Technical Writing
What is a "doc lint" tool?
It lists exact software versions, accounts, permissions, or knowledge the reader must have before the rest of the doc will work, preventing wasted effort.
Plain language uses familiar words, direct sentences, and clear structure without sacrificing accuracy.
ASCII art is text-drawn diagrams. It is acceptable in code comments, READMEs, and terminal contexts where rendered images are not an option, and when simplicity is the goal.
A doc lint tool (markdownlint, Vale, alex) automatically checks prose for style issues, banned words, broken links, and accessibility problems in CI.
Technical Writing
What is a strong technical-writing principle?
A "WHY" comment explains the reasoning behind a non-obvious line ("// Use 2 retries, not 3: 3 fails the SLO under packet loss"). It is the most valuable kind of code comment.
For most features, ship a short overview, a quickstart, and a reference for every public option. Other doc types come later, once support feedback shows gaps.
A doc lint tool (markdownlint, Vale, alex) automatically checks prose for style issues, banned words, broken links, and accessibility problems in CI.
Make the next right action obvious and the surrounding complexity easier to navigate.
Technical Writing
Why should jargon be controlled carefully?
It is the shortest example that still demonstrates the documented behavior, with no unrelated features, no decorative config, and no copy-paste noise.
Expert terms can be useful, but unnecessary jargon increases cognitive load and excludes readers.
A code block presents runnable or copy-pasteable commands, configuration, or output with preserved whitespace, syntax, and no prose wrapping.
Shared docs make decisions, process, and system behavior easier to reference consistently across the team.
Technical Writing
What is a "runbook" in technical writing?
A runbook is an operational document giving on-call responders exact steps to diagnose, mitigate, and resolve a specific production issue.
A style guide records team-specific decisions (capitalization, date format, voice, terminology) so writers do not re-debate them per page and the site feels uniform.
Voice is the personality the document projects: direct, helpful, neutral, sometimes warm. A consistent voice builds trust across pages.
It is a widely adopted public style guide covering word choice, formatting, voice, and inclusive language for technical content. Many teams adopt it as a baseline.
Technical Writing
What is the "F-shape" reading pattern?
A design doc explains a proposed technical change: the problem, options considered, the chosen approach, trade-offs, risks, and rollout plan.
A known issue section lists active bugs or limitations the user should be aware of, with workarounds, the affected versions, and a link to tracking.
It describes how web readers scan: horizontally across the top, then a shorter horizontal sweep further down. Docs use headings, lists, and bold to feed this pattern.
A technical reviewer is an engineer or subject-matter expert who verifies that the doc's claims about the system are correct, and that examples actually run.
Technical Writing
What is a "README" supposed to do?
Start with a Tutorial or Quickstart, then Concepts, then How-to guides, then Reference. This matches the journey of a learning reader.
Inclusive language avoids alienating or stereotyping readers, widens the audience, and reflects the team's values; specific words ("allowlist", "primary/replica") are usually easy swaps.
A README answers five questions fast: what this is, who it is for, how to install, how to run, and where to learn more. Anything else belongs in linked docs.
Broken examples erode trust quickly and make readers doubt the rest of the document.
Technical Writing
Why should short paragraphs be favored in docs?
Short paragraphs are easier to scan, especially when readers are under time pressure.
It is a widely adopted public style guide covering word choice, formatting, voice, and inclusive language for technical content. Many teams adopt it as a baseline.
A deprecation notice tells users a feature will be removed, states the removal version, explains the migration path, and links to the replacement.
It describes how web readers scan: horizontally across the top, then a shorter horizontal sweep further down. Docs use headings, lists, and bold to feed this pattern.
Technical Writing
Why should imperative mood be used in procedural steps?
Procedural writing gives readers step-by-step instructions to complete a task successfully.
Clear technical writing is specific, structured, concise, and easy to scan.
Imperative mood ("Click Save", "Run the script") puts the action on the reader without an ambiguous subject, making each step short, parallel, and easy to translate.
On mobile, long paragraphs, wide tables, and side-by-side code blocks break. Write for narrow screens first; wide layouts are a progressive enhancement.
Technical Writing
What is the "WHY" comment pattern?
Lead with the conclusion or main action of the paragraph, then add the conditions, edge cases, and background, so scanners get the point in the first sentence.
A "WHY" comment explains the reasoning behind a non-obvious line ("// Use 2 retries, not 3: 3 fails the SLO under packet loss"). It is the most valuable kind of code comment.
It is the tendency, once an author understands a topic, to forget which parts are hard, and to skip the steps and definitions a beginner actually needs.
Cognitive load is the mental effort a reader must spend to understand a passage. Docs minimize it with structure, examples, and consistent terminology.
Technical Writing
What is the "single source of truth" principle for docs?
Information should live in exactly one place. Other pages link to it, never copy-paste it, so updates propagate automatically and never contradict each other.
Inclusive language avoids alienating or stereotyping readers, widens the audience, and reflects the team's values; specific words ("allowlist", "primary/replica") are usually easy swaps.
A feature flag is a runtime switch that enables a feature for some users. Docs should describe the flag, its default, and which audiences currently see the feature.
A known issue section lists active bugs or limitations the user should be aware of, with workarounds, the affected versions, and a link to tracking.
Technical Writing
Why is the order of fields in an API reference important?
A docs review is a structured pass that checks accuracy (does it match the product?), completeness (are edge cases covered?), and clarity (would a new reader get it?). Engineers, writers, and ideally a fresh reader all contribute.
Ordering the most important and most-frequently-used fields first helps readers find the answer faster, even if the underlying schema is alphabetical.
An error code reference lists every error the system can return, with a unique identifier, a plain-language meaning, common causes, and remediation steps.
Docs as code stores documentation in the same repository as the product, in plain text (Markdown, AsciiDoc), reviewed via pull requests, and built by a static site generator.
Technical Writing
What is a tutorial?
A tutorial guides the reader through learning by doing, usually in a beginner-friendly sequence.
A design doc explains a proposed technical change: the problem, options considered, the chosen approach, trade-offs, risks, and rollout plan.
Conceptual writing explains how something works or why it matters before asking the reader to act.
An Architecture Decision Record captures one significant decision: the context, the choice made, the consequences, and the alternatives rejected, signed by date and author.
Technical Writing
What is a "decision record" (ADR)?
Task success is the rate at which a reader with a goal completes it using the doc alone. The single most important metric for procedural content.
Information should live in exactly one place. Other pages link to it, never copy-paste it, so updates propagate automatically and never contradict each other.
Versioned with the product, reviewed by engineers, branched per release, searchable as text, and CI-checkable for broken links and lint rules.
An Architecture Decision Record captures one significant decision: the context, the choice made, the consequences, and the alternatives rejected, signed by date and author.