[Insight-developers] itkBasicArchitectureTest won't compile under gcc/linux

[Paul Hughett]

As of this morning's update, itkBasicArchitectureTest would not compile
successfully under gcc 2.95.2 on linux:

c++  -g -O2   -I/home/hughett/work/Insight -I/home/hughett/work/Insight/Code/Insight3DParty/vxl -I/home/hughett/work/Insight/Code/Common -I/home/hughett/work/Insight/Code/Insight3DParty/vxl    -c itkBasicArchitectureTest.cxx -o itkBasicArchitectureTest.o
In file included from /home/hughett/work/Insight/Code/Common/itkShrinkImage.h:130,
                 from itkBasicArchitectureTest.cxx:21:
/home/hughett/work/Insight/Code/Common/itkShrinkImage.txx: In method `void itk::ShrinkImage<TInputImage,TOutputImage>::ThreadedGenerateData(typename itk::ShrinkImage<TInputImage,TOutputImage>::OutputImageRegion &, int)':
/home/hughett/work/Insight/Code/Common/itkShrinkImage.txx:90: name lookup of `i' changed for new ANSI `for' scoping
/home/hughett/work/Insight/Code/Common/itkShrinkImage.txx:76:   using obsolete binding at `i'
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx: In function `void itk::WriteDataArray<itk::Image<itk::Scalar<float>,2,itk::ValarrayImageContainer<long unsigned int,itk::Scalar<float> > > >(ofstream &, itk::Image<itk::Scalar<float>,2,itk::ValarrayImageContainer<long unsigned int,itk::Scalar<float> > > *, int, int)':
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx:126:   instantiated from `itk::WriteRawImage<itk::Image<itk::Scalar<float>,2,itk::ValarrayImageContainer<long unsigned int,itk::Scalar<float> > > >::WriteData()'
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.h:49:   instantiated from here
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx:60: warning: choosing `itk::Scalar<float>::operator float &()' over `itk::Scalar<float>::operator float() const'
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx:60: warning:   for conversion from `itk::Scalar<float>' to `float'
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx:60: warning:   because conversion sequence for the argument is better
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx:126:   instantiated from `itk::WriteRawImage<itk::Image<itk::Scalar<float>,2,itk::ValarrayImageContainer<long unsigned int,itk::Scalar<float> > > >::WriteData()'
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.h:49:   instantiated from here
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx:69: warning: choosing `itk::Scalar<float>::operator float &()' over `itk::Scalar<float>::operator float() const'
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx:69: warning:   for conversion from `itk::Scalar<float>' to `float'
/home/hughett/work/Insight/Code/Common/itkWriteRawImage.txx:69: warning:   because conversion sequence for the argument is better
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx: In function `void itk::WriteDataArray<main()::FloatImage2DType>(ostream *, main()::FloatImage2DType *, int)':
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx:313:   instantiated from `itk::WriteVTKImage<itk::Image<itk::Scalar<float>,2,itk::ValarrayImageContainer<long unsigned int,itk::Scalar<float> > > >::WriteVTKImageData(ostream *, main()::FloatImage2DType *)'
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx:105:   instantiated from `itk::WriteVTKImage<itk::Image<itk::Scalar<float>,2,itk::ValarrayImageContainer<long unsigned int,itk::Scalar<float> > > >::WriteData()'
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.h:49:   instantiated from here
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx:228: warning: choosing `itk::Scalar<float>::operator float &()' over `itk::Scalar<float>::operator float() const'
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx:228: warning:   for conversion from `itk::Scalar<float>' to `float'
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx:228: warning:   because conversion sequence for the argument is better
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx:236: warning: choosing `itk::Scalar<float>::operator float &()' over `itk::Scalar<float>::operator float() const'
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx:236: warning:   for conversion from `itk::Scalar<float>' to `float'
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.txx:236: warning:   because conversion sequence for the argument is better
/home/hughett/work/Insight/Code/Common/itkShrinkImage.txx: In method `void itk::ShrinkImage<itk::Image<itk::Scalar<float>,2,itk::ValarrayImageContainer<long unsigned int,itk::Scalar<float> > >,itk::Image<itk::Scalar<float>,2,itk::ValarrayImageContainer<long unsigned int,itk::Scalar<float> > > >::ThreadedGenerateData(const itk::ImageRegion<2> &, int)':
/home/hughett/work/Insight/Code/Common/itkWriteVTKImage.h:49:   instantiated from here
/home/hughett/work/Insight/Code/Common/itkShrinkImage.txx:90: name lookup of `i' changed for new ANSI `for' scoping
/home/hughett/work/Insight/Code/Common/itkShrinkImage.txx:76:   using obsolete binding at `i'
m