Re: typed arrays

["M. Edward (Ed) Borasky" <znmeb@...>]

On Fri, Sep 28, 2012 at 2:08 PM, Tom <tmbdev@gmail.com> wrote:
>> that only works for byte blobs.  ... anything further has to have some
>> semantic agreements to work
>
>
> No, it works for arbitrary multidimensional arrays.  Semantic agreement for
> how to represent multidimensional numerical arrays has existed for over half
> a century: a 1D typed arrays together with a concrete representation of
> collections of bounds, offsets, and strides.  BLAS, PIL, LINPACK, NumPy,
> OpenCV, ImageMagick, and thousands of other libraries and tools use this
> representation. And because they all use the same representation in memory,
> they can all easily talk to each other.

Well ... assuming one has endian-ness and integer / float / double
sorted out and IEEE 754 mandated, yes. It took much of that half
century for "Computational Darwinism" to eliminate word addressing,
decimal arithmetic, 36 and 48 and 60 bit floating point words, ones
complement, sign-magnitude and various bizarre floating point schemes.
;-)

-- 
Twitter: http://twitter.com/znmeb; Computational Journalism Publishers
Workbench: http://j.mp/QCsXOr

How the Hell can the lion sleep with all those people singing "A weem
oh way!" at the top of their lungs?