Clause model options

["Peter Meyer" <pmeyer@elkera.com.au>]

Dear TC members,


1. INTRODUCTION

1.1 Background

At last week's TC phone conference, it was agreed that anyone with comments
on Jason's proposal presented on 17 November should present comments as soon
as possible.

I found Jason's latest proposal (simple paragraph model) interesting and a
good contribution to the process. The challenge we now face is how to
identify the best option for the TC from the contributions received.

I think it will be important that the TC is actively involved in the
decisions and gains a solid understanding of the issues involved so that the
resulting work genuinely reflects the TC's goals, rather than those of
particular members.

In recognition of the questions raised by John McClure about proposed XHTML
2.0, I think we now need to use that as a reference point for our
discussions. Reasons for this were set out in my posting to the list on 23
November 2003.

1.2 Some observations

I would like to offer some general observations on the issues and propose a
framework for going forward.

The options
I think there are now 3 basic options:

* XHTML 2.0 (in draft only)

* Item and grammatical paragraph model (submitted by P Meyer)

* Item and simple paragraph model (submitted by Jason Harrop).

The 3 models have some quite distinctive features which I want to highlight
for comparative purposes. It is possible that we will need to modify any one
of these or combine aspects of them to achieve our objectives.

I assume that each of the models can be used to markup pretty much anything
that is found in contract documents, after some refinement. There are some
issues about how to do inline lists etc but this is not the critical issue
at present. Each of them can capture the basic clause pattern in some way.
However, there are important differences that should be understood.

None of the models is a perfect model. From a technical perspective, it is
possible to look at each and find points that we don't like or holes in
their area of coverage. Many of these depend on our frame of reference. Are
we more focused on developer issues, authoring issues, publishing issues or
metadata issues?