Re: TMQL: Direct (super|sub)types?

On Mon, Feb 28, 2011 at 11:32:15PM +0100, Lars Heuer wrote:

> Today, my attention was drawn to
> <http://www.itscj.ipsj.or.jp/sc34/open/1586.pdf>.
> 
> Even if I try to restrain me: What are *direct* supertypes/subtypes?
> Doesn't it withdraw everything we (should) know about TMDM,
> type/instance relationships and about super-/subtypes relationships in
> general?

Today is Groundhog day[1]. Yes, we have been there before. :-)

Yes, you're correct: a "direct/indirect" distinction breaks TMDM
semantics. At least the little as there is.

OTOH, the need to make this distinction is legitimate, as folks who
want to treat the map "as-is" (editors come to mind) need to
distinguish between direct neighborhood and the inferred one.

This is yet another instance where TMQL will have to ... let's call it
"supplement" TMDM.

In one of the ISO meetings - which I so fondly remember - we came up
with the idea to introduce a pragma which tells the machinery which
subtyping inference it should use:

  http://kill.devc.at/system/files/tmql-2008-07-15.html#Pragmas

Ha! The good old days... back to RDF.

\rho

[1] "Groundhog Day" isa movie
    http://en.wikipedia.org/wiki/Groundhog_Day_(film)
    comment: ... it's about doing the same thing every day ...