<?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: Thoughts on the Spring Framework</title>
	<atom:link href="http://krook.net/archives/208/feed" rel="self" type="application/rss+xml" />
	<link>http://krook.net/archives/208</link>
	<description>Daniel Krook is an IBM Certified IT Specialist based in the greater New York City area. He has over ten years experience in Web application development and presently builds a cloud infrastructure for IBM using Java EE, DB2, REST, and BlackBerry. He holds certifications in cloud archtecture, PHP, Java EE, BlackBerry, DB2, and Solaris. He occasionally writes PHP-related articles for IBM developerWorks and co-authored the IBM Redbook &#039;Developing PHP Applications for IBM Data Servers.&#039;</description>
	<lastBuildDate>Sat, 17 Mar 2012 10:08:22 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	
	<item>
		<title>By: Leandro</title>
		<link>http://krook.net/archives/208/comment-page-1#comment-78697</link>
		<dc:creator>Leandro</dc:creator>
		<pubDate>Mon, 07 Dec 2009 23:31:04 +0000</pubDate>
		<guid isPermaLink="false">http://krook.net/archives/208#comment-78697</guid>
		<description>&quot;Second, you hide the implementation details if you decide to move your site in 5 years to some newer technology.&quot;
This is very interesting, because all references to the old pages would be lost. I&#039;ll try to avoid to use extensions. =)</description>
		<content:encoded><![CDATA[<p>&#8220;Second, you hide the implementation details if you decide to move your site in 5 years to some newer technology.&#8221;<br />
This is very interesting, because all references to the old pages would be lost. I&#8217;ll try to avoid to use extensions. =)</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Mike</title>
		<link>http://krook.net/archives/208/comment-page-1#comment-45201</link>
		<dc:creator>Mike</dc:creator>
		<pubDate>Sun, 18 May 2008 15:49:02 +0000</pubDate>
		<guid isPermaLink="false">http://krook.net/archives/208#comment-45201</guid>
		<description>@ Daniel:
Very nice article.  Well thought out and well written.

@ Kallin
Actually a web browser does not &quot;request&quot; HTML, it requests the result of an URL.  The primary mode of displaying the result is HTML.  While this may seem like nitpicking, it is an important distinction, particularly in this day and age where it may not even be a browser requesting the URL.

In any case, exposing an extension of any kind is just messy IMO.</description>
		<content:encoded><![CDATA[<p>@ Daniel:<br />
Very nice article.  Well thought out and well written.</p>
<p>@ Kallin<br />
Actually a web browser does not &#8220;request&#8221; HTML, it requests the result of an URL.  The primary mode of displaying the result is HTML.  While this may seem like nitpicking, it is an important distinction, particularly in this day and age where it may not even be a browser requesting the URL.</p>
<p>In any case, exposing an extension of any kind is just messy IMO.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Kallin Nagelberg</title>
		<link>http://krook.net/archives/208/comment-page-1#comment-27604</link>
		<dc:creator>Kallin Nagelberg</dc:creator>
		<pubDate>Wed, 07 Nov 2007 00:11:37 +0000</pubDate>
		<guid isPermaLink="false">http://krook.net/archives/208#comment-27604</guid>
		<description>I&#039;ve been using Spring for sometime now. Even if you don&#039;t use it&#039;s uber-functionality, you may find that it includes many libraries for things such as IO that are just way easier than doing it through the java API.

I do have a slight gripe with the &quot;.htm is foolishness&quot; argument though. In my view, the browser is making a request for HTML so it makes sense for .htm to be in the url. If you were requesting a .gif, you would probably have blah.com/myimage.gif. Why not the same for any other content? You don&#039;t care if the image is coming out of a database, being decrypted and run through a set of filters, you only care about what you get. I would even say it should be .html, but that&#039;s just nitpicking now :).</description>
		<content:encoded><![CDATA[<p>I&#8217;ve been using Spring for sometime now. Even if you don&#8217;t use it&#8217;s uber-functionality, you may find that it includes many libraries for things such as IO that are just way easier than doing it through the java API.</p>
<p>I do have a slight gripe with the &#8220;.htm is foolishness&#8221; argument though. In my view, the browser is making a request for HTML so it makes sense for .htm to be in the url. If you were requesting a .gif, you would probably have blah.com/myimage.gif. Why not the same for any other content? You don&#8217;t care if the image is coming out of a database, being decrypted and run through a set of filters, you only care about what you get. I would even say it should be .html, but that&#8217;s just nitpicking now :).</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Daniel Krook</title>
		<link>http://krook.net/archives/208/comment-page-1#comment-27253</link>
		<dc:creator>Daniel Krook</dc:creator>
		<pubDate>Sat, 03 Nov 2007 16:12:28 +0000</pubDate>
		<guid isPermaLink="false">http://krook.net/archives/208#comment-27253</guid>
		<description>Hi K,

Thanks for your comments.

The three points about Spring that I highlighted for the group were intended to convey the main ideas at a high level.   So my goal there was to introduce the terms that are commonly used, without having to go into too much detail about the technical differences.  

That&#039;s why I left that bullet as both &quot;Inversion of Control / Dependency Injection,&quot; since the audience may hear one or more of those terms used to describe the same concept.  In that case it seems more helpful to keep them together.

To add some more detail about the &quot;.htm&quot; convention... You may have noticed that this blog itself uses WordPress, a PHP application.  It comes with the option to choose URLs that don&#039;t show the file extension at all, thanks to a simple &lt;a href=&quot;http://en.wikipedia.org/wiki/Mod_rewrite&quot; rel=&quot;nofollow&quot;&gt;mod_rewrite&lt;/a&gt; rule.  

If I bothered to configure my Web server not to show its signature and this blog wasn&#039;t so obviously build on WordPress, that would theoretically make it safer from hackers.   

This is a nice goal, but when applied to the using &quot;.htm&quot; to mask the server side implementation argument, it&#039;s pretty weak... there are better ways.   

First off, if you are going to try to hide your implementation from search engines and hackers alike, just drop a file extension completely.  This makes your URL smaller and more likely to contain a higher ratio of relevant semantic information.  Second, you hide the implementation details if you decide to move your site in 5 years to some newer technology.

Unfortunately, there are only two methods for mapping servlets by wildcard in your Web deployment descriptor... either provide a base piece of the URL, such as /servlet/*, or map everything before the file extension, such as *.do.   Dropping the extension means you have to figure out some other pattern that needs to be repeated in your servlet URLs, or manually specify each servlet in web.xml by name.

There is some leeway with &lt;a href=&quot;http://krook.net/ppt/poster-daniel-krook-12-01-2005-1.ppt&quot; rel=&quot;nofollow&quot;&gt;URL structures that can be performed via servlet filters&lt;/a&gt;, but these can only be applied after the mapping in web.xml.

So what do I recommend in place of the &quot;.htm&quot; extension?  No extension at all, if possible.   Then, maybe some mod_rewrite at the Web server level to dodge the other required URL pattern and maybe the context root.  I must meditate on this some more...

Anyway, thanks for your comments K.

-Dan</description>
		<content:encoded><![CDATA[<p>Hi K,</p>
<p>Thanks for your comments.</p>
<p>The three points about Spring that I highlighted for the group were intended to convey the main ideas at a high level.   So my goal there was to introduce the terms that are commonly used, without having to go into too much detail about the technical differences.  </p>
<p>That&#8217;s why I left that bullet as both &#8220;Inversion of Control / Dependency Injection,&#8221; since the audience may hear one or more of those terms used to describe the same concept.  In that case it seems more helpful to keep them together.</p>
<p>To add some more detail about the &#8220;.htm&#8221; convention&#8230; You may have noticed that this blog itself uses WordPress, a PHP application.  It comes with the option to choose URLs that don&#8217;t show the file extension at all, thanks to a simple <a href="http://en.wikipedia.org/wiki/Mod_rewrite" rel="nofollow">mod_rewrite</a> rule.  </p>
<p>If I bothered to configure my Web server not to show its signature and this blog wasn&#8217;t so obviously build on WordPress, that would theoretically make it safer from hackers.   </p>
<p>This is a nice goal, but when applied to the using &#8220;.htm&#8221; to mask the server side implementation argument, it&#8217;s pretty weak&#8230; there are better ways.   </p>
<p>First off, if you are going to try to hide your implementation from search engines and hackers alike, just drop a file extension completely.  This makes your URL smaller and more likely to contain a higher ratio of relevant semantic information.  Second, you hide the implementation details if you decide to move your site in 5 years to some newer technology.</p>
<p>Unfortunately, there are only two methods for mapping servlets by wildcard in your Web deployment descriptor&#8230; either provide a base piece of the URL, such as /servlet/*, or map everything before the file extension, such as *.do.   Dropping the extension means you have to figure out some other pattern that needs to be repeated in your servlet URLs, or manually specify each servlet in web.xml by name.</p>
<p>There is some leeway with <a href="http://krook.net/ppt/poster-daniel-krook-12-01-2005-1.ppt" rel="nofollow">URL structures that can be performed via servlet filters</a>, but these can only be applied after the mapping in web.xml.</p>
<p>So what do I recommend in place of the &#8220;.htm&#8221; extension?  No extension at all, if possible.   Then, maybe some mod_rewrite at the Web server level to dodge the other required URL pattern and maybe the context root.  I must meditate on this some more&#8230;</p>
<p>Anyway, thanks for your comments K.</p>
<p>-Dan</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: KiLVaiDeN</title>
		<link>http://krook.net/archives/208/comment-page-1#comment-27173</link>
		<dc:creator>KiLVaiDeN</dc:creator>
		<pubDate>Fri, 02 Nov 2007 13:40:23 +0000</pubDate>
		<guid isPermaLink="false">http://krook.net/archives/208#comment-27173</guid>
		<description>Hi !

First, I&#039;d like to react to the Spring presentation you did : it&#039;s quite exact, except for the fact that you should make a difference between &quot;Inversion of Control&quot; and &quot;Dependancy Injection&quot; because it&#039;s not quite the same, but it&#039;s a detail. I also think that AOP is a very important point of Spring, but maybe one point you forgot, is the abstraction level of utilities classes, that make your code &quot;library independant&quot;, for example take the ORM package of Spring, you could use it in such a way that nobody would know what kind of persistance library you are using.

I wanted to post for an answer to the &quot;.htm&quot; debate. I totally agree with the fact that using &quot;.htm&quot; can be confusing if there is mixed static HTML content ( even though there is easy ways to get around this ) but then, does it really happen that often ?

I believe that extension is really not that important for referencing. I&#039;d even say that it&#039;s irrelevant. So the argument of people saying that &quot;.htm is good for referencing&quot; is quite out of its time.

BUT the fact that when a user goes to your website and doesn&#039;t know what kind of technology you use to display the dynamic content because of the &quot;.htm&quot; extension, is quite interesting, because the more your hide from the public, the less risk you take when dealing with hackers.

Cheers
K</description>
		<content:encoded><![CDATA[<p>Hi !</p>
<p>First, I&#8217;d like to react to the Spring presentation you did : it&#8217;s quite exact, except for the fact that you should make a difference between &#8220;Inversion of Control&#8221; and &#8220;Dependancy Injection&#8221; because it&#8217;s not quite the same, but it&#8217;s a detail. I also think that AOP is a very important point of Spring, but maybe one point you forgot, is the abstraction level of utilities classes, that make your code &#8220;library independant&#8221;, for example take the ORM package of Spring, you could use it in such a way that nobody would know what kind of persistance library you are using.</p>
<p>I wanted to post for an answer to the &#8220;.htm&#8221; debate. I totally agree with the fact that using &#8220;.htm&#8221; can be confusing if there is mixed static HTML content ( even though there is easy ways to get around this ) but then, does it really happen that often ?</p>
<p>I believe that extension is really not that important for referencing. I&#8217;d even say that it&#8217;s irrelevant. So the argument of people saying that &#8220;.htm is good for referencing&#8221; is quite out of its time.</p>
<p>BUT the fact that when a user goes to your website and doesn&#8217;t know what kind of technology you use to display the dynamic content because of the &#8220;.htm&#8221; extension, is quite interesting, because the more your hide from the public, the less risk you take when dealing with hackers.</p>
<p>Cheers<br />
K</p>
]]></content:encoded>
	</item>
</channel>
</rss>

