[fitsbits] CDELTn

[Phil Hodge]

Francois,

> I just found a curiosity: FITS files having CDELTn axis set to zero.
> This may look strange, but when the corresponding NAXISn is 1, this 
> should be harmless. Is the value zero actually forbidden in the 
> FITS definition ? (I did not find this restriction in the FITS definition).


Regardless of the value of NAXISn, computing the coordinate value at a 
pixel would involve dividing by CDELTn.  For example:  CRVALn + (1 - 
CRPIXn) / CDELTn.  So it's not as harmless as it may seem.

Phil