waft is a good complement to Cactus & HTTUnit. It extends HttpUnit and gives it a more natural interface and claims to support javascript (!!).
One more comment, though, it's been my experience that testing web sites that aren't designed to be tested is a real nightmare; start testing as soon as possible. You may even need to clean up your HTML code to make httpunit accept it; mozilla and IE will accept practically anything, test tools won't.
One more thing, never let tests fall behind or you'll loose them. That's especially true when testing web sites. Run them as often as possible. (see http://cruisecontrol.sourceforge.net ?)
The gesture recognition system lionhead uses can be quite simple, since it only have to recognize at most a dozen gestures or so.
In a photoshop-like, that would be more like a hundred. That means more arcane gestures and more mistakes.
It'd be a nightmare if it's not implementend right...
I think it's exactly the reason why you don't use voice-recognition system to use your system: too error-prone or too complicated.
IBM tried integrating voice-recognition in OS/2 years ago. ever heard about OS/2 ?
Slashdot Mirror
waft is a good complement to Cactus & HTTUnit. It extends HttpUnit and gives it a more natural interface and claims to support javascript (!!).
One more comment, though, it's been my experience that testing web sites that aren't designed to be tested is a real nightmare; start testing as soon as possible. You may even need to clean up your HTML code to make httpunit accept it; mozilla and IE will accept practically anything, test tools won't.
One more thing, never let tests fall behind or you'll loose them. That's especially true when testing web sites. Run them as often as possible. (see http://cruisecontrol.sourceforge.net ?)
The gesture recognition system lionhead uses can be quite simple, since it only have to recognize at most a dozen gestures or so.
In a photoshop-like, that would be more like a hundred. That means more arcane gestures and more mistakes.
It'd be a nightmare if it's not implementend right...
I think it's exactly the reason why you don't use voice-recognition system to use your system: too error-prone or too complicated.
IBM tried integrating voice-recognition in OS/2 years ago. ever heard about OS/2 ?