[AMRITA]

ASME CFD Policy

The document:

numerical-errors

was constructed in response to ASME's new CFD policy.

[?]

Comment

This document shows that PDF could be used to deliver worked examples as part of an effort to define computational standards. The motivation for the approach is given in:

Computational Science
  "Same Old Silence, Same Old Mistakes"
  "Something More Is Needed ..."

As it stands the present PDF should be viewed strictly as proof of concept, and since it stretches Adobe's format to the limit of what is currently possible, depending on your version of Acroread you might experience viewing problems. Specifically, some versions of Acroread are sluggish to refresh pages containing a large number of widgets, and this document packs 2000+ widgets onto a single page.

The take-home message, however, is that electronic document formats allow reporting possibilities well beyond the current thinking of ASME and other professional bodies.

You may well feel that it is fanciful to expect authors to package their work in the manner shown here. But computational authors have a scientific responsibility to ensure that their work is repeatable and so cultural resistance to AMRITA-sytle literate-programming will be eroded in time, especially once commerical packages pick up on the underlying principles.

Of course, without pro-active intervention, it could take 20 years, or more, for scientific journals to recognizise the educational potential of requiring authors to submit self-substantiating articles.

© 1988-2005 Website server donated by Kama'aina Productions