[CF-metadata] Add new integer types to CF?

From: Manning, Evan M <Evan.M.Manning>
Date: Tue, 26 Sep 2017 22:15:00 +0000

Sounds Good.

_Unsigned should also show up in table A.1

  -- Evan

On 9/26/17, 1:39 PM, "Charlie Zender" <zender at uci.edu> wrote:

    If there is strong opposition to replacing the valid_* method in
    CF <= 1.7 with the _Unsigned method in CF 1.8 then please speak-up.
    People seem to prefer _Unsigned (as described in NUG) to the
    previous CF standard method (valid_*) as way to allow CDF1, CDF2,
    and netCDF4-classic to indicate unsigned integers in CF 1.8.
    New datasets could still use the valid_* method but that would
    restrict their CF-compliance to <= 1.7. CDF1 and CDF2 and
    netCDF4-classic datasets intended for CF 1.8 compliance would have
    to use _Unsigned instead of valid_*.
    
    This plan could be enacted with wording like this:
    
    "The netCDF data types char, byte, unsigned byte, short, unsigned
    short, int, unsigned int, int64, unsigned int64, float or real,
    and double are all acceptable. The char type is not intended for
    numeric data. One byte numeric data should be stored using the byte
    or unsigned byte data type. Unsigned types should be used for unsigned
    data if possible. If the underlying file format does not support
    unsigned types, integer or byte variables that are to be interpreted
    as unsigned must have the attribute _Unsigned = "true". The convention
    explicitly distinguishes between signed and unsigned integer types
    only where necessary. Unless otherwise noted, int is interchangeable
    with unsigned int, int64, and unsigned int64 in this convention,
    including examples and appendices. Similarly short is interchangable
    with unsigned short, and byte with unsigned byte."
    
    Comments welcome...
    --
    Charlie Zender, Earth System Sci. & Computer Sci.
    University of California, Irvine 949-891-2429 )'(
    
    
Received on Tue Sep 26 2017 - 16:15:00 BST