> 100% anal C++ issues. If you look, the largest part is just
> changing 'i'->'j' or adding the class reference (eg: 'GenericChord::') in front
> of the various members so gcc47 will quit complaining. With the pickiness
> of gcc47, it makes you wonder "how the hell did this stuff even run before?" I
> guess it always knew which 'i' it was supposed to use, but now gcc makes you tell
> it explicitly.
> 
> Absolutely 'zarro' substantive change to the code. Where
> loops used to run:
> 
> for(this::i=x;i<something;i++) {
> 
>   for(i=y;i<something_else;i++) {
>     stuff
>   }
> 
> }
> 
>   The patched loops run like:
> 
> for(this::i=x;i<something;i++) {
> 
>   for(j=y;j<something_else;j++) {
>     stuff
>   }
> 
> }
> 
> The other changes were just to #include <unistd.h> where needed... Should be
> good to push.

Over my head, but to be expected when I have barely dented my C++ text book. :(

I just rebuilt rosegarden with your patch (gcc 4.4.4). The package starts without noticeable problems. I'll push the patch to GIT, although as I mentioned, I have no idea what or how to test.

We should have a basic usability test plan for these patches....

With that said, I touched the digikam patch with some preprocessor checks like I did in the gwenview patch. I don't know whether that will make a difference for you with libpng 1.5. I built digikam here with the updated patch against libpng 1.4.9. Basic usability testing indicates the package built and runs. Yet again, I don't know specifically what the patch affects and I can't test that aspect for usability.

Please test the updated digikam patch:

http://humanreadable.nfshost.com/trinity/patches/digikam/digikam-libpng15.diff

I'll be glad if the patch works but I won't be heart broken if the patch fails. Much like you, I'm a C++ blind guy walking in the dark....

Darrell