is honest and fact-driven. Focus on Netdata's strength while accurately communicating what Netdata can and cannot do, and emphasize technical accuracy over hard sells and marketing jargon.
- Passionate writing is strong and direct. Be a champion for the product or feature you're writing about, and let your unique personality and writing style shine.
- Playful writing is friendly, thoughtful, and engaging. Don't take yourself too seriously, as long as it's not at the expense of Netdata or any of its users.
- Respectful writing treats people the way you want to be treated. Prioritize giving solutions and answers as quickly as possible.
Netdata's tone is fun and playful, but clarity and conciseness comes first. We also tend to be informal, and aren't afraid of a playful joke or two.
While we have general standards for voice and tone, we do want every individual's unique writing style to reflect in published content.
Netdata is a global company in every sense, with employees, contributors, and users from around the world. We strive to communicate in a way that is clear and easily understood by everyone.
Here are some guidelines, pointers, and questions to be aware of as you write to ensure your writing is universal. Some of these are expanded into individual sections in the language, grammar, and mechanics section below.
- Would this language make sense to someone who doesn't work here?
- Could someone quickly scan this document and understand the material?
- Create an information hierarchy with key information presented first and clearly called out to improve scannability.
- Avoid directional language like "sidebar on the right of the page" or "header at the top of the page" since presentation elements may adapt for devices.
- Use descriptive links rather than "click here" or "learn more".
- Include alt text for images and image links.
- Ensure any information contained within a graphic element is also available as plain text.
- Avoid idioms that may not be familiar to the user or that may not make sense when translated.
- Avoid local, cultural, or historical references that may be unfamiliar to users.
- Prioritize active, direct language.
- Avoid referring to someone's age unless it is directly relevant; likewise, avoid referring to people with age-related descriptors like "young" or "elderly."
- Avoid disability-related idioms like "lame" or "falling on deaf ears." Don't refer to a person's disability unless it’s directly relevant to what you're writing.
- Don't call groups of people "guys." Don't call women "girls."
- Avoid gendered terms in favor of neutral alternatives, like "server" instead of "waitress" and "businessperson" instead of "businessman."
- When writing about a person, use their communicated pronouns. When in doubt, just ask or use their name. It's OK to use "they" as a singular pronoun.
Some of these guidelines were adapted from MailChimp under the Creative Commons license.
To ensure Netdata's writing is clear, concise, and universal, we have established standards for language, grammar, and certain writing mechanics. However, if you're writing about Netdata for an external publication, such as a guest blog post, follow that publication's style guide or standards, while keeping the preferred spelling of Netdata terms in mind.
Active voice is more concise and easier to understand compared to passive voice. When using active voice, the subject of the sentence is action. In passive voice, the subject is acted upon. A famous example of passive voice is the phrase "mistakes were made."
|Not recommended||When an alarm is triggered by a metric, a notification is sent by Netdata.|
|Recommended||When a metric triggers an alarm, Netdata sends a notification to your preferred endpoint.|
Use the second person ("you") to give instructions or "talk" directly to users.
In these situations, avoid "we," "I," "let's," and "us," particularly in documentation. The "you" pronoun can also be implied, depending on your sentence structure.
One valid exception is when a member of the Netdata team or community wants to write about said team or community.
|Not recommended||To install Netdata, we should try the one-line installer...|
|Recommended||To install Netdata, you should try the one-line installer...|
|Recommended, implied "you"||To install Netdata, try the one-line installer...|
Using words that imply the complexity of a task or feature goes against our policy of universal communication. If you claim that a task is easy and the reader struggles to complete it, you may inadvertently discourage them.
However, if you give users two options and want to relay that one option is genuinely less complex than another, be specific about how and why.
For example, don't write, "Netdata's one-line installer is the easiest way to install Netdata." Instead, you might want to say, "Netdata's one-line installer requires fewer steps than manually installing from source."
A particular word, phrase, or metaphor you're familiar with might not translate well to the other cultures featured among Netdata's global community. We recommended you avoid slang or colloquialisms in your writing.
In addition, don't use abbreviations that have not yet been defined in the content. See our section on abbreviations for additional guidance.
If you must use industry jargon, such as "mean time to resolution," define the term as clearly and concisely as you can.
Netdata helps you reduce your organization's mean time to resolution (MTTR), which is the average time the responsible team requires to repair a system and resolve an ongoing incident.
While the Netdata team is mostly not American, we still aspire to use American spelling whenever possible, as it is the standard for the monitoring industry.
See the word list for spellings of specific words.
Follow the general English standards for capitalization. In summary:
- Capitalize the first word of every new sentence.
- Don't use uppercase for emphasis. (Netdata is the BEST!)
- Capitalize the names of brands, software, products, and companies according to their official guidelines. (Netdata, Docker, Apache, NGINX)
- Avoid camel case (NetData) or all caps (NETDATA).
Whenever you refer to the company Netdata, Inc., or the open-source monitoring agent the company develops, capitalize Netdata.
However, if you are referring to a process, user, or group on a Linux system, use lowercase and fence the word in an
inline code block:
|Not recommended||The netdata agent, which spawns the netdata process, is actively maintained by netdata, inc.|
|Recommended||The Netdata Agent, which spawns the |
Document titles and page headings should use sentence case. That means you should only capitalize the first word.
If you need to use the name of a brand, software, product, and company, capitalize it according to their official guidelines.
Also, don't put a period (
.) or colon (
:) at the end of a title or header.
|Not recommended||Getting Started Guide |
Service Discovery and Auto-Detection:
Install netdata with docker
|Recommended||Getting started guide |
Service discovery and auto-detection
Install Netdata with Docker
Use abbreviations (including acronyms and initialisms) in documentation when one exists, when it's widely accepted within the monitoring/sysadmin community, and when it improves the readability of a document.
When introducing an abbreviation to a document for the first time, give the reader both the spelled-out version and the shortened version at the same time. For example:
Use Netdata to monitor Extended Berkeley Packet Filter (eBPF) metrics in real-time.
After you define an abbreviation, don't switch back and forth. Use only the abbreviation for the rest of the document.
You can also use abbreviations in a document's title to keep the title short and relevant. If you do this, you should still introduce the spelled-out name alongside the abbreviation as soon as possible.
When instructing users to take action, give them the context first. By placing the context in an initial clause at the beginning of the sentence, users can immediately know if they want to read more, follow a link, or skip ahead.
|Not recommended||Read the reference guide if you'd like to learn more about custom dashboards.|
|Recommended||If you'd like to learn more about custom dashboards, read the reference guide.|
The Oxford comma is the comma used after the second-to-last item in a list of three or more items. It appears just before "and" or "or."
|Not recommended||Netdata can monitor RAM, disk I/O, MySQL queries per second and lm-sensors.|
|Recommended||Netdata can monitor RAM, disk I/O, MySQL queries per second, and lm-sensors.|
Do not mention future releases or upcoming features in writing unless they have been previously communicated via a public roadmap.
In particular, documentation must describe, as accurately as possible, the Netdata Agent as of the latest commit in the GitHub repository. For Netdata Cloud, documentation must reflect the _current state of production.
Every link should clearly state its destination. Don't use words like "here" to describe where a link will take your reader.
|Not recommended||To install Netdata, click here.|
|Recommended||To install Netdata, read the installation instructions.|
Use links as often as required to provide necessary context. Blog posts and guides require less hyperlinks than documentation. See the section on linking between documentation for guidance on the Markdown syntax and path structure of inter-documentation links.
Contractions like "you'll" or "they're" are acceptable in most Netdata writing. They're both authentic and playful, and reinforce the idea that you, as a writer, are guiding users through a particular idea, process, or feature.
Contractions are generally not used in press releases or other media engagements.
Emoji can add fun and character to your writing, but should be used sparingly and only if it matches the content's tone and desired audience.
Configuration or maintenance of the Netdata Agent requires some system administration skills, such as navigating directories, editing files, or starting/stopping/restarting services. Certain processes
Netdata documentation often suggests that users switch from their normal user to the
netdata user to run specific
commands. Use the following command to instruct users to make the switch:
NODE instead of an actual or example IP address/hostname when referencing the process of navigating to a dashboard
or API endpoint in a browser.
|Not recommended||Navigate to |
|Recommended||Navigate to |
If you worry that
NODE doesn't provide enough context for the user, particularly in documentation or guides designed
for beginners, you can provide an explanation:
With the Netdata Agent running, visit
http://NODE:19999/api/v1/infoin your browser, replacing
NODEwith the IP address or hostname of your Agent.
When instructing users to run a Netdata-specific command, don't assume the path to said command, because not every Netdata Agent installation will have commands under the same paths. When applicable, help them navigate to the correct path, providing a recommendation or instructions on how to view the running configuration, which includes the correct paths.
For example, the configuration doc first teaches users how to find the Netdata config
directory and navigate to it, then runs commands from the
/etc/netdata path so that the instructions are more
Don't include full paths, beginning from the system's root (
/), as these might not work on certain systems.
|Not recommended||Use |
sudo before a command if you believe most Netdata users will need to elevate privileges to run it. This makes
our writing more universal, and users on
sudo-less systems are generally already aware that they need to run commands
For example, most users need to use
sudo with the
edit-config script, because the Netdata config directory is owned
netdata user. Same goes for restarting the Netdata Agent with
|Not recommended||Run |
Netdata's documentation uses Markdown syntax.
If you're not familiar with Markdown, read the Mastering Markdown guide from GitHub for the basics on creating paragraphs, styled text, lists, tables, and more.
The following sections describe situations in which a specific syntax is required.
Syntax standards (
The Netdata team uses
remark-lint for Markdown code styling.
- Use a maximum of 120 characters per line.
- Begin headings with hashes, such as
# H1 heading,
## H2 heading, and so on.
- Use dashes
-to begin an unordered list, and put a single space after the dash.
- Tables should be padded so that pipes line up vertically with added whitespace.
If you want to see all the settings, open the
remarkrc.js file in the
Every document must begin with frontmatter, followed by an H1 (
Unlike typical Markdown frontmatter, Netdata uses HTML comments (
-->) to begin and end the frontmatter block.
These HTML comments are later converted into typical frontmatter syntax when building Netdata
Frontmatter must contain the following variables:
titlethat quickly and distinctly describes the document's content.
descriptionthat elaborates on the purpose or goal of the document using no less than 100 characters and no more than 155 characters.
custom_edit_urlthat links directly to the GitHub URL where another user could suggest additional changes to the published document.
Some documents, like the Ansible guide and others in the
/docs/guides folder, require an
image variable as well. In
this case, replace
/img/seo, and then rebuild the remainder of the path to the document in question. End
the path with
.png. A member of the Netdata team will assist in creating the image when publishing the content.
For example, here is the frontmatter for the guide about deploying the Netdata Agent with Ansible.
Questions about frontmatter in documentation? Ask on our community forum.
Documentation should link to relevant pages whenever it's relevant and provides valuable context to the reader.
Links should always reference the full path to the document, beginning at the root of the Netdata Agent repository
/), and ending with the
.md file extension. Avoid relative links or traversing up directories using
For example, if you want to link to our node configuration document, link to
/docs/configure/nodes.md. To reference
the guide for deploying the Netdata Agent with Ansible, link to
When referencing a user interface (UI) element in Netdata, reference the label text of the link/button with Markdown's
**bold text**) tag.
Avoid directional language whenever possible. Not every user can use instructions like "look at the top-left corner" to find their way around an interface, and interfaces often change between devices. If you must use directional language, try to supplement the text with an image.
Don't rely on images to convey features, ideas, or instructions. Accompany every image with descriptive alt text.
In Markdown, use the standard image syntax,
!(/docs/agent/contributing), and place the alt text between the brackets
. Here's an example
using our logo:
Reference in-product text, code samples, and terminal output with actual text content, not screen captures or other images. Place the text in an appropriate element, such as a blockquote or code block, so all users can parse the information.
Our documentation site at learn.netdata.cloud uses
Prism for syntax highlighting. Netdata
documentation will use the following for the most part:
go. If no language is specified, Prism tries to guess the language based on its content.
Include the language directly after the three backticks (
```) that start the code block. For highlighting C
code, for example:
And the prettified result:
Prism also supports titles and line highlighting. See the Docusaurus documentation for more information.
The following tables describe the standard spelling, capitalization, and usage of words found in Netdata's writing.
|claimed node||A node that you've proved ownership of by completing the connecting to Cloud process. The claimed node will then appear in your Space and any War Rooms you added it to.|
|Netdata||The company behind the open-source Netdata Agent and the Netdata Cloud web application. Never use netdata or NetData. |
In general, focus on the user's goals, actions, and solutions rather than what the company provides. For example, write Learn more about enabling alarm notifications on your preferred platforms instead of Netdata sends alarm notifications to your preferred platforms.
|Netdata Agent||The free and open source monitoring agent that you can install on all of your distributed systems, whether they're physical, virtual, containerized, ephemeral, and more. The Agent monitors systems running Linux, Docker, Kubernetes, macOS, FreeBSD, and more, and collects metrics from hundreds of popular services and applications.|
|Netdata Cloud||The web application hosted at https://app.netdata.cloud that helps you monitor an entire infrastructure of distributed systems in real time. |
Never use Cloud without the preceding Netdata to avoid ambiguity.
|Netdata community||Contributors to any of Netdata's open-source projects, members of the community forum.|
|Netdata community forum||The Discourse-powered forum for feature requests, Netdata Cloud technical support, and conversations about Netdata's monitoring and troubleshooting products.|
|node||A system on which the Netdata Agent is installed. The system can be physical, virtual, in a Docker container, and more. Depending on your infrastructure, you may have one, dozens, or hundreds of nodes. Some nodes are ephemeral, in that they're created/destroyed automatically by an orchestrator service.|
|Space||The highest level container within Netdata Cloud for a user to organize their team members and nodes within their infrastructure. A Space likely represents an entire organization or a large team. |
Space is always capitalized.
|unreachable node||A connected node with a disrupted Agent-Cloud link. Unreachable could mean the node no longer exists or is experiencing network connectivity issues with Cloud.|
|visited node||A node which has had its Agent dashboard directly visited by a user. A list of these is maintained on a per-user basis.|
|War Room||A smaller grouping of nodes where users can view key metrics in real-time and monitor the health of many nodes with their alarm status. War Rooms can be used to organize nodes in any way that makes sense for your infrastructure, such as by a service, purpose, physical location, and more. |
War Room is always capitalized.
|filesystem||Use instead of file system.|
|preconfigured||The concept that many of Netdata's features come with sane defaults that users don't need to configure to find immediate value.|
|real time/real-time||Use real time as a noun phrase, most often with in: Netdata collects metrics in real time. Use real-time as an adjective: _Netdata collects real-time metrics from hundreds of supported applications and services.|