<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"
	xmlns:content="http://purl.org/rss/1.0/modules/content/"
	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/"
		>
<channel>
	<title>Comments on: Work with gzip, not against it</title>
	<atom:link href="http://www.karmagination.com/blog/2009/09/01/work-with-gzip-not-against-it/feed/" rel="self" type="application/rss+xml" />
	<link>http://www.karmagination.com/blog/2009/09/01/work-with-gzip-not-against-it/</link>
	<description>Creation begins with imagination</description>
	<lastBuildDate>Wed, 19 May 2010 13:01:48 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.0.1</generator>
	<item>
		<title>By: Paul Irish</title>
		<link>http://www.karmagination.com/blog/2009/09/01/work-with-gzip-not-against-it/comment-page-1/#comment-58</link>
		<dc:creator>Paul Irish</dc:creator>
		<pubDate>Fri, 04 Sep 2009 08:28:52 +0000</pubDate>
		<guid isPermaLink="false">http://www.karmagination.com/blog/?p=86#comment-58</guid>
		<description>Based on Ywg&#039;s comments on Andrea&#039;s post: http://webreflection.blogspot.com/2009/02/on-my-vicious-javascript-code.html

... I&#039;m starting to believe the dictionary reduction wouldn&#039;t be all that successful. However I would certainly think this should happen from within the YUI Compressor, and not manually.</description>
		<content:encoded><![CDATA[<p>Based on Ywg&#8217;s comments on Andrea&#8217;s post: <a href="http://webreflection.blogspot.com/2009/02/on-my-vicious-javascript-code.html" rel="nofollow">http://webreflection.blogspot.com/2009/02/on-my-vicious-javascript-code.html</a></p>
<p>&#8230; I&#8217;m starting to believe the dictionary reduction wouldn&#8217;t be all that successful. However I would certainly think this should happen from within the YUI Compressor, and not manually.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Kean</title>
		<link>http://www.karmagination.com/blog/2009/09/01/work-with-gzip-not-against-it/comment-page-1/#comment-56</link>
		<dc:creator>Kean</dc:creator>
		<pubDate>Thu, 03 Sep 2009 22:57:16 +0000</pubDate>
		<guid isPermaLink="false">http://www.karmagination.com/blog/?p=86#comment-56</guid>
		<description>@Paul The optimization that you talked about can possibly be done but I predict that they will run into unforeseen problems if they do that.</description>
		<content:encoded><![CDATA[<p>@Paul The optimization that you talked about can possibly be done but I predict that they will run into unforeseen problems if they do that.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Kean</title>
		<link>http://www.karmagination.com/blog/2009/09/01/work-with-gzip-not-against-it/comment-page-1/#comment-54</link>
		<dc:creator>Kean</dc:creator>
		<pubDate>Thu, 03 Sep 2009 22:51:54 +0000</pubDate>
		<guid isPermaLink="false">http://www.karmagination.com/blog/?p=86#comment-54</guid>
		<description>@Craig I agree with you, we should sit back, relax and let gzip do its job. This will make the code more readable and you don&#039;t have to do the extra work.</description>
		<content:encoded><![CDATA[<p>@Craig I agree with you, we should sit back, relax and let gzip do its job. This will make the code more readable and you don&#8217;t have to do the extra work.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Craig</title>
		<link>http://www.karmagination.com/blog/2009/09/01/work-with-gzip-not-against-it/comment-page-1/#comment-53</link>
		<dc:creator>Craig</dc:creator>
		<pubDate>Thu, 03 Sep 2009 18:33:17 +0000</pubDate>
		<guid isPermaLink="false">http://www.karmagination.com/blog/?p=86#comment-53</guid>
		<description>I&#039;m rather curious... have you guys heard of over-optimization?

You&#039;re making your code much harder to work with (in terms of maintenance and initial authoring) to save (literally!) a few bytes.

This conversation is fascinating from a computer science perspective (I love thinking about this stuff, too), but the cost is incredibly high. If anything, the fruits of this research should appear as improvement to YUI Compressor and other minifiers, and *definitely* not as a guide for day-to-day javascript programmers.</description>
		<content:encoded><![CDATA[<p>I&#8217;m rather curious&#8230; have you guys heard of over-optimization?</p>
<p>You&#8217;re making your code much harder to work with (in terms of maintenance and initial authoring) to save (literally!) a few bytes.</p>
<p>This conversation is fascinating from a computer science perspective (I love thinking about this stuff, too), but the cost is incredibly high. If anything, the fruits of this research should appear as improvement to YUI Compressor and other minifiers, and *definitely* not as a guide for day-to-day javascript programmers.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Paul Irish</title>
		<link>http://www.karmagination.com/blog/2009/09/01/work-with-gzip-not-against-it/comment-page-1/#comment-52</link>
		<dc:creator>Paul Irish</dc:creator>
		<pubDate>Thu, 03 Sep 2009 18:24:14 +0000</pubDate>
		<guid isPermaLink="false">http://www.karmagination.com/blog/?p=86#comment-52</guid>
		<description>More detail: The idea is to identify the javascript keywords that are in use and can be use legally as variable names (not reserved words or anything). Sort according to string size.  Use those up and then go back to A, B, C, D, etc.

So:
Input:
(function(htmlelem,className){htmlelem[className]= htmlelem[className].replace(/\bnojs\b/,&#039;js&#039;)})(document.documentElement,&#039;className&#039;);

current output:
(function(B,C){B[C]=B[C].replace(/\bnojs\b/,&#039;js&#039;)})(document.documentElement,&#039;className&#039;);

proposed output:
(function(document,replace){document[replace]=document[replace].replace(/\bnojs\b/,&#039;js&#039;)})(document.documentElement,&#039;className&#039;);

Based on gzip&#039;s technology I think this could be a measurable win for gzip-enabled agents. Thoughts?</description>
		<content:encoded><![CDATA[<p>More detail: The idea is to identify the javascript keywords that are in use and can be use legally as variable names (not reserved words or anything). Sort according to string size.  Use those up and then go back to A, B, C, D, etc.</p>
<p>So:<br />
Input:<br />
(function(htmlelem,className){htmlelem[className]= htmlelem[className].replace(/\bnojs\b/,&#8217;js&#8217;)})(document.documentElement,&#8217;className&#8217;);</p>
<p>current output:<br />
(function(B,C){B[C]=B[C].replace(/\bnojs\b/,&#8217;js&#8217;)})(document.documentElement,&#8217;className&#8217;);</p>
<p>proposed output:<br />
(function(document,replace){document[replace]=document[replace].replace(/\bnojs\b/,&#8217;js&#8217;)})(document.documentElement,&#8217;className&#8217;);</p>
<p>Based on gzip&#8217;s technology I think this could be a measurable win for gzip-enabled agents. Thoughts?</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Paul Irish</title>
		<link>http://www.karmagination.com/blog/2009/09/01/work-with-gzip-not-against-it/comment-page-1/#comment-51</link>
		<dc:creator>Paul Irish</dc:creator>
		<pubDate>Thu, 03 Sep 2009 04:12:23 +0000</pubDate>
		<guid isPermaLink="false">http://www.karmagination.com/blog/?p=86#comment-51</guid>
		<description>If this is true, then I&#039;d say we&#039;d get optimal results if we, as developers, munged our strings, but our minifiers rename the variables to (un-minifiable) strings already in use:
constructor, length, nodeName, script, break, slice, etc.

Would make the compressed source reallllly hard to follow but I think it&#039;d lead to savings after gzip, yes?</description>
		<content:encoded><![CDATA[<p>If this is true, then I&#8217;d say we&#8217;d get optimal results if we, as developers, munged our strings, but our minifiers rename the variables to (un-minifiable) strings already in use:<br />
constructor, length, nodeName, script, break, slice, etc.</p>
<p>Would make the compressed source reallllly hard to follow but I think it&#8217;d lead to savings after gzip, yes?</p>
]]></content:encoded>
	</item>
</channel>
</rss>
