These are images taken on La Silla's danish telescope and uploaded by them.
(originally to /home/silla/upload).

While they were uploading, bin/fetchdata && gavoimp -u danish/red ran in a
nightly cron job, where fetchdata would copy the reduced files from the upload
directory to the RD's input directory (if such a thing were to run again,
you'd have to make sure bin/fixupHeaders.py is called before gavoimp).

In 2009, a set of DVDs came in; its content is in cds-2009-09.  To move these
files, we ran bin/fetchdata.py cds-2009-09.

Later, it turned out that the data on the DVDs was incomplete, so Christine
Liebig downloaded more data from a danish server.  That data is now in
ftp_2009-10.

In 2010, Janine Fohlmeister obtained data from some online the web page
http://comp.astro.ku.dk/Twiki/view/MiNDTSTEp (Data area->objects).  This
input is kept in mindstep2010 and was copied into the main data tree using
bin/fetchdata.py mindstep2010/.

Some or the purportedly reduced data files for 2010 data are in fact
flatfields or something similar; this includes (in the notation generated
by fetchdata) HE_0047-1756_20100710T31838, HE_0047-1756_20100710T31838.fits,
HE_0047-1756_20100716T29545.fits, HE_0047-1756_20100716T29877.fits

The files coming in from la silla are botched (invalid headers, weird date-obs
format, etc).   Thus, you need to run bin/fixupHeaders.py when new images are
in data.  This is idempotent, i.e., fixed files will not be touched by it.

The silla directory in here is the upload user's home.  It contains raw data
in addition to the reduced data that's in the DB.  It can probably be deleted
when all is done, but that should be discussed with Robert Schmidt.  The same
goes for the cd* and ftp* input directories.



Removed data
============

An image called Q2237+0305_20090907T23571.fits was removed.  It evidently
contained a solar system image.