Hi Yannick,
About a year ago the CADC switched from sending .fz files by default
to sending .gz files. This was done so that ALADIN and ds9 could
directly load images from the archive URL with no need to pass the
file through the local filesystem. This change has greatly improved
our interaction with these VO compliant tools and made the collection
more accessible, in general.
At the same time, we appreciated that many people want to continue to
use the .fz format, essentially anyone who keeps many files on local
disk and wants to see the headers without having decompress the file.
So, we continue to offer the files in .fz format but this is done in
what we call 'Archive Specific Format' for the download. To get the
asf version of the file you should include the asf=true tag in the
URL, for example:
http://www.cadc.hia.nrc.gc.ca/getData?file_id=883362p&archive=CFHT&asf=true
Since you are retrieving many 100s of files I surprised if only a one
of them is in .gz format, but perhaps there was a syntax error in the
construction of the URL for just this one file?
Have a look at the URL below for a bit more detail on the use of the
CADC proxy delivery system.
http://www.cadc-ccda.hia-iha.nrc-cnrc.gc.ca/authProxy
Hope this clears up your issue,
JJ
On 19-Apr-08, at 12:46 AM, Yannick Mellier wrote:
>
>
> Dear JJ
>
> for an unknown reason, it is impossible to get 883362p.fits.fz using
> curl (W2 image).
> It works for all other images, but this one. I then went to the
> CADC interface and saw this image if not .fz but .gz compressed. Curl
> did not work neither using this filename. I was obliged to get it
> "by hand" using wget and CADC interface for that single image.
>
> Why this image is gzipped compressed?
>
> Cheers
>
> yannick
>
>
Received on Mon Apr 21 2008 - 06:43:53 HST