Comments on: What Problem Would Next-Generation Functional Testing Tools Solve? http://testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/ Elisabeth Hendrickson's thoughts on Agile, Testing, and Agile Testing. Sat, 04 Sep 2010 11:42:24 +0000 hourly 1 http://wordpress.org/?v=3.0 By: John Lockhart http://testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/#comment-240 John Lockhart Sat, 24 Mar 2007 06:49:46 +0000 http://www.testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/#comment-240 I have just completed an initial automation project for the web booking engine for Air New Zealand, using Canoo Webtest. While challenging, I believe it sucessfully answered many of the issues above. Using an Ant script gave the benefits of Ant and XML and the (mixed blessing) of being able to run without driving a "real" browser or having to modify anything in the web app or server stack. By careful nesting of clearly named xml entities for duplicated function (similar to action words but simpler and perhaps more crude) and the use of ant/java properties for mapping to the web pages and for test data values as well as setting and following clear guidelines we achieved a very useful result very quickly (compared to traditional automation projects, which I've been involved in) and have left a legacy that can be maintained (although not by completely non-technical staff - but is that ever the case with automation? Despite what some companies like Worksoft say, I doubt it). I am not a developer, though I was one in a previous life/decade (but not a java one - boy it's a different world new!), and while challenging both techically and as a paradigm shift, I found the tool useable. regards John (P.S. I met you before this all started at the Wellington conference in Aug 06 Elisabeth). I have just completed an initial automation project for the web booking engine for Air New Zealand, using Canoo Webtest. While challenging, I believe it sucessfully answered many of the issues above. Using an Ant script gave the benefits of Ant and XML and the (mixed blessing) of being able to run without driving a “real” browser or having to modify anything in the web app or server stack. By careful nesting of clearly named xml entities for duplicated function (similar to action words but simpler and perhaps more crude) and the use of ant/java properties for mapping to the web pages and for test data values as well as setting and following clear guidelines we achieved a very useful result very quickly (compared to traditional automation projects, which I’ve been involved in) and have left a legacy that can be maintained (although not by completely non-technical staff – but is that ever the case with automation? Despite what some companies like Worksoft say, I doubt it).

I am not a developer, though I was one in a previous life/decade (but not a java one – boy it’s a different world new!), and while challenging both techically and as a paradigm shift, I found the tool useable.

regards
John (P.S. I met you before this all started at the Wellington conference in Aug 06 Elisabeth).

]]> By: Jerrad Anderson http://testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/#comment-239 Jerrad Anderson Wed, 28 Feb 2007 14:40:27 +0000 http://www.testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/#comment-239 I started a long comment that ended up being to long for a comment so I posted it here: http://www.jerradanderson.com/blog/index.php?/archives/61-Functional-test-tools-direction.html#extended I started a long comment that ended up being to long for a comment so I posted it here:

http://www.jerradanderson.com/blog/index.php?/archives/61-Functional-test-tools-direction.html#extended

]]> By: Patrick Lightbody http://testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/#comment-238 Patrick Lightbody Wed, 28 Feb 2007 04:25:49 +0000 http://www.testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/#comment-238 I really enjoyed your "next gen" posts, as well as this. As the founder of OpenQA (home of Watir and Selenium), and a develop on Selenium, I have a lot of thoughts on this. In fact, I even created a product called HostedQA that tries to address these very issues (the company/product was recently sold and will be announced soon). I'm not sure why I didn't have you in my RSS reader sooner, but you're there now :) I've written up an initial response to some of your thoughts here in my blog entry, with the plan to follow up with two more posts in the coming days on my take on how next gen tools (like HostedQA, I hope) can help. You can find the entry here: http://blogs.opensymphony.com/plightbo/2007/02/next_gen_testing_tools.html I really enjoyed your “next gen” posts, as well as this. As the founder of OpenQA (home of Watir and Selenium), and a develop on Selenium, I have a lot of thoughts on this. In fact, I even created a product called HostedQA that tries to address these very issues (the company/product was recently sold and will be announced soon). I’m not sure why I didn’t have you in my RSS reader sooner, but you’re there now :)

I’ve written up an initial response to some of your thoughts here in my blog entry, with the plan to follow up with two more posts in the coming days on my take on how next gen tools (like HostedQA, I hope) can help. You can find the entry here: http://blogs.opensymphony.com/plightbo/2007/02/next_gen_testing_tools.html

]]> By: Paul Rogers http://testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/#comment-237 Paul Rogers Thu, 22 Feb 2007 21:18:13 +0000 http://www.testobsessed.com/2007/02/22/what-problem-would-next-generation-functional-testing-tools-solve/#comment-237 I completely agree with this, and also your earlier posts on the next generation of test tools. In many places that Ive done automation, Ive been asked "So what does your automation test?" My usual answer of "check the code" isnt usually the answer people want. When there are subtle differences, this becomes more important ( eg, 2 links to get to a help page, does the automation need to test both, how do we easily tell people which link we use) Ive attempted to do this by tying the watir tests to an app like test director. It required alot of work, and still really didnt acheive what we wanted. I completely agree with this, and also your earlier posts on the next generation of test tools.

In many places that Ive done automation, Ive been asked “So what does your automation test?” My usual answer of “check the code” isnt usually the answer people want.
When there are subtle differences, this becomes more important ( eg, 2 links to get to a help page, does the automation need to test both, how do we easily tell people which link we use)

Ive attempted to do this by tying the watir tests to an app like test director. It required alot of work, and still really didnt acheive what we wanted.

]]>