Boost GIL


Todo List
Member boost::gil::detail::pixel_to_tuple (Pixel const &p, boost::mp11::index_sequence< I... >) -> decltype(std::make_tuple(p[I]...))
With C++14 and using auto we don't need the decltype anymore
Member boost::gil::detail::tuple_to_tuple (Tuple const &t, boost::mp11::index_sequence< I... >) -> decltype(std::make_tuple(std::get< I >(t)...))
With C++14 and using auto we don't need the decltype anymore
Member boost::gil::otsu
TODO
Member boost::gil::threshold
TODO
Member boost::gil::zero
TODO
Class default_color_converter_impl< gray_t, cmyk_t >
FIXME: Where does this calculation come from? Shouldn't gray be inverted? Currently, white becomes black and black becomes white.
Class dynamic_xy_step_type< View >
TODO: Is Locator allowed or practical to occur?
Class file_stream_device< FormatTag >
switch to a sane interface as soon as there is something good in boost. I.E. the IOChains library would fit very well here.
Member file_stream_device< FormatTag >::read (byte_t *data, std::size_t count) -> std::size_t

: change byte_t* to void*

: add compiler symbol to turn error checking on and off.

Class is_read_supported< Pixel, FormatTag >
the name is_supported is to generic, pick something more IO realted.
Member virtual_2d_locator< DerefFn, IsTransposed >::is_1d_traversable (x_coord_t) const
TODO: is there no gap at the end of each row? i.e. can we use x_iterator to visit every pixel instead of nested loops?