Re: DKRs without the "K"

From: N. C a r r o l l (
Date: Mon Feb 12 2001 - 15:53:02 PST

From: Eugene Eric Kim <>
To: ohs-dev <>

> On Mon, 12 Feb 2001, N. C a r r o l l wrote:
> > Spoke to Lee today, and he made an observation that
> > clarified my view of an OHS. He referred to the lower
> I'll throw in another twist, which will be further explicated in my tech
> notes (to be posted in the next few days, hopefully). I think that the
> underlying structure for these global DKRs will be defined primarily by
> the link databases.


> associations, etc.) without forcing us to adopt some sort of universal
> document type (i.e. the so-called "I-file.").

I'm not wild about the idea of permanent I-files, if only because of
storage issues. And versioning problems. Well, we'll get there.

> If you believe this, I would also argue that the most important elements
> of the OHS are the link database and a metalanguage for defining document
> data models (such as SGML groves property sets).

I don't know about "most important" -- I'm a whole-system-
works-together fan -- but data structures are far more established
than knowledge handling, so it's reasonable that they can be
more concrete than the middle knowledge level.

Had a look at groves. It didn't make my day, but then SGML
rarely does.


Nicholas Carroll

This archive was generated by hypermail 2.0.0 : Tue Aug 21 2001 - 17:58:01 PDT