<?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: Array Performance</title>
	<atom:link href="http://jacksondunstan.com/articles/529/feed" rel="self" type="application/rss+xml" />
	<link>http://jacksondunstan.com/articles/529</link>
	<description>Mastering AS3</description>
	<lastBuildDate>Thu, 29 Jul 2010 17:30:45 +0000</lastBuildDate>
	<generator>http://wordpress.org/?v=2.9.2</generator>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
		<item>
		<title>By: jpauclair</title>
		<link>http://jacksondunstan.com/articles/529/comment-page-1#comment-452</link>
		<dc:creator>jpauclair</dc:creator>
		<pubDate>Wed, 16 Dec 2009 01:46:48 +0000</pubDate>
		<guid isPermaLink="false">http://jacksondunstan.com/?p=529#comment-452</guid>
		<description>I just added it: https://bugs.adobe.com/jira/browse/FP-3477</description>
		<content:encoded><![CDATA[<p>I just added it: <a href="https://bugs.adobe.com/jira/browse/FP-3477" rel="nofollow">https://bugs.adobe.com/jira/browse/FP-3477</a></p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jackson</title>
		<link>http://jacksondunstan.com/articles/529/comment-page-1#comment-451</link>
		<dc:creator>jackson</dc:creator>
		<pubDate>Tue, 15 Dec 2009 01:41:25 +0000</pubDate>
		<guid isPermaLink="false">http://jacksondunstan.com/?p=529#comment-451</guid>
		<description>Crazy! When you get that Adobe bug filed, reply here and we can vote it up. Thanks for finding this.</description>
		<content:encoded><![CDATA[<p>Crazy! When you get that Adobe bug filed, reply here and we can vote it up. Thanks for finding this.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jpauclair</title>
		<link>http://jacksondunstan.com/articles/529/comment-page-1#comment-450</link>
		<dc:creator>jpauclair</dc:creator>
		<pubDate>Tue, 15 Dec 2009 00:10:24 +0000</pubDate>
		<guid isPermaLink="false">http://jacksondunstan.com/?p=529#comment-450</guid>
		<description>Hey! the reslut you had was not coherent from what I did read in Tamarin source.
Check this out!
http://jpauclair.wordpress.com/2009/12/14/tamarin-part-i-i-as3-array-bug/</description>
		<content:encoded><![CDATA[<p>Hey! the reslut you had was not coherent from what I did read in Tamarin source.<br />
Check this out!<br />
<a href="http://jpauclair.wordpress.com/2009/12/14/tamarin-part-i-i-as3-array-bug/" rel="nofollow">http://jpauclair.wordpress.com/2009/12/14/tamarin-part-i-i-as3-array-bug/</a></p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Tamarin Part I.I &#8211; AS3 Array (bug) &#171; jpauclair</title>
		<link>http://jacksondunstan.com/articles/529/comment-page-1#comment-449</link>
		<dc:creator>Tamarin Part I.I &#8211; AS3 Array (bug) &#171; jpauclair</dc:creator>
		<pubDate>Tue, 15 Dec 2009 00:08:39 +0000</pubDate>
		<guid isPermaLink="false">http://jacksondunstan.com/?p=529#comment-449</guid>
		<description>[...] a comment &#187;  Thanks to the analysis Jackson Dunstan, I was able to find a &#8220;bug&#8221; in the Array [...]</description>
		<content:encoded><![CDATA[<p>[...] a comment &raquo;  Thanks to the analysis Jackson Dunstan, I was able to find a &#8220;bug&#8221; in the Array [...]</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jackson</title>
		<link>http://jacksondunstan.com/articles/529/comment-page-1#comment-448</link>
		<dc:creator>jackson</dc:creator>
		<pubDate>Mon, 14 Dec 2009 17:17:45 +0000</pubDate>
		<guid isPermaLink="false">http://jacksondunstan.com/?p=529#comment-448</guid>
		<description>The test app shows that step C is required by the testing times. Iterating is initially fast (dense reads). I then delete element 0 and iterating becomes slow (sparse reads). I then fill the undefined value at 0 with a non-undefined value (33), but iterating is still slow (indicating sparse). I then fill with an overlap by filling in element 0 and element 1, but iterating is still slow (sparse). I then fill in all the elements again and iterating is fast again (dense restored).

Good point about the use of &lt;tt&gt;null&lt;/tt&gt;. That would be a good choice for an &lt;tt&gt;Array&lt;/tt&gt; where &lt;tt&gt;null&lt;/tt&gt; is not a naturally-occurring element. For an array of class instances though, it just might. Perhaps in that case it would be better to use -1 and in the case of an &lt;tt&gt;Array&lt;/tt&gt; of &lt;tt&gt;int&lt;/tt&gt;, &lt;tt&gt;Number&lt;/tt&gt;, or the like to use &lt;tt&gt;null&lt;/tt&gt;. Thanks for the suggestion!</description>
		<content:encoded><![CDATA[<p>The test app shows that step C is required by the testing times. Iterating is initially fast (dense reads). I then delete element 0 and iterating becomes slow (sparse reads). I then fill the undefined value at 0 with a non-undefined value (33), but iterating is still slow (indicating sparse). I then fill with an overlap by filling in element 0 and element 1, but iterating is still slow (sparse). I then fill in all the elements again and iterating is fast again (dense restored).</p>
<p>Good point about the use of <tt>null</tt>. That would be a good choice for an <tt>Array</tt> where <tt>null</tt> is not a naturally-occurring element. For an array of class instances though, it just might. Perhaps in that case it would be better to use -1 and in the case of an <tt>Array</tt> of <tt>int</tt>, <tt>Number</tt>, or the like to use <tt>null</tt>. Thanks for the suggestion!</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: jpauclair</title>
		<link>http://jacksondunstan.com/articles/529/comment-page-1#comment-447</link>
		<dc:creator>jpauclair</dc:creator>
		<pubDate>Mon, 14 Dec 2009 12:33:04 +0000</pubDate>
		<guid isPermaLink="false">http://jacksondunstan.com/?p=529#comment-447</guid>
		<description>Thanks a lot for adding to my article. I appreciate it. 

Though I&#039;m not sure of your conclusion.
Why is step C required? Are you talking about 
an array with multiple undefined values?
You only need to fill &quot;undefined&quot; value. so when you remove the first element, you must fill that one only to get back the whole dense array. tamarin is doing the iterative work to transfert all values (until first undefined) to the dense part.

as for the &quot;-1&quot; value, it can also be &quot;null&quot;. since null is a valid atom for tamarin, you can use null for filling holes.
If you have an array with 75% of valid values in it, I would suggest filling all others with &quot;null&quot; values to make your array faster. 
If you read my next article you&#039;ll see flash is allocating memory in block of 4K, so filing empty values or nulll-initialisation of values is not heavy on memory, it will only take some more process at the beginning. http://jpauclair.wordpress.com/2009/12/05/tamarin-part-ii-more-on-array-and-vector/</description>
		<content:encoded><![CDATA[<p>Thanks a lot for adding to my article. I appreciate it. </p>
<p>Though I&#8217;m not sure of your conclusion.<br />
Why is step C required? Are you talking about<br />
an array with multiple undefined values?<br />
You only need to fill &#8220;undefined&#8221; value. so when you remove the first element, you must fill that one only to get back the whole dense array. tamarin is doing the iterative work to transfert all values (until first undefined) to the dense part.</p>
<p>as for the &#8220;-1&#8243; value, it can also be &#8220;null&#8221;. since null is a valid atom for tamarin, you can use null for filling holes.<br />
If you have an array with 75% of valid values in it, I would suggest filling all others with &#8220;null&#8221; values to make your array faster.<br />
If you read my next article you&#8217;ll see flash is allocating memory in block of 4K, so filing empty values or nulll-initialisation of values is not heavy on memory, it will only take some more process at the beginning. <a href="http://jpauclair.wordpress.com/2009/12/05/tamarin-part-ii-more-on-array-and-vector/" rel="nofollow">http://jpauclair.wordpress.com/2009/12/05/tamarin-part-ii-more-on-array-and-vector/</a></p>
]]></content:encoded>
	</item>
</channel>
</rss>
