How these pages treat links
- A link goes somewhere; a button does something. Every link is a real
<a href>, and nothing a link does ever changes data. - Following a link is a request. The pages show what it sends, using
/api/echo, which replies with exactly the request it received. - A link is a promise about its destination. Its text says where it goes, its address is honest, and it keeps working.
- The browser already does most of the work. New tabs, history, the back button and keyboard access come free with real links; a script that takes a link over has to keep all of them working.
Demos use the /go/ addresses on cse134.site, which answer with every kind of redirect, a missing page and a download, so you can watch each one in DevTools.
The pages
- 01The link
Hypertext and the
<a>element, links around blocks, link text as the link's name, and link versus button. - 02Where a link goes
How an
hrefresolves, absolute and relative links,<base>, fragments and text fragments, and schemes such asmailto:andtel:. - 03What a click does
The request a link sends, the Referer header and referrer policies, redirects and other answers, new tabs and the opener, downloads and pings, history, and faster clicks.
- 04Links between documents
<link rel>andrelon links: stylesheets and icons, canonical and alternate pages, preloading, and the values search engines read. - 05Links people can use
Link text that makes sense alone, links that look like links, focus and skip links, new windows and files, image links and image maps.
- 06Links you can trust
Reading a URL before following it, short links, tracking parameters, link rot, embedding with
<iframe>, and where JavaScript takes links next.