[fitsbits] 64-bit integer comments

[Mark Calabretta]

On Thu 2005/05/12 10:55:58 +0200, Preben Grosbol wrote
in a message to: fitsbits at donar.cv.nrao.edu

>I still don't see a convincing argument for adding BITPIX=64.  That it

The argument for maintaining consistency with 'K' TFORMs and Q array
descriptors seems, if not compelling, then at least persuasive.  Another
point is that FITS standard always seems to play catchup with what
people have done in practice, perhaps it should try to be proactive.

While we're counting bits, I note that the precision of keyvalue data
types is essentially undefined.  Specifically, I believe that most
generic header parsers that encounter

PSRTIME =       88402115102000 / [us] Time since T0
PULSENO =           9465023873 / Pulsar pulse number since T0

will pack them into 32-bit ints and lose essential precision.  The FITS
standard is silent on this issue, perhaps it should address it also. 

Mark Calabretta
ATNF