Re: Decentralized RDF Distribution

Danny Ayers wrote:

> Some form of bracketing will be needed as Frank suggests - otherwise you
> could have a scenario very comparable to the transaction classic of debiting
> one bank account and crediting another - if you stop halfway you've got an
> invalid (or at least undesirable) state. Some form of imposing ACID on a
> group of triples? Ideally some form of existing grouping (or one that could
> be easily produced using an existing tool) - but what about putting the set
> of triples (or even URIs) that should behave together atomically in a bag?
> Or is there a neater way of grouping (potentially distributed) items?

How to put statements in bags is a hotly contested issue in these waters.  But
if we allow the statement Id  to be a URI and to be the ID mentioned in 6.6 of
http://www.w3.org/TR/REC-rdf-syntax/#grammar , then it could work nicely.  That
way we can draw context arcs directly to statements as depicted in this
picture:  http://robustai.net/mentography/arcarc.jpg  where P2 might be labeled
"contextContains" and S1 is a Context.  But Ora seemed to think this was just a
cute diagram with nothing to do with RDF ... so go figure.

> The transaction itself shouldn't be too problematic - the relevant part of
> the database made read-only for the duration of the transaction (snapshot
> taken), all (grouped) changes made, read-write reinstated - I see problems
> with queueing where the sequence of the transactions is important, maybe
> better to reject an incoming update (return to sender?) if an update is
> already in progress.

When talking of things like "read only" we cannot be talking about the semantic
cloud, rather we must be talking about a closed world collection of triples or
quadruples .. that is we've got to start  distinguishing between a local
coherent semantic memory and the greater semantic cloud.

It seems I have started talking about "The Semantic Cloud" rather than the
"Semantic Web".  Perhaps this is because it has slowly dawned on me (us?) that
this thing will not be a web like the www was at all.  If you think about it,
you might agree.  With the www, everybody in the world can go to a URL and if
they have access to it they will see pretty much the same thing [and yes I know
of the many notable exceptions to this].  But this will not be the case with the
semantic cloud.  We will all see these triples through the eyes of our own
semantic memories .. it will be quite different than a web.  Anyway I coined a
new acronym and can now say something about it:

<semTran language="Semenglish">
TSC acronymOf "The Semantic Cloud"; (everything in the) (collection of
Quadrapules).
ASM acronymOf "A Semantic Memory"; (a particular local collection from) TSC, a
CWM.
CWM acronymOf "Closed World Machine";
     seeUrl "... woops my semantic memory fails at this point ...".
semTran a (transaction to TSC).
</semTran>

Seth

Received on Monday, 19 February 2001 14:13:32 UTC