RE: HumanMarkup: Paved With Good Intentions

[Manos Batsis <m.batsis@bsnet.gr>]

Before I go on, I would like to thank the people of this list that
posted their comments to www-patentpolicy-comment.

I am very sad to say I agree with all of Sean's points and more.

It strucks me to see how the list "ignores" Sean's resignation. Sean has
been by far the most advanced technical expert of this group.

It seems that the majority of the list cannot comprehend that markup
like <happiness level="someValue"/> is complete and *explicit* junk (I
never understood how one could *ever* reduce miscommunication by putting
an explicit value to express something even the person is experiencing
it cannot fully grasp and describe). Furthermore, emotions are hardly
the critical issue in communication these days.

The list, for quite some time, is stuck on conversations and random
ideas on "emotions", "emoticons", "gestures" etc. All these, would fit
in a markup that could be named cartoonML. The list is continuously
focused on what I have kept calling rendering layer, which is irrelevant
to our scope.

The list as I see it is divided into two parts. Half of us think through
XSD: Explicit mapping of concepts to elements/attributes, datatyping,
vocabulary design. The "descriptive" vocabulary discussed here is, IMHO,
useless, complex and not extensible. Although this part of the list is
by far the most active, the only thing I would like to keep from this
activity (and I'm not sure I want to do that either, I would be happy to
find a way to ban literals from the HumanML core. Still, they are
unavoidable in implementations) is datatypes.

From time to time, Sean, me and many others have tried to turn this
group towards the only possible technical approach, which is RDF(S) and
extensions to these. The idea, again, is to build a set of tools in the
form of properties that will enable one to build a custom classification
system for a specific purpose, based on our vocabulary (meaning the
properties, core classes, etc) that will be build on top of RDF(S),
DAML+OIL, DC (well, a modificated version, Dublin core has flaws) etc.
In other words, use RDF as a shortcut (it is one of the few well defined
models suitable for our purpose, if others exist) and build our model
one level on top of it. Then implementors could just use our HumanML
core to do whatever. In short, HumanML will be a metalanguage. Most of
the people in this list use XML to build vocabularies. Well HumanML will
be used (heh, if ever) to build classifications of meaning definitions,
vertical most of the time.

Another flaw, in our organizational level, is rush. Dates of useless
deliverables and tons of documentation around our random thoughts. I
guess the internet does this to you as well.

Finally I'd like to say, or rather, admit that since I was one of the
three initial proposers along with Ranjeeth and Joe, with the extra
responsibility of being more technical oriented than they where back
then, it's probably my fault more than anyone else's. I must admit that
ironically, in a group that tried to reduce miscommunication, I have
failed to keep the connection alive between the core members and the
worst part is that some of the reasons of this where completely
personal.

If there are people in this list that agree or disagree with me, it
would be a good time to show it. It is a rather critical point for
HumanMarkup. Better late than never...

Always with my kindest regards,

Manos





> -----Original Message-----
> From: Sean B. Palmer [mailto:sean@mysterylights.com] 
> Sent: Wednesday, October 10, 2001 3:45 AM
> To: humanmarkup-comment@lists.oasis-open.org
> Cc: Ranjeeth Kumar Thunga
> Subject: HumanMarkup: Paved With Good Intentions
> 
> 
> I've written an article which I suggest all HumanMarkup members and
> contributors read:-
> 
>    http://infomesh.net/2001/10/humanml/
> 
> It contains my note of resignation from the HumanMarkup TC.
> 
> Cheers,
> 
> --
> Kindest Regards,
> Sean B. Palmer
> @prefix : <http://webns.net/roughterms/> .
> :Sean :hasHomepage <http://purl.org/net/sbp/> .
> 
> 
> ----------------------------------------------------------------
> To subscribe or unsubscribe from this elist use the subscription
> manager: <http://lists.oasis-open.org/ob/adm.pl>
>