compressed fits format

[Alain MAURY]

Since my original message was posted, I was able to read the document
recommended by Dr Schlesinger, and found it to be a very good basis to
start from. I have also tried the Hcompress algorithm, and with
a factor of compression of 10, it does not seem to loose much
as far as faint stars are concerned if you blink the original image
and the compressed-decompressed one. This is rather subjective, a
more detailed analysis will be done here. Still, sounds good. The
compression time may be a concern, but we should run this algorithm
on a Pentium and see what we get. It seems that 2 minutes of compression
time for a 2K image ( i.e. 8.5 megabytes ) is doable.
What is the procedure for some modifications to the FITS standard to be 
registered by the IAU FITS WG. Next general assembly ( next year ? ),
or a separate meeting ?
Finally, after possible "registration", would compressed fits files still
have the .fit ot .fts extension, or would a new extension like .ftc ( fits
compressed ) be used ?
Alain