[Webtest] [dev] migration to htmlunit issues

[Marc Guillemot]

Hi,

>>what should be the behavior? How to have a deterministic behavior 
>>if 2 nested frames contain subframes with identical 
>>names (Httpunit javadoc doesn't say anything about this)?
> 
> Yes. I posted a poll about this issue.
> Nobody answered that he uses nested frames with same names.
> (actually that's two issues: a nests b and c. it may be that
> b.name==c.name or a.name==b.name)
> Since the behaviour is currently indeterministic in such a case,
> I added a note to the documentation about that.
> http://webtest.canoo.com/webtest/manual/syntax.html#followframe
> 
> I would prefer to keep this behaviour (i.e. reimplement the
> httpunit 1.6 behaviour for htmlunit) to avoid "going back".
> 
> It may be a good idea to log a warning (or even fail?)
> in case of doubled names.
> 
> As stated in the poll, I currently see no sensible use of nested frames
> with same names anyway...


=> in the case of doubled names I would just return one frame without saying why this one and could return the other one 
by the next run. Right?


>>I've explicitely added type="submit" to avoid problems.
> 
> Ok. Thanks for that. 
> I just recognized that
> http://www.w3.org/TR/REC-html40/interact/forms.html#h-17.5
> states that "type" is not implied.
> So omitting this attribute wouldn't be valid anyway.

As far as I understand it, omitting this attribute was valid.
Type has (or should have, cf IE) a default value as written in the attribute definitions:

"submit: Creates a submit button. This is the default value."

Marc.