To: Phyllis Jackson From: Eric Greisen Date: 1 July 1993 Subject: Quarterly Report on AIPS (1993 Q2) The 15APR93 release of "Classic AIPS" has been replaced with a 15JUL93 release which we expect to occur approximately on the nominal date. The delay is due primarily to the addition of support for Sun's new operating system, Solaris. For this addition, most routines written in C in AIPS were revised to meet the ANSI standard for the C language and many were revised to conform to basic Posix standards. In this way, we actually reduced the number of host-specific routines while increasing those that apply to all UNIX systems. The system procedure files were also revised to be more portable, general, and reliable. With help from other institutions, our support for Hewlett Packard, DEC, and Cray systems was improved. The television server XAS was brought up to the standards of MIT Release 4 of X Windows and the message server was made more reliable and transparently fast. In anticipation of upcoming major VLA surveys, two new tasks were added. The first (WFCLN) does wide-field, wide-band image deconvolution taking into account frequency effects (in source spectral index, single-dish primary beam, erroneous center frequencies) while using an improved weighting scheme. The second (OHGEO) regrids the images to correct the geometry to be normal to the phase center and to correct effects of the variation of the primary antenna gain over the observed bandpass. Another new and useful task (DTSUM) provides matrix and other listings summarizing the contents of visibility data sets. Several data manipulation tasks were added. These are BLWUP to blow up an image by pixel replication, IMTXT to convert an image into a text form useful for non-AIPS programs, MULIF to add or increase an IF axis in a data set, UV2MS to append single-source visibility data correctly in a multi-source visibility file, and SPECR to regrid a visibility data set in frequency using an FFT interpolation method. A new canned procedure VLBA may be used to do a quick, but detailed examination of data coming from the VLBA Correlator. The most serious bugs which we corrected were in the "gridded FFT" method of computing source models. The method did not work when there were both rotations and shifts. The AIPS test suite called DDT is being changed to test for this and assure future reliability. Baseline-based fringe fitting received numerous improvements and corrections. The standard fringe-fitting task was corrected to allow division by a clean component model and changes were made to allow such models to be used more easily in normal gain calibration. Task ANCAL which applies user-supplied, antenna-based calibrations to the data was overhauled and the general FITS reader FITLD was corrected for a number of problems including failure to catch parts of non-AIPS headers. UVPLT and other tasks were corrected to stop them from plotting every point twice. The AIPS tasks to convert plots and images to Postscript were generalized in the grey level of overlay lines (LWPLA) and in the handling of color lookup tables (TVCPS). Finally, a new verb TVPHLAME was added to do flame-like pseudo coloring on AIPS' TV displays. File: /home/primate/egreisen/AIPS/reports/QRep93b.TXT