plain-site › type

Type without CSS

The browser ships a complete typographic system: six heading levels, a dozen inline elements that each look different, monospace, quotation marks that follow the language, and furigana. Before reaching for <font>, use the element that means the thing. It renders differently because it is different.


■ Inline elements that look like something

All of these are current, conforming HTML. Each has a default rendering in every browser.


■ Headings

h1 is 2em, h6 is 0.67em. Since 2025 browsers no longer shrink an <h1> nested inside <section> or <article>, so the level you write is the size you get.


■ <font size> and <big>

size takes 1 to 7 (3 is normal) or a relative +n / -n. <big> is the obsolete twin of <small>. Both are obsolete and both still render everywhere.


■ <font face>

face is a font list, like font-family. The generic names always resolve to something on every system. Named fonts only work where they are installed, so always end the list with a generic.


■ lang: the font switch nobody uses

The lang attribute changes rendering. With lang="ja" the browser picks a Japanese font for CJK text (otherwise it may fall back to Chinese forms of the same characters), and <q> uses 「」 brackets instead of “”.


■ Ruby: furigana and glosses

<ruby> puts small text above characters. Built for Japanese reading aids, it works for any annotation. <rp> holds the parentheses that show only in browsers without ruby support.


■ Glyphs are the ornaments

Japanese web pages decorate with typography, not images: section marks, brackets, notes, arrows. They are text, so they are free, they scale, they copy and paste, and they read the same with styles off. Keep a glyph next to a word, never instead of it.


■ <pre>: the oldest layout engine

Preformatted text keeps every space and line break. Box-drawing characters turn it into diagrams, tables, and banners that survive copy-paste into an email or a terminal. <pre wrap> (obsolete) wraps long lines.


■ <xmp>: code samples with no escaping

<xmp> is obsolete, and it is the most useful obsolete element for a documentation site. Its content is raw text: tags inside it are shown, not parsed, so you paste HTML in and never write &lt;. The only thing it cannot contain is its own closing tag. <listing> renders like <pre> but does parse tags.


■ Keeping words together

<nobr> (obsolete) never breaks its content. <wbr> (conforming) marks where a long word may break. &nbsp; glues two words.



▲ top ・ plain-site ・ previous: colour ・ next: layout