Branch: refs/heads/master
Home: https://github.com/Starlink/starlink
Commit: 09153f5e10c6071238ca9a57fdf286fdae44b7fc
https://github.com/Starlink/starlink/commit/09153f5e10c6071238ca9a57fdf286fdae44b7fc
Author: Daniel Berke <[log in to unmask]>
Date: 2013-07-26 (Fri, 26 Jul 2013)
Changed paths:
M applications/smurf/data/csofit2.dat
Log Message:
-----------
smurf: Update csofit2.dat to fix missing CSO fits.
Several more instances where CSO fits did not extend for the entire
observing period have been found and fixed.
Commit: 642d9b9801b6cb816184d85a8012fbb65a8f8947
https://github.com/Starlink/starlink/commit/642d9b9801b6cb816184d85a8012fbb65a8f8947
Author: Tim Jenness <[log in to unmask]>
Date: 2013-07-28 (Sun, 28 Jul 2013)
Changed paths:
M applications/smurf/libsmf/smf_calc_wvm.c
Log Message:
-----------
smurf: Ignore errors if the cache free fails
There are reports of CALCQU triggering errors during the cleanup
of the cache. I am guessing that CALCQU does not attempt to fix
up header metadata so does not initialise an early AST object in
the main thread. Later on the WVM data are calculated in threads
such that when the object is freed in the main thread at task
exit the object is invalid even though the pointer is defined.
The error message is:
!! AST: Error at line 160 in file smf_calc_wvm.c.
! Invalid Object pointer given (value is 390951).
! This pointer has been annulled, or the associated Object deleted.
! Application exit status AST__OBJIN, Object invalid
which hints that my interpretation may not be the whole story.
Commit: 25c1b74db1f859e19da11a7968c8d60d6025b543
https://github.com/Starlink/starlink/commit/25c1b74db1f859e19da11a7968c8d60d6025b543
Author: Tim Jenness <[log in to unmask]>
Date: 2013-07-28 (Sun, 28 Jul 2013)
Changed paths:
M applications/smurf/libgsd/Makefile.am
M applications/smurf/libgsd/gsdac.h
A applications/smurf/libgsd/gsdac_getRealInstrumentName.c
M applications/smurf/libgsd/gsdac_putFits.c
Log Message:
-----------
smurf: Translate GSD frontend string to the real instrument name
It turns out that GSD frontend strings are not very consistent
with the worst offender being the switch of RXB3CU to RXB instead
of just using RXB3.
This mainly affects the B-band instruments and RxW now gets the
band appended to the instrument name to be consistent with ACSIS
usage.
A full table of instrument history is included in the routine
gsdac_getRealInstrumentName.c
Compare: https://github.com/Starlink/starlink/compare/76e5e5661e48...25c1b74db1f8
|