Proposal for a new metric for WG11 standards

[Martin Hardwick]

All,

I would like to propose that in the future all Wg11 standards
should be measured against a set of STEP patterns that commonly
appear in APs.

For example, how well can the method encode common patterns like
the geometric representation of a rectangular block, the product
identification information (product/pdf/pd), a mass property,
and an assembly relationship (product structure, pd_relationship).

The requirement would be for each standard to show the encoding
of their implementation method against these patterns in an
Appendix. 

The SC4 member countries would then be able to judge each proposed
implementation method using the patterns. The judgement would not
be absolute in the sense that there might be good reasons why a 
method does not look good with respect to one of the patterns.

However, if a project produced a method that took 10 pages to
implement a pattern when another project only needed 2 pages,
then the countries would be able to ask questions and if the
reasons were not good enough this could be used as a reason
for rejecting the proposed standard, just like today an 
implementation method is usually rejected if it does not cover
the whole of EXPRESS (this would continue under this proposal).

My hope is that a proposal like this would lead to a gradual
improvement of the Wg11 standards with respect to STEP.

Are there any other opinions on this issue? Should we debate this
proposal during the Wg11 plenary in San Diego?

Martin Hardwick