NXUenterdata

[Freddie Akeroyd]

I think i'm basically in favour of an NXwritedata along the lines
suggested. One advantage, however, of splitting the "makedata" and
"opendata" was that if you expected an SDS to be there you only used
"opendata" and got an error if it wasn't; i wouldn't like the default
behaviour to be "create" if something is missing. A way round it
without having to use lots of function calls would be to add a "flags"
argument that could be NXACC_RDWR (modify an existing set) or
NXACC_CREATE (create a new set with the given name) - just like
"status=old" and "status=new" on a FORTRAN OPEN(). A question would
be whether "create" should return an error if a set already exists? If
so, we should probably then add a "status=unknown" flag.

Freddie

-- 
Freddie Akeroyd                        Email:  Freddie.Akeroyd at rl.ac.uk
ISIS Facility                          Tel:    +44 1235 445457
Rutherford Appleton Laboratory         Fax:    +44 1235 445720
Chilton, DIDCOT, OX11 OQX, GB          WWW:    http://www.isis.rl.ac.uk/