> -----Original Message-----
> From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED]]On
> Behalf Of Geir Magnusson Jr.
>
> Paulo Gaspar wrote:
>
> [SNIP]
>
> > But it does not work because of that wrapper object. I only went
> > as far as to find out that it was some private hard-to-identify
> > object - I could not see its members in the debugger- but I
> > still did not trace the problem to its foundation. That is what
> > I want to do now.
>
> I know the problem - it was a decision that I had to make a while ago,
> and did to ensure backwards compatibility. As time has shown, it's more
> important to go the other way. We've been through this before.
Again, give me some pointers to the source code please. I would like to
bettern understand this.
> Later today, I'll post a summary and change the code in HEAD, so you can
> give it a whirl.
Thanks man.
> > BTW, I tried to see the wrapper members because I was trying to
> > find how to get the wrapped object. Since the debugger could not
> > even get the class from the object, I kept not knowing where to
> > start. Is it some anonymous inner class or something like that?
> >
> > Now, I think that
> > 1. The #if() directive should resolve a reference that returns
> > null to as being false;
> > 2. Macros should unwrap arguments before passing them to methods
> > called inside themselves.
> >
> > While 2 looks clear to me, but maybe you think different about 1.
> > Anyway, I think 1 is the most intuitive way to go.
> >
>
> The issue with 1) is that #if() should behave correctly in cases like
> this. 2), no, because no one should know.
Sorry, but
- I do not understand if you agree or not with 1).
- I do not understand at all what you mean about 2).
Could you just elaborate a little bit?
Thanks,
Paulo
