<?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: Don&#8217;t Unit Test Anymore&#8230; No, Really!</title>
	<atom:link href="http://cwash.org/2009/02/17/dont-unit-test-anymore-no-really/feed/" rel="self" type="application/rss+xml" />
	<link>http://cwash.org/2009/02/17/dont-unit-test-anymore-no-really/?utm_source=rss&amp;utm_medium=rss&amp;utm_campaign=dont-unit-test-anymore-no-really</link>
	<description>+= construction + craftsmanship;</description>
	<lastBuildDate>Fri, 26 Feb 2010 06:16:32 +0000</lastBuildDate>
	<sy:updatePeriod>hourly</sy:updatePeriod>
	<sy:updateFrequency>1</sy:updateFrequency>
	<generator>http://wordpress.org/?v=3.0.1</generator>
	<item>
		<title>By: Gregory Kornblum</title>
		<link>http://cwash.org/2009/02/17/dont-unit-test-anymore-no-really/comment-page-1/#comment-3080</link>
		<dc:creator>Gregory Kornblum</dc:creator>
		<pubDate>Thu, 19 Feb 2009 01:08:07 +0000</pubDate>
		<guid isPermaLink="false">http://cwash.org/?p=133#comment-3080</guid>
		<description>You definitely have something there when you mention the fact of how doing these &quot;tests&quot; relate to cost. Spending the man hours bending your development to follow the &quot;rules&quot; of unit testing is costly. I have always tested my development as soon as I had something to the point of where it can be tested and eventually full end to end testing in the past 13 years. In the end I have achieved the same results and spent much less time doing so. I continue to do so and save the company money to be spent on bigger and better things. I also use my own brain and not follow along like the rest of the sheep. Nowadays it seems the programmers are being programmed.</description>
		<content:encoded><![CDATA[<p>You definitely have something there when you mention the fact of how doing these &#8220;tests&#8221; relate to cost. Spending the man hours bending your development to follow the &#8220;rules&#8221; of unit testing is costly. I have always tested my development as soon as I had something to the point of where it can be tested and eventually full end to end testing in the past 13 years. In the end I have achieved the same results and spent much less time doing so. I continue to do so and save the company money to be spent on bigger and better things. I also use my own brain and not follow along like the rest of the sheep. Nowadays it seems the programmers are being programmed.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: bob</title>
		<link>http://cwash.org/2009/02/17/dont-unit-test-anymore-no-really/comment-page-1/#comment-3079</link>
		<dc:creator>bob</dc:creator>
		<pubDate>Thu, 19 Feb 2009 00:52:13 +0000</pubDate>
		<guid isPermaLink="false">http://cwash.org/?p=133#comment-3079</guid>
		<description>TDD is a joke.  Make it stop.</description>
		<content:encoded><![CDATA[<p>TDD is a joke.  Make it stop.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: dlm</title>
		<link>http://cwash.org/2009/02/17/dont-unit-test-anymore-no-really/comment-page-1/#comment-3077</link>
		<dc:creator>dlm</dc:creator>
		<pubDate>Wed, 18 Feb 2009 17:28:04 +0000</pubDate>
		<guid isPermaLink="false">http://cwash.org/?p=133#comment-3077</guid>
		<description>I used to do write software and do software test, long ago, for a company whose bread and butter at that location was Independent Software V&amp;V.  We had thick, detailed books of requirements to test against.  After a while it became obvious what an inefficient thing it was, to write software, THEN test it.  Shouldn&#039;t the things it was going to have to be able to do be foremost in your mind as you were writing the code?  Prevent errors, instead of catching them?  When I started reading about TDD and unit testing, it sounded like exactly what we&#039;d been thinking of.  With all the same issues of misinterpreting requirements, miscoding, and partitioning the problem - 2X.  Still sounds good, though, to at least be defensively focusing on the requirements when you start to write.</description>
		<content:encoded><![CDATA[<p>I used to do write software and do software test, long ago, for a company whose bread and butter at that location was Independent Software V&amp;V.  We had thick, detailed books of requirements to test against.  After a while it became obvious what an inefficient thing it was, to write software, THEN test it.  Shouldn&#8217;t the things it was going to have to be able to do be foremost in your mind as you were writing the code?  Prevent errors, instead of catching them?  When I started reading about TDD and unit testing, it sounded like exactly what we&#8217;d been thinking of.  With all the same issues of misinterpreting requirements, miscoding, and partitioning the problem &#8211; 2X.  Still sounds good, though, to at least be defensively focusing on the requirements when you start to write.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Kevin Hazzard</title>
		<link>http://cwash.org/2009/02/17/dont-unit-test-anymore-no-really/comment-page-1/#comment-3076</link>
		<dc:creator>Kevin Hazzard</dc:creator>
		<pubDate>Wed, 18 Feb 2009 16:30:33 +0000</pubDate>
		<guid isPermaLink="false">http://cwash.org/?p=133#comment-3076</guid>
		<description>Semantics are important but I would be happy to just get the average developer into a testing mindset. Most organizations, if they make the mistake of displacing integration and system testing with just the tests that developers write, will come back to their senses eventually. And the unit-test-only bigots who led them down that primrose path will mature.

Honestly, though, I’m not worried about the semantics or the test-happy developers. Having test-happy developers is a good thing, even if they don’t yet understand what they are doing because they have the right idea (and maybe the wrong execution). I’m worried that we’re a solid decade into understanding the best practices for unit testing and the majority of developers I come into contact with don’t have a clue how to how to build developer based testing into their methodology. And I’m not even a TDD extremist. I’ll settle for TED (Test Eventually Development).</description>
		<content:encoded><![CDATA[<p>Semantics are important but I would be happy to just get the average developer into a testing mindset. Most organizations, if they make the mistake of displacing integration and system testing with just the tests that developers write, will come back to their senses eventually. And the unit-test-only bigots who led them down that primrose path will mature.</p>
<p>Honestly, though, I’m not worried about the semantics or the test-happy developers. Having test-happy developers is a good thing, even if they don’t yet understand what they are doing because they have the right idea (and maybe the wrong execution). I’m worried that we’re a solid decade into understanding the best practices for unit testing and the majority of developers I come into contact with don’t have a clue how to how to build developer based testing into their methodology. And I’m not even a TDD extremist. I’ll settle for TED (Test Eventually Development).</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Joerg</title>
		<link>http://cwash.org/2009/02/17/dont-unit-test-anymore-no-really/comment-page-1/#comment-3075</link>
		<dc:creator>Joerg</dc:creator>
		<pubDate>Wed, 18 Feb 2009 12:34:23 +0000</pubDate>
		<guid isPermaLink="false">http://cwash.org/?p=133#comment-3075</guid>
		<description>&lt;b&gt;My understanding always (at least since my studies in the nineties) has been, that &quot;unit tests&quot; are tests against isolated sofware units (ie. modules or functions).&lt;/b&gt; These are usually created and executed/scheduled in regression by developers. The concept is older than the popular frameworks.
Nevertheless, I think you should not confuse unit testing (which are developer practices) with QA tests of a software as a whole (end-to-end if you like).
Both are necessary, if you want your software to reach a certain degree of quality. I&#039;d recommend to also add integration testing to your software process - this makes for higher quality test candidates for the QA  as it allows combined testing of a whole bunch of units in concert.
By the way, the QA role in some cases might be very well made up of members of a developers team. So the term &quot;developer test&quot; can be just as misleading or ambiguous. Just my 2 ct.</description>
		<content:encoded><![CDATA[<p><b>My understanding always (at least since my studies in the nineties) has been, that &#8220;unit tests&#8221; are tests against isolated sofware units (ie. modules or functions).</b> These are usually created and executed/scheduled in regression by developers. The concept is older than the popular frameworks.<br />
Nevertheless, I think you should not confuse unit testing (which are developer practices) with QA tests of a software as a whole (end-to-end if you like).<br />
Both are necessary, if you want your software to reach a certain degree of quality. I&#8217;d recommend to also add integration testing to your software process &#8211; this makes for higher quality test candidates for the QA  as it allows combined testing of a whole bunch of units in concert.<br />
By the way, the QA role in some cases might be very well made up of members of a developers team. So the term &#8220;developer test&#8221; can be just as misleading or ambiguous. Just my 2 ct.</p>
]]></content:encoded>
	</item>
	<item>
		<title>By: Janusz Gorycki</title>
		<link>http://cwash.org/2009/02/17/dont-unit-test-anymore-no-really/comment-page-1/#comment-3074</link>
		<dc:creator>Janusz Gorycki</dc:creator>
		<pubDate>Wed, 18 Feb 2009 11:02:36 +0000</pubDate>
		<guid isPermaLink="false">http://cwash.org/?p=133#comment-3074</guid>
		<description>I think you may actually be right that we should just stop using the term &quot;unit tests&quot;, because the meaning of this term has been twisted and distorted to pretty the point where it lost its original meaning. Re-badging it to &quot;Developer tests&quot; could be beneficial, as it better describes what these tests are for. Also - it makes it clear that these tests are a natural part of what every developer has no choice but do anyway when they write code, regardless of whether they do TDD or not - it is just that if you don&#039;t use automated testing frameworks for the purpose, you end up doing it all manually, and likely with worse results.</description>
		<content:encoded><![CDATA[<p>I think you may actually be right that we should just stop using the term &#8220;unit tests&#8221;, because the meaning of this term has been twisted and distorted to pretty the point where it lost its original meaning. Re-badging it to &#8220;Developer tests&#8221; could be beneficial, as it better describes what these tests are for. Also &#8211; it makes it clear that these tests are a natural part of what every developer has no choice but do anyway when they write code, regardless of whether they do TDD or not &#8211; it is just that if you don&#8217;t use automated testing frameworks for the purpose, you end up doing it all manually, and likely with worse results.</p>
]]></content:encoded>
	</item>
</channel>
</rss>
