Comments on: How Much Detail Should I Write In My Test Case? http://mattarcherblog.wordpress.com/2008/07/24/how-much-detail-should-i-write-in-my-test-case/ Mon, 24 Aug 2009 11:30:51 +0000 http://wordpress.com/ hourly 1 By: Matt Archer http://mattarcherblog.wordpress.com/2008/07/24/how-much-detail-should-i-write-in-my-test-case/#comment-29 Matt Archer Thu, 28 Aug 2008 09:51:35 +0000 http://mattarcherblog.wordpress.com/?p=100#comment-29 Hi Bruce, Liz, thanks for you comments, I know what you mean about GUI level test automations. There are a lot of barriers to adoption and some good reason why you just wouldn't consider it in the first place. I was reading Brian Marick's blog the other day about <a href="" rel="nofollow">Barriers to acceptance-test driven design</A> which I think highlight (even with a more agile spin) that GUI (or just below GUI) level automation isn't something to rush into. Hi Bruce, Liz, thanks for you comments,

I know what you mean about GUI level test automations. There are a lot of barriers to adoption and some good reason why you just wouldn’t consider it in the first place.

I was reading Brian Marick’s blog the other day about Barriers to acceptance-test driven design which I think highlight (even with a more agile spin) that GUI (or just below GUI) level automation isn’t something to rush into.

]]> By: Bruce http://mattarcherblog.wordpress.com/2008/07/24/how-much-detail-should-i-write-in-my-test-case/#comment-28 Bruce Wed, 27 Aug 2008 19:54:48 +0000 http://mattarcherblog.wordpress.com/?p=100#comment-28 Hi Matt, This is a great analysis. Thanks for posting it. I also agree with Liz that automated testing is still very much a holy grail. Automation for unit testing works well and can be achieved with relative low cost. In fact it is a good idea to write the unit tests before the actual code. Automated testing of user interfaces is slowly getting there. There are a number of tools available to test both native GUIs and as well as web applications. Unfortunately these take almost as much effort to program as the application itself. Changes to the application can invalidate large blocks of automated test scripts very quickly. The real problems start when you need to test across multiple integrated systems to validate the end-to-end business processes. Very few tools even allow for testing across multiple UIs and APIs so you end up having to write an interface for the test automation tool as well. Hi Matt,

This is a great analysis. Thanks for posting it.

I also agree with Liz that automated testing is still very much a holy grail.

Automation for unit testing works well and can be achieved with relative low cost. In fact it is a good idea to write the unit tests before the actual code.

Automated testing of user interfaces is slowly getting there. There are a number of tools available to test both native GUIs and as well as web applications. Unfortunately these take almost as much effort to program as the application itself. Changes to the application can invalidate large blocks of automated test scripts very quickly.

The real problems start when you need to test across multiple integrated systems to validate the end-to-end business processes. Very few tools even allow for testing across multiple UIs and APIs so you end up having to write an interface for the test automation tool as well.

]]> By: Simple Documentation for Software Testing « Matt Archer’s Blog http://mattarcherblog.wordpress.com/2008/07/24/how-much-detail-should-i-write-in-my-test-case/#comment-20 Simple Documentation for Software Testing « Matt Archer’s Blog Fri, 08 Aug 2008 12:51:28 +0000 http://mattarcherblog.wordpress.com/?p=100#comment-20 [...] How Much Detail Should I Write In My Test Case? [...] [...] How Much Detail Should I Write In My Test Case? [...]

]]> By: LizM http://mattarcherblog.wordpress.com/2008/07/24/how-much-detail-should-i-write-in-my-test-case/#comment-19 LizM Sun, 03 Aug 2008 18:43:23 +0000 http://mattarcherblog.wordpress.com/?p=100#comment-19 Hullo! Wise words :-) The 'snake oil' link is particularly useful as someone was recently here talking up automated testing - a case of reckless assumptions I fear :-) Automation is a fine idea and absolutely the right way to go - if only it worked! Well, automation can work, but I've got a feeling that it's of limited benefit when a system is subject to a large amount of change (i.e. early in development). Post development when you move to lower change instability then automated testing is valuable for regression test efficiency. We'll be reading about the snake oil with great interest here ;o) Liz Hullo!
Wise words :-) The ’snake oil’ link is particularly useful as someone was recently here talking up automated testing – a case of reckless assumptions I fear :-)
Automation is a fine idea and absolutely the right way to go – if only it worked! Well, automation can work, but I’ve got a feeling that it’s of limited benefit when a system is subject to a large amount of change (i.e. early in development). Post development when you move to lower change instability then automated testing is valuable for regression test efficiency.
We’ll be reading about the snake oil with great interest here ;o)
Liz

]]>