I'm beginning to look at testing the framework itself 
using a Mock Objects approach.

Could you (Christian) expand on the risks and benefits 
of this kind of change.  I'm not familiar with HttpUnit; 
perhaps you could explain its overall premise.

The risk is that, in some cases, client-side scripting 
may be reliant on explicitly setting the id attribute of 
particular elements.

--
[EMAIL PROTECTED]

http://tapestry.sf.net
> Chrisitan Noack wrote:
> > Al lot of HTML elements can be given an id. You can use this id very well
> > to test your application with HttpUnit. Some of the ids are already set by
> > tapestry automatically and some are not. Wouldn't it be good to have all
> > comonent ids automatically set as HTML ids wherever that is possible?
> 
> Are you really sure you want to rely on automatically generated ids in 
> your HTTPUnit test cases? What if your HTML changes? Do the generated 
> ids change as well? In that case you can dump your testcase and redo it, 
> which is something I would't really like to do on a frequent basis.
> 
> OTOH, having some automated test support would be nice. I currently 
> don't do HTTPUnit testing because of the very same dilemma (in WebObjects).
> 
> Just my 0.02 EUR,
> 
> -dirk
> 
> 
> 
> -------------------------------------------------------
> This sf.net email is sponsored by:ThinkGeek
> Gadgets, caffeine, t-shirts, fun stuff.
> http://thinkgeek.com/sf
> _______________________________________________
> Tapestry-developer mailing list
> [EMAIL PROTECTED]
> https://lists.sourceforge.net/lists/listinfo/tapestry-developer


-------------------------------------------------------
This sf.net email is sponsored by:ThinkGeek
Gadgets, caffeine, t-shirts, fun stuff.
http://thinkgeek.com/sf
_______________________________________________
Tapestry-developer mailing list
[EMAIL PROTECTED]
https://lists.sourceforge.net/lists/listinfo/tapestry-developer

Reply via email to