<!DOCTYPE html> <html class="reftest-wait"> <head> <meta charset="utf-8"> <title>CSS Test: Table Layout and Viewport Resizing</title> <link rel="author" title="Chris Rebert" href="http://chrisrebert.com"> <link rel="help" href="https://drafts.csswg.org/css2/tables.html#auto-table-layout"> <link rel="help" href="https://drafts.csswg.org/mediaqueries-3/#width"> <link rel="help" href="https://drafts.csswg.org/mediaqueries-4/#width"> <link rel="match" href="../reference/ref-filled-green-100px-square.xht"> <meta name="flags" content="dom"> <meta name="assert" content="Resizing a page which toggles the `display` of elements between `block` and `table-cell` based on the viewport width should not cause unnecessary wrapping of the table."> <style> iframe { border: 0; } </style> </head> <body> <p>Test passes if there is a filled green square and <strong>no red</strong>.</p> <iframe id="toy" width="100" height="300" src="support/min-width-tables-001-iframe.html"></iframe> <!-- See min-width-tables-001-iframe.html for the derivation of the 100px value --> <!-- We use 300px height so the incorrect stacking is visible in failure cases --> <!-- This test is not about iframes specifically. It's just that resizing an iframe is more reliable than resizing the window, given browser security restrictions. --> <script> window.addEventListener('load', function () { var PAINT_MS = 250;/* Assume the browser takes about this long to layout/paint this whole page */ var iframe = document.getElementById('toy'); window.setTimeout(function () { iframe.width = 64;/* <100px ; toggle media query off */ window.setTimeout(function () { iframe.width = 100;/* >=100px ; toggle media query on; back to initial width */ // Take the reftest screenshot after the last relayout/repaint finishes window.setTimeout(function () { document.documentElement.className = ''; }, PAINT_MS); }, PAINT_MS); }, PAINT_MS); }, false); </script> </body> </html>