5 messages in com.canoo.lists.webtestRe: [Webtest] Dynamic field naming| From | Sent On | Attachments |
|---|---|---|
| Erik...@alltel.com | 01 May 2002 11:56 | |
| Dierk König | 02 May 2002 02:46 | |
| Ben Cox | 03 May 2002 15:04 | |
| Dierk König | 06 May 2002 00:56 | |
| testing Auto | 06 May 2002 07:55 |
| Subject: | Re: [Webtest] Dynamic field naming![]() |
|---|---|
| From: | testing Auto (test...@yahoo.com) |
| Date: | 05/06/2002 07:55:33 AM |
| List: | com.canoo.lists.webtest |
Ben, I'm sorry I haven't followed this thread very closely so if you've already answered these questions I apologize.
What is the exact failing or failings that you have with WebTest? What features of HttpUnit do you wish to have access to? Be specific, don't just say all.
I'm torn between the two sides. On the one side I think that WebTest is limiting (See posting "Extending WebTest..."). And on the other its lot of work to start from the bottom and start over. (Dierk how much time has been committed to getting WebTest to this point?)
I'm not sure that its feasible to make something that meets everyones needs for an acceptance test tool. I feel that WebTest has a basic framework and that for the more specific acceptance test challenges it will have to be customized. Of course I hear what you are saying in that the customer wants to pay for functionality in their product, not in the testing framework. But sometimes it is worth the price in long run for such an effort.
Already after investing a week or so in customizing WebTest it is starting to pay off. We've moved from 0 automated web tests to 32,000+ and found 4 new bugs.
Although finding 4 new bugs doesn't sound like much, its verifies that the unit/functional testing done by the developers has paid off. And from the Customer's perspective they have now verified for themselves that what they wanted is there and acceptable. And they will be able to repeat these tests every time the developers give them a new build.
Additionally, the customization was made with a domain knowledgeable Customer in mind. In other words he/she will be able to quickly create additional tests in XML using terms that they are familiar with.
With another one week investment we could eliminate the need for working in the XML file as well, creating a Web Gui. (This hasn't been done yet. We don't do XP, but the Customer does drive what gets done. They want to work w/ what has been created and if they feel that the XML is to hard to get through then we will move on to the GUI component. I'm betting they'll want it though!)
Chris
__________________________________________________ Do You Yahoo!? Yahoo! Health - your guide to health and wellness http://health.yahoo.com




