The parameters are: observation number, group number, object name, observation type, UT start time, exposure time, number of coadds, read mode and speed, filter, start airmass, frame dimensions in pixels, base equatorial co-ordinates, and data-reduction recipe name.

• The $<$date$>$ comes from the internal header keyword ORAC_UTDATE.

• The logfile created by this primitive does not follow the standard ORAC-DR naming convention (log.xxxx) since it can be used to write log files to directories other than $ORAC_DATA_OUT and unique file names are required.

• Fudges missing or old headers.

• Uses user header ORAC_INSTRUMENT to specify the file name.

• Specification provided by Sandy Leggett.

• The text log file $ORAC_DATA_IN/$<$date$<$.nightlog, where $<$date$>$ is the UT date, unless the OUT argument is set, whereupon the log is in $ORAC_DATA_OUT. This enables a separate on-the-fly log. For the multi-mode instruments UIST, Michelle, IRIS2, and ISAAC the file is $ORAC_DATA_IN/$<$date$>$_im.nightlog. For LCOGT data, we make use of the DAY-OBS header keyword to use as the observation date (and hence log root) to prevent writing two logs after we go over the UTC midnight boundary. The on-the-fly log in $ORAC_DATA_OUT is always appended to, being created only if it doesn’t exist. Thus multiple entries for the same observation may exist in the on-the-fly log if the pipeline is rerun. The "clean" log file in F$$<$ORAC_DATA_IN$>$ is re-started if the observation number equals 1 and is appended to otherwise, being created as necessary.