gzipped images in SIAP 1.0

[Rob Seaman]

> RFC4047 says:
>
>    A FITS file described with the media type "image/fits" SHOULD be
>    principally intended to communicate the single data array in the
>    PHDU.

"Should be principally intended"?  This is completely unconstraining  
- subjunctive conditional good intentions.  I'll have to remember  
that for my next performance review...

Mark Taylor wrote:

> as I understand it tile-compressed FITS images are stored in a  
> BINTABLE
> extension (hence not the PHDU), so they should properly be described
> as application/fits and not image/fits.

The point I was trying to make was that a tile compressed FITS files  
remains an image.  It is represented (that is, "encoded") as a  
bintable.  Any FITS can be application/fits.  An image/fits says "I  
represent a single image".  A content encoding could then be applied  
to that to turn the stream of bytes into anything else, including  
(potentially) a bintable.

Rob