Bug list

Number 2.4-4 Added September 2001 nap_info command gave poor and confusing error messages
when used without a parameter. Fixed in version 2.5.4.

Number 2.4-3 Added September 2001 Known problems using bit operators on signed integers.
Bit shift operations could result in missing/null values being formed. Use only unsigned types in bit
operations and also make sure missing/null values are not set for the variable. Problem is current
in all versions of CAPS.

Number 2.4-2 Added September 2001 Occasional errors when large data arrays were
remapped using the warping. This problem was associated with an internal incorrectly calculated
coordinate variable step size causing rounding problems. Fixed in version 2.5.3+.

Number 2.4-1 Added September 2001 The get_atsr_gbt is known to have had several faults including:
1. Errors reading latitude and longitude values in night GBT files (all versions).
2. PC version read erroneous data because the BINARY flag was not set on file open resulting in data
values that represent DOS control characters upseting the read process.
Fixed in version 2.5.3+.

Number 2.1.2-1 Severity 1 Date June 15 2000 Windows version unable to read calibration files
Description Solar calibration fails because the ancillary files are unreadable. The problem has been identified as a corruption of the ancillary data files. Extra carriage returns have been generated in the ancillary files in the process of going from the CAPS CVS repository under linux to the Windows environment.

Solution Edit the files and remove the blank lines introduced. (aux_data\filtflux.tab etc)

Number 2.1.1-1 Severity 1 Date May 2000 Windows Version unable to read HRPT files
Description When attempting to read a Telonics format HRPT file get_HRPT_hdr reported an incorrect file format. This problem has been identfied as a Windows only problem associated with the need to read binary data with explicit read binary calls.

Solution Version 2.1.2 released for Windows.

Number 2.0.1-2 Severity 1 Date 2 Dec 1999 Solar radiance and reflectance calculation too high Status: Fixed V2.0.3
Description AVHRR channel 1 and 2 radiance and reflectance values for NOAA 14 from 1 February 1999 are calculated approximately 7% and 2% respectively too high. The algorithm for calculation the calibration responsivity as a function of time uses a reference time from launch date instead of from the calibration date. The error only effects NOAA 14 from 1 February because all other calibration times are from the same date as the launch.

Work Around
The auxilary file n14avhrr.res can be edited so that the launch date corresponds to the date of the last calibration entry. This will give the correct results for NOAA 14 data after 31 January 1999. However, the launch date must be set back to normal from processing NOAA 14 before 1 February 1999, or for processing any other AVHRR data.

Number 2.0.1-1    Severity 3   Date 23 Nov 1999    NAO with zero reference count not destroyed    Status:     active
Description     If a ragged array is given as an argument to the shape function the ragged array is padded unnecessarily and
the resultant padded version of the input array is not destroyed. This can lead to memory being absorbed by the
zero reference count results.

Example:
                caps warp {
                        out_warpedNearest_avhrr_1
                }
#
# warpedNearest_avhrr is a ragged array. Use of this array as an arguement of shape generates an unnamed
# 0 reference count NAO of the padded dimensions and type of warpedNearest_avhrr_1
#
                nap "x = shape(warpedNearest_avhrr_1)"
#
# work around
#
                nap "x = shape(pad(warpedNearest_avhrr_1)"