<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	xmlns:wfw="http://wellformedweb.org/CommentAPI/"
	xmlns:dc="http://purl.org/dc/elements/1.1/"
	xmlns:atom="http://www.w3.org/2005/Atom"
	xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
	xmlns:slash="http://purl.org/rss/1.0/modules/slash/"
	>

<channel>
	<title>Learning the World &#187; http-request</title>
	<atom:link href="http://learningtheworld.eu/tag/http-request/feed/" rel="self" type="application/rss+xml" />
	<link>http://learningtheworld.eu</link>
	<description></description>
	<lastBuildDate>Tue, 06 Nov 2012 00:17:33 +0000</lastBuildDate>
	<language>en-US</language>
		<sy:updatePeriod>hourly</sy:updatePeriod>
		<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.9.1</generator>
	<item>
		<title>Better Foreground Sprites</title>
		<link>http://learningtheworld.eu/2008/better-foreground-sprites/</link>
		<comments>http://learningtheworld.eu/2008/better-foreground-sprites/#comments</comments>
		<pubDate>Thu, 26 Jun 2008 18:00:17 +0000</pubDate>
		<dc:creator><![CDATA[Martin Kliehm]]></dc:creator>
				<category><![CDATA[accessibility]]></category>
		<category><![CDATA[conferences]]></category>
		<category><![CDATA[downloads]]></category>
		<category><![CDATA[web development]]></category>
		<category><![CDATA[background-image]]></category>
		<category><![CDATA[CSS]]></category>
		<category><![CDATA[css sprites]]></category>
		<category><![CDATA[foreground sprites]]></category>
		<category><![CDATA[Google]]></category>
		<category><![CDATA[http-request]]></category>
		<category><![CDATA[performance]]></category>
		<category><![CDATA[T.V. Raman]]></category>

		<guid isPermaLink="false">http://learningtheworld.eu/?p=106</guid>
		<description><![CDATA[A while ago I wrote about using <acronym title="Cascading Style Sheets">CSS</acronym> Sprites in <code>img</code> tags&#160;&#8212; <strong>Foreground Sprites</strong>. Thus you avoid HTTP requests, but the page turns really ugly when <acronym>CSS</acronym> is switched off because the sprite image will be displayed in its full size. Now Google&#8217;s accessibility specialist T.V. Raman explained their idea of using sprites for foreground images&#160;&#8212; a much better solution!]]></description>
				<content:encoded><![CDATA[<p>A while ago I wrote about using <a href="http://www.alistapart.com/articles/sprites/"><acronym title="Cascading Style Sheets">CSS</acronym> Sprites</a> in <code>img</code> tags, calling it <strong><a href="/2007/foreground-sprites/">Foreground Sprites</a></strong>. Thus you avoid performance-eating HTTP requests, but the page turns really ugly when <acronym>CSS</acronym> is switched off because the sprite image will be displayed in its full size. <img src="http://learningtheworld.eu/wp-includes/images/smilies/icon_sad.gif" alt=":(" class="wp-smiley" /> </p>

<p>What are the alternatives? Using background-images with off-screen text? Bad code:</p>

<ol class="code bad" title="HTML">
<li><code>&lt;a href=&quot;foo&quot; class=&quot;button&quot;&gt;</code></li>
<li class="indent"><code>&lt;span&gt;Alternative Text&lt;/span&gt;</code></li>
<li><code>&lt;/a&gt;</code></li>
</ol>

<ol class="code bad" title="CSS">
<li><code>a.button, button {</code></li>
<li class="indent"><code>background: url(foo.gif) 0 &minus;26px no-repeat;</code></li>
<li class="indent">display: block;</li>
<li class="indent">height: 52px;</li>
<li class="indent">overflow: hidden;</li>
<li class="indent">position: relative;</li>
<li class="indent">width: 150px;</li>
<li><code>}</code></li>
<li><code>span {</code></li>
<li class="indent"><code>left: &minus;9999px;</code></li>
<li class="indent"><code>position: absolute;</code></li>
<li class="indent"><code>top: auto;</code></li>
<li><code>}</code></li>
</ol>

<p><strong>That&rsquo;s a bad idea with accessibility issues.</strong> Screen readers will be able to read the text, but if somebody with low vision has high-contrast custom stylesheets, it is likely that the <code>background-image</code> won&rsquo;t be displayed. Still the alternative text will be invisible off-screen, so the link or <code>button</code> becomes unusable.</p>

<h3>A better solution</h3>

<p>Now Google&rsquo;s accessibility specialist <strong>T.V. Raman</strong> explained their idea for <a href="http://googlewebmastercentral.blogspot.com/2008/05/design-patterns-for-accessible.html">replacing  images with sprites</a>. Here are screenshots of Google search results with images, and with background images disabled:</p>

<p><img src="/wp-content/uploads/2008/06/screenshot-google-search-result-with-bgimage.png" width="400" height="84" alt="Screenshot displaying the Google logo" class="screenshot" />
<img src="/wp-content/uploads/2008/06/screenshot-google-search-result-wo-bgimage.png" width="400" height="84" alt="Screenshot with text visible instead of the logo" class="screenshot" /></p>

<p>Google puts the alternative text <strong><em>behind</em> the background image</strong> simply by attaching it on the <code>span</code> element, not the anchor. So when <acronym>CSS</acronym> or background images are turned off, the text just reappears.</p>

<ol class="code" title="XHTML">
<li><code>&lt;a id=&quot;logo&quot; title=&quot;Go to Google Home&quot; href=&quot;http://www.google.com/&quot;&gt;</code></li>
<li class="indent">Google</li>
<li class="indent"><code><strong>&lt;span&gt;<ins datetime="20080627T174700">&lt;/span&gt;</ins></strong></code></li>
<li><code>&lt;/a&gt;</code></li>
</ol>

<ol class="code" title="CSS">
<li><code>#logo {</code></li>
<li class="indent">display: block;</li>
<li class="indent">height: 52px;</li>
<li class="indent">overflow: hidden;</li>
<li class="indent">position: relative;</li>
<li class="indent">width: 150px;</li>
<li><code>}</code></li>
<li><code>#logo span {</code></li>
<li class="indent"><code>background: url(foo.gif) 0 &minus;26px no-repeat;</code></li>
<li class="indent"><code>height: 100%;</code></li>
<li class="indent"><code>left: 0;</code></li>
<li class="indent">position: absolute;</li>
<li class="indent">top: 0;</li>
<li class="indent"><code>width: 100%;</code></li>
<li><code>}</code></li>
</ol>

<p>There&rsquo;s only an issue if you have <em>a lot</em> of alternative text, perhaps combined with text zoom, so that it doesn&rsquo;t fit in the reserved space. But I can live with that.</p>

<p>Congratulations to Google for this ellegant solution, and it&rsquo;s <em>so</em> simple! Why didn&rsquo;t <em>I</em> think of it? <img src="http://learningtheworld.eu/wp-includes/images/smilies/icon_wink.gif" alt=";)" class="wp-smiley" /> </p>
]]></content:encoded>
			<wfw:commentRss>http://learningtheworld.eu/2008/better-foreground-sprites/feed/</wfw:commentRss>
		<slash:comments>3</slash:comments>
		</item>
		<item>
		<title>Website Performance Tweaks, Part Two</title>
		<link>http://learningtheworld.eu/2007/performance-2/</link>
		<comments>http://learningtheworld.eu/2007/performance-2/#comments</comments>
		<pubDate>Sun, 24 Jun 2007 14:20:19 +0000</pubDate>
		<dc:creator><![CDATA[Martin Kliehm]]></dc:creator>
				<category><![CDATA[conferences]]></category>
		<category><![CDATA[web development]]></category>
		<category><![CDATA[@media]]></category>
		<category><![CDATA[atmedia]]></category>
		<category><![CDATA[atmedia07]]></category>
		<category><![CDATA[atmedia2007]]></category>
		<category><![CDATA[book:ean=9780596529307]]></category>
		<category><![CDATA[book:isbn=0596529309]]></category>
		<category><![CDATA[cdn]]></category>
		<category><![CDATA[cpu]]></category>
		<category><![CDATA[CSS]]></category>
		<category><![CDATA[energy efficiency]]></category>
		<category><![CDATA[etag]]></category>
		<category><![CDATA[expires header]]></category>
		<category><![CDATA[gzip]]></category>
		<category><![CDATA[http-request]]></category>
		<category><![CDATA[JavaScript]]></category>
		<category><![CDATA[Nate Koechley]]></category>
		<category><![CDATA[performance]]></category>
		<category><![CDATA[steve souders]]></category>
		<category><![CDATA[tenni theurer]]></category>
		<category><![CDATA[yahoo]]></category>
		<category><![CDATA[yslow]]></category>
		<category><![CDATA[yui]]></category>

		<guid isPermaLink="false">http://learningtheworld.eu/2007/performance-2/</guid>
		<description><![CDATA[Nate Koechley presented the research results of the Yahoo! Exceptional Performance Team two weeks ago in London. The traditional focus of <strong>performance optimization</strong> has been on the backend, i.e. system efficiency. But comparing a number of high profile websites, the Yahoo! team found that frontend performance is responsible for 80-98% of the perceived response time. Therefore doubling the frontend performance gains more than doubling the backend performance. [&#8230;]]]></description>
				<content:encoded><![CDATA[<p class="vcard"><a href="http://www.flickr.com/photos/drewm/538822354/in/set-72157600330136671/" title="See larger version on flickr"><img alt="Nate Koechley" src="/wp-content/uploads/2007/07/nate-koechley" width="240" height="160" class="floatleft photo" /></a> <strong><a href="http://nate.koechley.com" class="fn url" rel="co-worker met acquaintance">Nate Koechley</a></strong> presented the <a href="http://nate.koechley.com/blog/2007/06/12/high-performance-web-sites">research results</a> of the Yahoo! Exceptional Performance Team two weeks ago in London (<a href="http://www.htmldog.com/atmedia2007/highperformancewebpages.mp3" type="audio/mp3">podcast</a>). Like Yahoo! shares I would like to share that knowledge with you for those who couldn&rsquo;t attend.</p>

<p>The traditional focus of <strong>performance optimization</strong> has been on the backend, i.e. system efficiency. But comparing a number of high profile websites, the Yahoo! team found that frontend performance is responsible for 80-98% of the perceived response time. Therefore doubling the frontend performance gains more than doubling the backend performance. In case studies <em>Yahoo! Search</em> became 40-50% faster, the <em>Yahoo! Mail</em> web application gained 70-100%. Of course there are ways to increase backend performance without throwing in more hardware, but better frontend performance reduces traffic and saves resources.</p>

<p>Saving resources on the <em>client</em> side, particularly <strong>CPU usage</strong>, also pays off in speed. <a href="http://icant.co.uk/sandbox/eventdelegation/">Event delegation</a> is faster than a large number of event handlers. Likewise we know that <a href="/2007/performance/">reducing the number of HTTP requests</a> through techniques like CSS sprites, sliding doors, or file aggregation increases speed. The reason is the limit of two parallel requests <em>per host</em> imposed by HTTP 1.1. That results in a download queue of two requests at a time, increasing the perceived response time of a page. By configuring additional host aliases for your server you can <a href="http://yuiblog.com/blog/2007/04/11/performance-research-part-4/">increase the number of parallel requests</a>&nbsp;&mdash; but more than 2-4 also increase DNS lookups resulting in higher CPU usage and slower response times.</p>

<p>I wonder when Yahoo! will present us another impressive calculation <strong>how many gigawatts have been preserved</strong> by reducing CPU usage in client PCs and in their <a href="http://www.technewsworld.com/story/55792.html" title="Study: Data Center Power Usage Exploding">data</a> <a href="http://brand.yahoo.com/forgood/environment/energy_conservation.html" title="Yahoo! Energy Conservation Program">centers</a>, as one participant asked in the <acronym title="questions and answers">Q&amp;A</acronym> part. <a href="http://www.ecologee.net">Energy efficient servers</a> are the next big thing, but are there any concrete suggestions for <a href="http://www.addsimplicity.com/adding_simplicity_an_engi/2007/01/compute_power_i.html">greener programming</a>? Is <acronym title="Asynchronous JavaScript and XML">AJAX</acronym> destroying the ozone layer? <img src="http://learningtheworld.eu/wp-includes/images/smilies/icon_wink.gif" alt=";)" class="wp-smiley" /> </p>

<p>Environmental issues aside, here&rsquo;s the <strong><a href="http://developer.yahoo.com/performance/rules.html">list of rules</a></strong>. I&rsquo;ll keep it short where I have written about it in <a href="/2007/performance/">my previous article</a>. See the <a href="http://stevesouders.com/hpws/">examples and testcases</a> by Steve Souders.</p>

<ol>
<li id="rule-1"><strong>Make <a href="/2007/performance/">fewer HTTP requests</a>:</strong> This also affects <a href="http://yuiblog.com/blog/2007/03/01/performance-research-part-3/">cookies</a>. Eliminate unnecessary cookies, keep them small, set them at granular domain levels (e.g. <code>finance.yahoo.com</code> instead of <code>.yahoo.com</code>), and set an appropriate Expires date.</li>
<li id="rule-2"><strong>Use a content distribution network (<acronym>CDN</acronym>)</strong> like <a href="http://www.akamai.com">Akamai</a> where your (static) content is served from distributed data centers located nearer to your client. Even if your website is not as big as Google you can profit from faster response times by using the <a href="http://yuiblog.com/blog/2007/02/22/free-yui-hosting"><acronym title="Yahoo! User Interface">YUI</acronym> library&rsquo;s own <acronym>CDN</acronym></a>.</li>
<li id="rule-3"><strong>Add an Expires header</strong> not just <a href="/2007/performance/#enforce-caching" title="Enforce image caching">for images</a>, but also for JavaScript and stylesheet files.</li>
<li id="rule-4"><strong>Enable gzip:</strong> 90%+ of browsers support compression, and <code>gzip</code> is better supported and compresses more than <code>deflate</code>. Gzip <acronym title="Hypertext Markup Language">HTML</acronym> files, <acronym title="Cascading Stylesheets">CSS</acronym>, scripts, <acronym title="Extensible Markup Language">XML</acronym>, <acronym title="JavaScript Object Literal Notation">JSON</acronym>&nbsp;&mdash; <em>no</em> images or <acronym title="Portable Data Format">PDF</acronym>s.</li>
<li id="rule-5"><strong>Put <acronym>CSS</acronym> at the top</strong>, avoid <code>@import</code> as it loads <em>last</em>, even <em>after</em> the images!</li>
<li id="rule-6"><strong>Move scripts to the bottom</strong> as they block parallel downloads even across hostnames and block rendering of any code below them.</li>
<li id="rule-7"><strong>Avoid <acronym>CSS</acronym> expressions</strong> as they execute many times and cost CPU.</li>
<li id="rule-8">
<p><strong>Use external JavaScript and <acronym>CSS</acronym> files.</strong> <a href="/2007/performance/#inline-css">Inline <acronym>CSS</acronym></a> is apparently faster for a user&rsquo;s start page, but not on subsequent pages. After the page has finished loading, use the time to <strong>preload scripts</strong> to speed up secondary pages.</p>
<ol class="code">
<li><code>window.onload = downloadComponents;</code></li>
<li><code>function downloadComponents() {</code></li>
<li class="indent"><code>var elem = document.createElement(&quot;script&quot;);</code></li>
<li class="indent"><code>elem.src = &quot;http://.../file1.js&quot;;</code></li>
<li class="indent"><code>document.body.appendChild(elem);</code></li>
<li><code>}</code></li>
</ol></li>
<li id="rule-9"><strong>Reduce <acronym title="Domain Name Server">DNS</acronym> lookups</strong> for the reasons stated above. Use 1-4 hosts and the <code>keep alive</code> setting.</li>
<li id="rule-10"><strong><a href="/2007/performance/#file-aggregation">Minify JavaScript</a></strong> with JSMin&nbsp;&mdash; inline scripts, too.</li>
<li id="rule-11"><strong>Avoid redirects</strong> as they are the worst form of blocking. Set Expires headers for redirects to enable caching.</li>
<li id="rule-12"><strong>Remove duplicate files:</strong> this is self-explanatory, but it can happen in large teams with many scripts and stylesheets.</li>
<li id="rule-13"><p><strong>Mind the <acronym title="Entity Tag">ETag</acronym>:</strong> Now this was something I never paid attention to. ETags are unique identifiers to distinguish files that share a <acronym title="Uniform Resource Identifier">URI</acronym>. They are transmitted in the HTTP header. The default server setting uses the <a href="http://en.wikipedia.org/wiki/Inode">INode</a>, the bytesize and the modification date of a file to calculate a unique ID. Unless servers in a cluster are identical, ETags differ, therefore the files are <strong>not cached</strong>. Fortunately <a href="http://httpd.apache.org/docs/2.0/mod/core.html#fileetag">ETags can be configured</a> in Apache, so it should be possible to match them across different servers.</p><ol class="code"><li><code>FileETag MTime Size</code></li></ol>
<p>Note that the ETag is also <strong>relevant for <acronym title="Really Simple Syndication">RSS</acronym> feeds</strong>. For example, currently the <a href="http://www.w3.org/2004/08/TalkFiles/Talks.rss" type="application/rss+xml"><acronym title="World Wide Web Consortium">W3C</acronym> talks feed</a> is more or less unusable: some feed readers and services apparently regard the ETag, the feed is mirrored on many servers, so the same news entry from a different server is shown as new and unread multiple times every day&hellip;</p>
</li>
<li id="rule-14"><strong>Make <acronym title="Asynchronous JavaScript and XML">AJAX</acronym> cacheable and small</strong>. Some data like a user&rsquo;s address book or buddy list change infrequently and should be requested via GET, cached, and set with a <code>Last-modified</code> timestamp and gzipped.</li>
</ol>

<p><img src="/wp-content/uploads/2007/06/book-high-performance-web-sites" alt="Book cover: High Performance Web Sites" width="120" height="158" class="floatleft book" /> These are a lot of rules, and they will be published in a O&rsquo;Reilly book by Steve Souders and Tenni Theurer in September 2007. Anyway, don&rsquo;t be overwhelmed by their mass, instead you can start with the easy things: <strong>&ldquo;<q>harvest the low hanging fruit</q>.&rdquo;</strong> Enable caching with the Expire date setting and reduce the number of HTTP requests. You can deal with the rest later.</p>

<p>Finally Nate Koechley announced a Yahoo! performance tool called <strong><a href="http://developer.yahoo.com/yslow/">YSlow</a></strong> as a plugin for the indespensible <a href="http://www.getfirebug.com">Firebug</a> extension. He also recommended the commercial <a href="http://alphaworks.ibm.com/tech/pagedetailer">IBM Page Detailer</a>, and <a href="http://livehttpheaders.mozdev.org">LiveHTTPHeaders</a> to visualize what&rsquo;s happening in your browser.</p>
]]></content:encoded>
			<wfw:commentRss>http://learningtheworld.eu/2007/performance-2/feed/</wfw:commentRss>
		<slash:comments>3</slash:comments>
<enclosure url="http://www.htmldog.com/atmedia2007/highperformancewebpages.mp3" length="26276424" type="audio/mpeg" />
		</item>
		<item>
		<title>Website Performance Tweaks</title>
		<link>http://learningtheworld.eu/2007/performance/</link>
		<comments>http://learningtheworld.eu/2007/performance/#comments</comments>
		<pubDate>Thu, 25 Jan 2007 20:00:35 +0000</pubDate>
		<dc:creator><![CDATA[Martin Kliehm]]></dc:creator>
				<category><![CDATA[downloads]]></category>
		<category><![CDATA[web development]]></category>
		<category><![CDATA[book:isbn=0596529309]]></category>
		<category><![CDATA[cache]]></category>
		<category><![CDATA[caching]]></category>
		<category><![CDATA[concat]]></category>
		<category><![CDATA[css sprites]]></category>
		<category><![CDATA[development]]></category>
		<category><![CDATA[Douglas Crockford]]></category>
		<category><![CDATA[Ed Eliot]]></category>
		<category><![CDATA[file aggregation]]></category>
		<category><![CDATA[http-request]]></category>
		<category><![CDATA[JSMin]]></category>
		<category><![CDATA[mozilla]]></category>
		<category><![CDATA[Nate Koechley]]></category>
		<category><![CDATA[performance]]></category>
		<category><![CDATA[sliding doors]]></category>
		<category><![CDATA[techniques]]></category>
		<category><![CDATA[yahoo]]></category>
		<category><![CDATA[yui]]></category>

		<guid isPermaLink="false">http://learningtheworld.eu/2007/performance/</guid>
		<description><![CDATA[In the last six months I became more aware of techniques for optimizing website performance. I learned about memory leaks and JavaScript performance, but what impressed me most was Nate Koechleyâ€™s presentation about large scale website performance issues in â€œYahoo! <abbr title="versus">vs.</abbr> Yahoo!&#8221; at the @media conference 2006. In the meantime there have been more blog posts about particular aspects of performance optimization, so I wrote a summary.&#160;[&#8230;]]]></description>
				<content:encoded><![CDATA[<p>In the last six months I became more aware of techniques for <strong>optimizing website performance</strong>. I learned about <a href="http://msdn.microsoft.com/library/default.asp?url=/library/en-us/IETechCol/dnwebgen/ie_leak_patterns.asp" title="Microsoft Developer Network: Understanding and solving Internet Explorer leak patterns">memory</a> <a href="http://outofhanwell.com/ieleak/" title="Drip: A memory leak detector for Internet Explorer">leaks</a> and <a href="http://blogs.msdn.com/ie/archive/2006/08/28/728654.aspx" title="IE + JavaScript performance recommendations &ndash; part 1">JavaScript</a> <a href="http://blogs.msdn.com/ie/archive/2006/11/16/ie-javascript-performance-recommendations-part-2-javascript-code-inefficiencies.aspx" title="IE + JavaScript performance recommendations &ndash; part 2">performance</a>, but what impressed me most was <a href="http://nate.koechley.com/blog/2006/07/12/my_atmedia_2006_slides/" rel="met colleague">Nate Koechley&rsquo;s presentation</a> about large scale website performance issues in &ldquo;<a href="http://learningtheworld.eu/2006/atmedia-day-two/#koechley" title="See my notes about his talk">Yahoo! <abbr title="versus">vs.</abbr> Yahoo!</a>&rdquo; at the @media conference 2006. In the meantime there have been more blog posts about particular aspects of performance optimization, and I&rsquo;d like to sum them up:</p>

<p id="file-location"><strong>Parsing JavaScript</strong> freezes the browser. Therefore put <acronym title="Cascading Stylesheets">CSS</acronym> in the <code>head</code> and JavaScript near to the <code>&lt;/body&gt;</code> so that it is parsed when the page has been rendered.</p>

<p id="http-requests">The arch enemy of performance are <strong><a href="http://yuiblog.com/blog/2006/11/28/performance-research-part-1/" title="YUI Blog: Performance research &ndash; what the 80/20 rule tells us about reducing HTTP requests"><acronym title="Hypertext Transfer Protocol">HTTP</acronym> requests</a></strong>. Many browsers still can&rsquo;t handle more than two or four requests at a time. Keep the number of files down, your website will be faster.</p>

<p>There are several techniques with the aim to reduce the number of files:</p>

<ol><li id="inline-css"><p><strong>&ldquo;<q>A single large file is fastest.</q>&rdquo;</strong> (<cite>Nate Koechley</cite>) That&rsquo;s why Yahoo! <em>apparently</em> has such an amount of <a href="http://www.robertnyman.com/2007/01/24/with-these-web-sites-would-you-say-the-web-standards-war-is-won/">inline <acronym title="Cascading Stylesheets">CSS</acronym></a>. They found out <a href="http://yuiblog.com/blog/2007/01/04/performance-research-part-2/" title="YUI Blog: Performance research, part 2: browser cache usage &ndash; exposed!">browser caching</a> is not as effective as they thought, in particular not on a user&rsquo;s start page. So they deliver &ldquo;inline&rdquo; <acronym>CSS</acronym>. Actually writing inline <acronym>CSS</acronym> is a maintenance nightmare, but delivering <acronym>CSS</acronym> content inline doesn&rsquo;t mean the files can&rsquo;t have separate lives on the server: concatenate the files with a server side technique of your choice.</p>
<p><strong>Update:</strong> A couple of months later <a href="/2007/performance-2/#rule-8">Nate explained that further</a>: when your page is likely to be a user&rsquo;s start page, caching plays a minor role, thus &ldquo;inline&rdquo; <acronym>CSS</acronym> is faster. Otherwise use external files, aggregate them, and make sure they are cached (see below).</p></li>
<li id="enforce-caching"><p><strong>Enforce caching.</strong> Another <a href="http://www.bazon.net/mishoo/articles.epl?art_id=958"><acronym title="Internet Explorer">IE</acronym> bug</a> prevents image caching. Add the following to your <code>.htaccess</code>, <code>httpd.conf</code> or <code>vhost.conf</code> settings:</p>
<ol class="code">
<li><code>&lt;IfModule mod_expires.c&gt;</code></li>
<li class="indent"><code>ExpiresActive On</code></li>
<li class="indent"><code>ExpiresByType image/jpg &quot;access plus 1 day&quot;</code></li>
<li class="indent"><code>ExpiresByType image/jpeg &quot;access plus 1 day&quot;</code></li>
<li class="indent"><code>ExpiresByType image/gif &quot;access plus 1 day&quot;</code></li>
<li class="indent"><code>ExpiresByType image/png &quot;access plus 1 day&quot;</code></li>
<li><code>&lt;/IfModule&gt;</code></li></ol></li>
<li id="background-images"><p><strong>Reduce the number of background images</strong> with techniques like <a href="http://www.alistapart.com/articles/sprites/">CSS Sprites</a> or <a href="http://www.alistapart.com/articles/slidingdoors/">Sliding Doors</a>. Instead of four images of rounded corners you <a href="http://www.fiftyfoureleven.com/sandbox/sliding-doors-one-image/" title="Example">only need one</a> and get the mouseover state for free! The green download button on <a href="http://www.mozilla.com">mozilla.com</a> is based on that technique. And <a href="http://www.yahoo.com">Yahoo!</a> uses <acronym>CSS</acronym> Sprites to combine a huge number of icons.</p>

<p><img src="/wp-content/uploads/2007/01/mozilla-button.jpg" class="centered screenshot" width="300" height="149" alt="Download button on mozilla.com using the Sliding Doors technique" /></p>

<p>Please note this approach is only for <em>decorational background images</em> that degrade gracefully. <del>It&rsquo;s not for <code>img</code> elements.</del> <ins>Be careful when you use it for <a href="/2007/foreground-sprites/">foreground images</a>.</ins> And if text comes as a graphical representation, it can become inaccessible for screen reader users, zoom readers, or people with stylesheets switched off. Use real text instead.</p>

<p>Also note changing the <code>background-position</code> causes <acronym>IE6</acronym> to flicker, related to the caching bug above. To avoid it, simply add the following:</p>
<ol class="code">
<li><code>&lt;script type=&quot;text/javascript&quot;&gt;</code></li>
<li class="indent"><code>try { document.execCommand(<span class="codeSpace">&nbsp;</span>&quot;BackgroundImageCache&quot;, false, true); } catch(e) {};</code></li>
<li><code>&lt;/script&gt;</code></li></ol></li>
<li id="file-aggregation"><p><strong>Aggregate files.</strong> Ed Eliot wrote a nice <a href="http://www.ejeliot.com/blog/72" title="Automatic merging and versioning of CSS/JS files with PHP">script to merge JavaScript or <acronym>CSS</acronym> files</a>, bonus respect for the advanced versioning and caching features.</p>

<p>But remember the cases when it doesn&rsquo;t make sense to merge <acronym>CSS</acronym> files: your <acronym title="Internet Explorer">IE</acronym> bugfixes still belong in conditional comments. If you use the <code>@import</code> rule to filter antique browsers from getting advanced styles, you can&rsquo;t drop it. And if you want to merge stylesheets for different media (<abbr title="for example">e.g.</abbr> print), make sure the code is enclosed in something like</p>


<ol class="code">
<li><code>@media print {</code></li>
<li class="indent"><code>/* style sheet for print goes here */</code></li>
<li><code>}</code></li></ol>


<p>In an <a href="http://www.ejeliot.com/blog/73" title="Adding JSMin to the CSS/JS merging script">updated version</a> Ed added <a href="http://javascript.crockford.com/jsmin.html">JSMin</a> to strip comments and excess whitespace. JSMin works like a charm for JavaScript files. But it cuts a few space characters too much so that the syntax of <acronym>CSS</acronym> selectors changes <del>therefore for now I have abandoned the idea to compress them too</del>. <ins>See <a href="#comment-6045">Jens Meiert&rsquo;s comment</a> below for a recommendation to minimize <acronym>CSS</acronym>.</ins></p>

<p>His original code requires the C version of JSMin with PHP <code>safe_mode</code> turned off. If you prefer a pure PHP version, get the <a href="http://javascript.crockford.com/jsmin2.php.txt">PHP version of JSMin</a> and my <a href="/examples/combine-jsmin.phps" type="text/plain">adapted version of the script</a>.</p>
</li></ol>

<p>I&rsquo;m still in awe how fast one of my own websites became! Thanks to the guys at Yahoo! for the inspiration and for most of the research this article is based upon. Even JSMin was written by an employee of Yahoo! Speaking about Yahoo! employees: <a href="http://wait-till-i.com/" title="Christian Heilmann" rel="met colleague">Chris</a>, I hope there are still enough topics for your <a href="http://www.thinkvitamin.com/features/dev/enhance-your-page-performance" title="Chris Heilmann: Enhance your (page) performance!">Vitamin article</a>. I wanted to write about performance anyway, and to my surprise I <a href="http://www.robertnyman.com/2007/01/24/with-these-web-sites-would-you-say-the-web-standards-war-is-won/#comment-29959">read yesterday</a> that you have similar plans. See ya in <a href="/2007/brain-food/#e-accessibility" title="First European e-Accessibility Forum">Paris</a>. <img src="http://learningtheworld.eu/wp-includes/images/smilies/icon_wink.gif" alt=";)" class="wp-smiley" /> </p>
]]></content:encoded>
			<wfw:commentRss>http://learningtheworld.eu/2007/performance/feed/</wfw:commentRss>
		<slash:comments>18</slash:comments>
		</item>
	</channel>
</rss>
