Previous: Data Formats Up: Data Calibration
M.A. Garrett explained that MkIII data processing was still not a straigh forward task although improvements in the AIPS package had recently improved the situation. The main packages in use were the Caltech VLBI package and AIPS. However, a new package, DIFMAP had recently been written/released by Martin Sheperd (CIT). This included a particularly nice implementation of the difference mapping algorithm, previously only available in Jodrell Bank's OLAF-MAP package. It was ideally suited to single channel VLBI data. The editing facilities were superb and easy to use. The program had been ported to Sun and HP workstations. The only drawback was that the data were required in Caltech Merge format. It was hoped that UV-FITS files might replace the Merge format in the future. Most PI's were now turning to the AIPS package for fringe-fitting and visibility calibration. Several new tasks of interest to VLBI users had appeared in the JUL93 version of AIPS, including BLING (a baseline oriented Fringe Fitting program) and UBAVG (average time can be set for each paricular baseline). The data editor IBLED had also been improved.
P. Burgess asked if JIVE would undertake any software development. H.S. Sanghera replied that no development would be undertaken but the main VLBI packages (AIPS, CIT and now Difmap) would be supported. J.D. Bregman noted that it was a general policy of JIVE not to get involved in developemnt projects if they could be avoided. H.S. Sanghera expected PI's to be supported from session 3 onwards.
A. Mujunen described the port of AIPS to LINUX (BSD4.3 running on top of MACH kernel). The AT&T fortran to C converter had been used (this was now widely available on the Internet). There were still problems with tapes and the X-display server. M.A. Garrett suggested that the JUL93 version of AIPS might be easier to port than OCT92. It was certainly an interesting development.