Steve and all, hello.
[replying with rather a lag]
[...and this is potentially a tangent to the subsequent direction of the
discussion]
On 29 Apr 2016, at 18:30, Steve Newcomb wrote:
> It is also necessary to know that <x>, just as it sits there, /is in
> fact also both <y> and <z>,/ so, dear maintainers of these data:
> "BEWARE, lest you screw them up!" With AFs, a maintainer can always
> test whether those who need <z> have been inadvertently
> disenfranchised , even if the maintainer doesn't have a clue what <z>
> is all about, really, as long as the constraints imposed by the <z>
> interpretation are specified adequately.
Yes -- my impression of AFs was that they were much more concerned with
the idea of loose or even reluctant collaborations, rather than the more
'contractual' atmosphere that surrounds DTDs. So this idea of AFs being
able to manage overlapping contracts, and keep everyone honest, hadn't
struck me before.
One could examine (or qualify or expand) that idea more, but the thing
that struck me at the time about AFs was this idea that _a document
could be more than one thing_ -- different people could have very
different ideas about the important content, and actual structure, of a
document.
After I'd failed to persuade my community (astronomical data management
and software) that AFs were a wonderful thing, I spend about the same
number of years trying to persuade them that the Semantic Web was a good
thing, with the same intuition in mind: Alice saying that this
particular piece of data is an X need not preclude Bob saying that it is
also a Y, if all Xs are Ys _as far as Bob is concerned_. Alice may not
care about Ys, and even if she does may not agree that all Xs are Ys,
but that needn't stop Bob using this document in a way useful to him.
Now, AFs are at least partly concerned with syntax, whereas the Semantic
Web wasn't, at all ('you go and parse your stuff, then come back and
we'll talk about what it means'). That means that fewer guarantees are
possible in the Semantic Web: while the 'Semantic' bit is cold logic,
the 'Web' part of the name invokes the 'style' of the web, with 404s,
where stuff happens, and we try to make the best of it. But where a
Semantic Web approach is appropriate, it's probably more realistic, in
the sense that the world is complicated and doesn't always reduce to
syntax or contracts. Loose coordination, again.
(I didn't, in the end, have much success in persuading my colleagues
that the Semantic Web was the wonderful solution I thought it was,
because its strengths don't really apply to the technical problems that
were actually blocking progress for us, but that's another story).
Best wishes,
Norman
--
Norman Gray : https://nxg.me.uk
SUPA School of Physics and Astronomy, University of Glasgow, UK
