Comments on the IFIP working group 2.5 proposal on "parameterization of the environment for transportable numerical software"

  • Authors:
  • S. I. Feldman

  • Affiliations:
  • Murray Hill, New Jersey

  • Venue:
  • ACM SIGNUM Newsletter
  • Year:
  • 1979

Quantified Score

Hi-index 0.00

Visualization

Abstract

IFIP Working Group 2.5 has proposed a set of parameters for describing the computing environment. This proposal appeared in the SIGNUM Newsletter [1], and simultaneously in several other journals. It is unfortunate that this paper is presented as a consensus view, when no reference is made to earlier published work on similar parametrizations (see especially [2] and [3]). There certainly is a clear need for a way to parametrize details of the computing environment. At present, each large numerical project uses its own set of parameters. A generally used and available set of such quantities is certainly long overdue. The numerical community should certainly propose and even adopt a set of its own, and it is laudable that the IFIP Working Group 2.5 has taken this task upon itself. However, I disagree with many of the details of the proposal.