> -----Original Message-----
> From: Stephen McConnell [mailto:[EMAIL PROTECTED]
> Sent: Monday, December 22, 2003 3:51 AM
> Olaf Bergner wrote:
>
> >>-----Original Message-----
> >>From: Stephen McConnell [mailto:[EMAIL PROTECTED]
> >>Sent: Monday, December 22, 2003 2:10 AM
> >>
> >>
> >>>I was thinking likewise, and since my debugging output
> >>>
> >>>INFO    2003-12-19 18:39:33.990 [main    ] (): Resolved appliance:
> >>>[EMAIL PROTECTED] Type: $Proxy3
> >>>
> >>>strongly hints at my component being proxied I am the more
> >>>
> >>>
> >>confused by the
> >>
> >>
> >>>problem at hand. Well, seems like I have to dig a little deeper.
> >>>
> >>>Anyways: thanks for your help,
> >>>
> >>>
> >>>
> >>Olaf:
> >>
> >>Just a quick followup - did you manage to resolve the problem?
> >>
> >>
> >
> >Yes, I did. But don't ask me how! I spent quite a few hours going through
> >the config and code and comparing it to the code/config in the hello
> >tutorial. Found no differences. I stuffed hello.block into my custom made
> >embedded kernel, and it worked, i.e. the returned component was
> castable to
> >Hello. Then I took the test case in the merlin-unit package and
> adapted it
> >to run my sample.block. Failure: component not castable to Sample (I
> >inserted a check).
> >
> >I ran the whole lot through a debugger and found out that a call to
> >Model.getInterfaces() (I think in DefaultAppliance) returned an
> empty array.
> >But why? Well, I basically gave up and decided to postpone the
> issue and go
> >on refactoring. After refactoring - Sample/SampleComponent got a
> new home -
> >I ran the same test: it worked, the component could be cast to Sample! I
> >think it was some kind of glitch in the space-time-continuum. I'll check
> >back on the issue, maybe in the end I'll solve this riddle ...
> >
>
> Ok  .. it didn't make sense to me at the time either but I know that the
> sands have been moving under our feet (at least on the bleeding edge).
> Wait a week and we'll have a reference point.
>
> Cheers, Steve,

No, I am pretty sure that's just some stupid mistake of mine. Right now, I
don't have any problems running blocks in my embedded kernel. Must have
overlooked something really obvious.

Cheers,

Olaf


---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to