[foaf-dev] Suggestion for WOT vocab - PGP Word List fingerprints

[Earle Martin]

Forwarding this to the list, as I suspect Graham didn't mean it to
come directly to me. (Dan, could I put in a plea for reply-to munging?
*)

Earle

* Not intending to drag the list into one of the oldest holy wars on
the web, sorry.

---------- Forwarded message ----------
From: Graham Higgins <gjh at bel-epa.com>
Date: 2008/8/6
Subject: Re: [foaf-dev] Suggestion for WOT vocab - PGP Word List fingerprints
To: Earle Martin <earle at downlode.org>


-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

On 6 Aug 2008, at 18:49, Earle Martin wrote:

> my feeling is that a FOAF document, in
> the sense that most people use it, is a human artifact; so shouldn't
> require rigid mathematical purity (i.e., containing only axioms from
> which all other information can be derived). Hence the existence of
> foaf:name, rather than a mandatory and complex attempt to model the
> components of names -

Human artefact or not, I'm a little uncomfortable about the
practicalities of using foaf:name to model complex personal names,
e.g. (hypothetical but close-to-real example)

<foaf:name>General Sir St John Richard James Addison KCB, CBE, DSO</foaf:name>

(where "St John" is pronounced "/sɪndʒɪn/")

There are three main reasons for my discomfort.

Firstly, I think it's semantically dubious to use the general term
"name" for the character string:

"General Sir St John Richard James Addison KCB, CBE, DSO"

because, although it's semantically true in a general sense,

"St John Richard James Addison"

seems to be the more appropriate value to use as this person's "true name" [1].

But, that seems to lose quite a lot that is both crucial and distinctive.

Secondly, I'm not going to be able to get away with

<foaf:givenname>St John</foaf:givenname>
<foaf:givenname>Richard</foaf:givenname>
<foaf:givenname>James</foaf:givenname>
<foaf:surname>Addison</foaf:surname>

without wrapping the foaf:givennames in an rdf:Seq to preserve the
inherent ordering for use in subsequent serialization. However, I
understand that following such an approach renders the modelling
largely unsuitable for integration with LOD data (avoid rdf:Seq),
which is a shame.

Thirdly, as foaf does not allow me to express either the prefixes or
the suffixes (or their respective orderings), I must recourse to a
different vocabulary in order to model those semantics.

> although foaf:givenName and foaf:surname do
> exist, there is no requirement to use them,

Again, as a human artefact, both are essential if a list of
foaf-modelled personal names is to be sorted into the familiar and
expected (if admittedly culture-biased) surname-then-givenname order.

If anyone can constructively ease my discomfort, I'd be extremely appreciative.

[1] http://en.wikipedia.org/wiki/Personal_name

Cheers,

Graham Higgins

-----BEGIN PGP SIGNATURE-----

iEYEARECAAYFAkiZ9eQACgkQOsmLt1Nhivwj3ACg6rqyz9L55f3f7NP6z/U53J6G
3a8AnjCJIPEnjphD6JUA2b3XJGqRs4xOiQCVAgUBSJn15FnrWVZ7aXD1AQKRHwP/
YuPdztOpXwyIXuUEOPz0rToWhK8x+0eAHbWMxnaZxJd4gHmjByUAYCJ/Lx9dCvLT
VRK0hYNocewdQv0NunCgEuVI6UsnxEEjL/5T1TPaigyDxh2thlFRxCJk1VaID7yK
KroRudC7REykArcR773P3xSdYcDKvuKJOxlN49mbHW4=
=5KF/
-----END PGP SIGNATURE-----




-- 
我歌月徘徊，我舞影零亂。
Earle Martin | http://downlode.org/