From: David Martin (martin@ai.sri.com)
Date: 04/03/02
one additional point -
David Martin wrote:
>
> Pat Hayes wrote:
> >
> > >ASSERTION: Some folks, who go to www.daml.org to be convinced they
> > >should start using DAML+OIL, are not finding the right kind of material
> > >to convince them.
> > >
> > >I base this statement on recent experience. SRI has an integrator role
> > >in DARPA's BioSPICE program. Part of this program's mission is to
> > >promote interoperation of data and software components used in
> > >biological research. Program participants are currently considering
> > >issues about ontologies and languages for use in exchanging data. Some
> > >of them are interested in learning about DAML, and have visited the Web
> > >site. But they are not coming away with a clear picture of why they
> > >should use DAML, as opposed to crafting some data exchange standards
> > >using (just) XML.
> >
> > Well, good point. Why SHOULD they be using DAML rather than crafting
> > some data exchange standards using XML? Can you answer that question?
> > I can't.
>
> Actually, I think I can answer that question reasonably well, and I
> think most folks reading this list can answer that question reasonably
> well. My claim is just that there's (practically) no place on daml.org
> where the answer is stated clearly. I'm hoping to write up a few
> paragraphs for the BioSPICE folks, and if and when I get that done, I'll
> share it with this list.
>
> I note, as an aside, that (as far as I can tell) none of our (DAML
> researchers') papers about DAML or semantic Web are linked from
> daml.org. It may be that simply linking in existing papers would go a
> significant ways toward addressing this need.
>
> >
> > >My claim is that (believe it or not) the answer to their question ("why
> > >should I use DAML+OIL") is not clearly stated anywhere on the Web site,
> > >or if it is, it's not easily found. ***What I think is needed is
> > >something like a technical white paper of the sort that technical
> > >companies typically make available on their Web sites, the main purpose
> > >of which is to state the attractive features and advantages of their
> > >technology, in a way that resonates with the audience of potential
> > >customers.***
> > >
> > >I'm not finding much like that on www.daml.org. The closest thing is
> > >perhaps the Scientific American article, which is great, but not at the
> > >right level for these potential "customers". (It's a good start, but to
> > >"make the sell", something less visionary is needed, which talks about
> > >the pragmatic benefits that might be expected over the short-term or
> > >medium-term, and possibly focuses on the concerns of a particular domain
> > >or a particular type of user.)
> >
> > BUt over the short term and forseeable medium term, there is no clear
> > advantage. The advantage, if there is one, will only come when a
> > large enough number of people use DAML to mark up their websites. I
> > don't see this as likely to happen in the immediate future.
>
> I agree, and that's an important issue. So I should reword my statement
> (above):
>
> Actually, I think I can answer that question reasonably well, in
> principle, and in also in pragmatic terms, *under certain assumptions
> about the widespread adoption of DAML*. If your main point is that the
> widespread adoption of DAML is unlikely in the immediate future, then
> there's no argument here.
I would add that getting the benefits of DAML+OIL doesn't actually
require "widespread adoption" across the entire Web. Widespread
adoption within a particular community of use (such as biology
researchers) would be enough for that community to benefit.
- David
> ... stuff omitted
This archive was generated by hypermail 2.1.4 : 03/26/03 EST