next up previous contents
Next: Revolution: the AIPS++ option Up: Evolution or Revolution ? Previous: Evolution or Revolution ?   Contents

Possible Options

GILDAS was developped and handled by a relatively small group of people. It is estimated between 0.5 to 1.5 Full Time Equivalent persons were working on the development of GILDAS. This manpower was devoted mostly to the application domain (CLIC, Tasks) and only very seldom to the ``system'' domain, where support for portability has been the most time consuming task. The total man-years is thus presumably of order 15.

Other interferometry/single-dish packages like MIRIAD and GIPSY, are being abandonned by their support observatories to be eventually replaced by AIPS++. In comparison with GILDAS, the AIPS++ project now has 17 FTEs working on the project, and has been on-going since early 1991. The estimated number of man years exceeds 100, and the first public release is foreseen in March 1999.

It is clear that the two softwares cannot be compared. However, the future of GILDAS, and of the IRAM data processing software in general, must be adressed now. IRAM faces several options: continuing support to GILDAS, joining AIPS++, or re-investing into other data handling packages, either commercial (like IDL) or non commercial (like IRAF). Several observatories have used IDL extensively for data reduction, mostly in Space programs (e.g. ISO), but have faced compatibility issues between different releases of the IDL suite.

Based on such experience, and given IRAM field of applications, we feel only the AIPS++ or GILDAS(++) options are realistic.


next up previous contents
Next: Revolution: the AIPS++ option Up: Evolution or Revolution ? Previous: Evolution or Revolution ?   Contents
Gildas manager 2014-07-01