Re: Bug in scenario

2007-09-24 Thread stanlick
That is correct! On 9/24/07, Dave Newton <[EMAIL PROTECTED]> wrote: > > I think the OP's issue is that there's a difference > between XWork object creation (it will create an > Action property if configured to) and ModelDriven > (won't). > > d. > > --- Ian Roughley <[EMAIL PROTECTED]> wrote: > > >

Re: Bug in scenario

2007-09-24 Thread Dave Newton
I think the OP's issue is that there's a difference between XWork object creation (it will create an Action property if configured to) and ModelDriven (won't). d. --- Ian Roughley <[EMAIL PROTECTED]> wrote: > This is the fundamental way the value stack works - > if it can't evaluate > an expres

Re: Bug in scenario

2007-09-24 Thread Ian Roughley
This is the fundamental way the value stack works - if it can't evaluate an expression at the topmost level, it starts to work it's way down until the expression can be evaluated. I guess my question really is "why are you trying to set value on a null object?" It sounds to be like you need t

Re: Bug in scenario

2007-09-24 Thread Dave Newton
--- [EMAIL PROTECTED] wrote: > I'm suggesting the S2 bean creation happens > if (model == null) in the same fashion it does > ordinarily. Otherwise, perhaps an exception should > be thrown? I'd suggest filing a JIRA enhancement request; seems like a reasonable idea. d. ---

Re: Bug in scenario

2007-09-24 Thread stanlick
I know a few developers who are fundamentally broken. I'm suggesting the S2 bean creation happens if(model==null) in the same fashion it does ordinarily. Otherwise, perhaps an exception should be thrown? What I witnessed was voodoo logic. Scott On 9/23/07, Dave Newton <[EMAIL PROTECTED]> wrote

Re: Bug in scenario

2007-09-23 Thread Dave Newton
--- [EMAIL PROTECTED] wrote: > I guess I'm wondering if the framework should have a > safety net to fall back on when an object is *not* > produced by user code. What would that look like; an extension to ModelDriven that included a createModel function, or...? If something implements ModelDri

Re: Bug in scenario

2007-09-23 Thread stanlick
When prepare/getModel misbehave! I guess I'm wondering if the framework should have a safety net to fall back on when an object is *not* produced by user code. On 9/23/07, Laurie Harper <[EMAIL PROTECTED]> wrote: > > [EMAIL PROTECTED] wrote: > > If your action class implements ModelDriven and Pr

Re: Bug in scenario

2007-09-23 Thread Laurie Harper
[EMAIL PROTECTED] wrote: If your action class implements ModelDriven and Preparable and your prepare method returns a null object, ParameterInterceptor tries to set the parms which are names appropriate to properties in model object that is expected to be on the top of stack. However, since mode