reshape can only rearrange the raw linearly ordered data into Julia’s default column-major order (or colexicographical order, for higher-dimensional arrays). Would it be worth adding an optional keyword argument that reshapes the data in row major (or generally lexicographical) order? This would be convenient e.g. for interfacing with data generated by or for other programming languages that use the row-major convention (e.g. here). It’s fairly easy to reshape an array into an n \times m row-major matrix by reshaping it into an m \times n column-major matrix and then taking the transpose, but this gets cumbersome for higher-dimensional arrays.
(Note that I’m only suggesting changing the visible behavior of
reshape, not the way that the resulting array is stored internally.)