Project 58: Requirements for documentation of numerical software  

Como 1987:
==========

Stetter briefly summarized problems aimed at by the project, and noted
that the activity on the project was low during the past year.

A discussion followed. It was noted that often it is not easy to
define applicability of a piece of software and document it, and that
often completeness and conciseness are mutually exclusive. The problem
of documentation and the interface between the documentation is very
often a human problem rather than a knowledge problem.  Hypertext was
given as an example of a system showing only information of interest
rather than the complete information available on the subject. It was
recognized that both "long" and "short" forms of software
documentation are necessary, but that it may be difficult to select
the correct proportions of the two. It was decided by the group that
the topic was of interest and will not be dropped, instead the
subgroup will attempt to produce a document, and then merge the
project with project [56] (Specification of knowledge bases for
numerical software). (Delves, Gaffney, Gentleman, Reid, Ford, Dekker,
Feldman, Vouk)

Stanford 1988:
==============

Document:	IFIP/WG 2.5 (Stanford-18) 1518, 1 page.

H. Stetter submitted report entitled "Requirements for Documentation
of Numerical Software" (1518) by mail. He suggested that the project
should be merged with project 56 (Knowledge Bases).

Project [58], Requirements for Documentation of Numerical Software, was 
completed and closed. Its activities will continue within project [56].