Primer proposal

Hello,

In the last telecon, I was given an action to propose how to proceed
with the primer.

I propose that a primer document be written for release with the other
public deliverables, but that it should not be expected to cover the
whole model. It will include one or multiple examples, as best suits
clear explanation, and high level intuition about the model that aims
to be easy to understand rather than normative. The examples will be
primarily based around the provenance of web-accessible documents, and
we will start from, but not be committed to, the data journalism
example. The coding of the examples will use the formal model
ontology. The file example currently in the conceptual model will be
moved to the primer, but may or may not be included in the released
version.

The rationale for the proposal is as follows.

I think there needs to be guidance somewhere, otherwise documents
released will be hard for others to review. Including this only in the
conceptual and formal model documents separately will break up the
guidance. Including only the conceptual model without some (real)
serialisation makes it harder to see how it will be used, and the
formal model provides a simple, readable and W3C-appropriate
serialisation as Turtle snippets. While existing documents require
high level intuition and examples themselves, they should aim to be
brief and normative, which can conflict with giving best guidance.
Putting primer information on the wiki means it will get lost and/or
messy, and would be hard to release with deliverables. The primer can
aim to cover all or most of the model, but recognising that the model
is still evolving, we should be prepared to cut out of the
September/October release anything that may not match the version of
the model released, or is too contentious or too minimal to be useful.
The documents are being released in a particular context, the W3C, and
so will be most easily understood if the examples are around web
resources, with web accessible documents such as web pages being the
most commonly observed case.

I understand the concerns of creating a primer while the model itself
still evolves, but preparing the primer could help in stabilising the
model both now and after the first release, as external reviews would
hopefully be more about the model itself rather than the way it is
written if a primer is available.

Thanks,
Simon

-- 
Dr Simon Miles
Lecturer, Department of Informatics
Kings College London, WC2R 2LS, UK
+44 (0)20 7848 1166

Received on Monday, 5 September 2011 20:38:16 UTC