<?xml version="1.0" ?>
<rss version="2.0">
	<channel>
		<title>Keith's Weblog: Comments on &quot;ScriptServer - client/server communication between Javascript and PHP&quot;</title>
		<description>Keith's Weblog: Comments on &quot;ScriptServer - client/server communication between Javascript and PHP&quot;, posted on October 6, 2004</description>
		<link>http://keithdevens.com/weblog/archive/2004/Oct/06/ScriptServer</link>

		<category>Programming</category>
		<language>en-us</language>
		<image>
			<link>http://keithdevens.com/weblog</link>
			<title>Keith Devens .com</title>
			<url>http://keithdevens.com/images/kbd.gif</url>
		</image>

		<item>
			<title>by Harry Fuecks</title>
			<link>http://keithdevens.com/weblog/archive/2004/Oct/06/ScriptServer#comment5804</link>
			<guid isPermaLink="false">http://keithdevens.com/weblog/5695#comment5804</guid>
			<pubDate>Wed, 06 Oct 2004 08:27:03 +0000</pubDate>
			<description>&lt;p class=&quot;st-markup&quot;&gt;Thanks for the announce. Think I'm confusing everyone because the Sourceforge site also has a fork of Simon's XML-RPC implementation (hence the site's name).&lt;/p&gt;

&lt;p class=&quot;st-markup&quot;&gt;ScriptServer actually sends strings to PHP (from Javascript) ready for unserialize() while in the opposite direction it generates Javascript wrapped in an anonymous function e.g. &lt;a href=&quot;http://xmlrpccom.sourceforge.net/scriptserver/scriptserver/examples/postoffice_server.php/colors/listcolors&quot;&gt;response from Colors::listColors&lt;/a&gt; - on the Javascript side use eval to &amp;quot;unserialize&amp;quot; the response. Basically uses inbuilt parsers already available in PHP and Javascript; parsing (e.g. XML) is usually alot harder than generating.&lt;/p&gt;

</description>
		</item>
		<item>
			<title>by Keith</title>
			<link>http://keithdevens.com/weblog/archive/2004/Oct/06/ScriptServer#comment5805</link>
			<guid isPermaLink="false">http://keithdevens.com/weblog/5695#comment5805</guid>
			<pubDate>Wed, 06 Oct 2004 08:29:51 +0000</pubDate>
			<description>&lt;p class=&quot;st-markup&quot;&gt;Even better! Who needs XML? Kick ass. Now I need an app to try this out on.&lt;/p&gt;

</description>
		</item>
		<item>
			<title>by Tim Wood</title>
			<link>http://keithdevens.com/weblog/archive/2004/Oct/06/ScriptServer#comment6015</link>
			<guid isPermaLink="false">http://keithdevens.com/weblog/5695#comment6015</guid>
			<pubDate>Sat, 23 Oct 2004 18:12:51 +0000</pubDate>
			<description>&lt;p class=&quot;st-markup&quot;&gt;It all sounds wonderful and good.  Problems: &lt;/p&gt;

&lt;p class=&quot;st-markup&quot;&gt;1) The docs linked above don't link to the .zip file they refer (the one with the nifty server code).  There are .tar.gz files at sourceforge.net/projects/xmlrpccom/&lt;/p&gt;

&lt;p class=&quot;st-markup&quot;&gt;2) The sample code in the docs have &lt;em style=&quot;font-style: normal; text-decoration: underline&quot;&gt;lots&lt;/em&gt; of missing quotes.&lt;/p&gt;

&lt;p class=&quot;st-markup&quot;&gt;3) Once I was able to fix all the problems I could find, The basic example in the quick start part still wouldn't work.&lt;/p&gt;

&lt;p class=&quot;st-markup&quot;&gt;I suspect it'll be a great tool when you don't spend an hour plus with the quickstart and still not even have a working 'hello world' app.  But, till then I'm going to make do with the code I found over at www.paranoidfish.org/boxes/2002/01/30/&lt;/p&gt;

</description>
		</item>
	</channel>
</rss>
