Call: Scope/Approach/Procedure HumanML Personal Summaries

["Emmanuil M. Batsis" <mbatsis@humanmarkup.org>]

This is just a post of my personal view on the subjects but most importantly a Call for
List Participants to do the same.

Scope
=====================

HumanML's initial scope was the attempt to:

1) Reduce human miscommunication.

To do that, we need to:

2) Create machine understandable taxonomies that describe various aspects as humans
anticipate them.

I think that does summarise it quite nicely for me. Note that the second <li> widens
usability beyond 1.


Approach
=====================

As Kurt Cagle points out and extends what we already have agreed upon back in Yahoogroups
(HumanML modules, ns methodology, "cool URIs"), modularization is to be the most important
aspect of our methodology.

This brings up some taxonomy issues. Modules established by HumanMarkup should not be
vertical (we'll never get the job done; not that HumanML will ever be finished but...).
Ok, maybe I should rephrase that... um... IMHO, modules need not be usable right out of
the box. Instead, they should be composed by classes everyone can work upon, to build a
more specific fit for his application. Of course, one should be able to use them as is,
but extensibility is the key and probably what we want from implementors too.

This will probably have the benefit DCMI spotted some time ago; when a third party reads
your data, it should be able to somewhat understand the meaning thanks to the upper class
in the hierarchy (the one we established).


Procedure
=====================

Throw module proposals on the table, clean up/redefine, sort their priority and get on it!


Kindest regards,

Manos