background: Andy Seaborne and I have been thinking about improving the SquishQL syntax. Andy has an implementation which is looking rather more professional...
Libby Miller
libby.miller@...
Sep 6, 2001 7:55 am
21
... Date: Wed, 1 Aug 2001 20:32:37 +0100 From: "Seaborne, Andy" <Andy_Seaborne@...> To: "Libby Miller (E-mail)" <Libby.Miller@...> ...
Libby Miller
libby.miller@...
Sep 6, 2001 7:58 am
22
I totally agree with this, and I have actually implemented it (without telling anyone ;) libby ... Date: Wed, 1 Aug 2001 20:29:41 +0100 From: "Seaborne, Andy"...
Libby Miller
libby.miller@...
Sep 6, 2001 7:58 am
23
... Date: Wed, 1 Aug 2001 20:34:21 +0100 From: "Seaborne, Andy" <Andy_Seaborne@...> To: "Libby Miller (E-mail)" <Libby.Miller@...> Cc:...
Libby Miller
libby.miller@...
Sep 6, 2001 8:01 am
24
[snip] Nice to have a substantial technical issue to wade into ;-) Here's my reason-to-date for preferring PSO syntaxes for a textual RDF query format (I don't...
Dan Brickley
daniel.brickley@...
Sep 6, 2001 9:08 am
25
... Yes it is :-) I prefer SPO for exactly the reasons Dan stated :-) My notions of RDF are graph-centric; I tend to read arcs in an SPO order, I'd like to see...
Jan Grant
jan.grant@...
Sep 6, 2001 9:25 am
26
... Correct; you either need to chain queries together (so that the identity of a variable can be passed from one query to the next) a la compound queries in...
Jan Grant
jan.grant@...
Sep 6, 2001 9:31 am
27
There is an alternative that's occurred to me, to provide explicit access to internal IDs by function. Lifetime considerations and portability issues as per...
Jan Grant
jan.grant@...
Sep 6, 2001 9:40 am
28
... It does make it more read-out-louadable too, yes. ... I don't like that; it introduces a context dependency: you need to jump to the namespace declaration...
Dan Brickley
daniel.brickley@...
Sep 6, 2001 9:46 am
29
... If it's called a "namespace" and it looks like a namespace in the only w3c-rec'ed syntax for RDF, then it should resemble a namespace in the query...
Jan Grant
jan.grant@...
Sep 6, 2001 9:55 am
30
[Seeing as there's traffic, I thought I might as well wade in :)] ... Just to let people know I've been working with Inkling recently, to refactor some of the...
Leigh Dodds
ldodds@...
Sep 6, 2001 9:59 am
31
re s, p, o/p,s,o, I don't care as long as we settle on one...good arguemnmets on both sides. re double/single colon, the problem I found (as sopmeone who has...
Libby Miller
libby.miller@...
Sep 6, 2001 11:28 am
32
RDQL is an implementation is SquishQL speciifcally for Jena: It's at: http://www-uk.hpl.hp.com/people/afs/RDQL/ the grammar is at ...
Seaborne, Andy
andy_seaborne@...
Sep 6, 2001 11:53 am
33
There is unlikely to be a sound logical argument either way on these sylistic issues so its down to the personal preferences and voting. FWIW my personal...
Dave Reynolds
der@...
Sep 6, 2001 12:15 pm
34
... Yeah, I've come around to that way of thinking too. As I was possibly the only PSO and :: (devil's) advocate, maybe this syntax question is resolved? To my...
Dan Brickley
daniel.brickley@...
Sep 6, 2001 1:59 pm
35
A very simple QL is good IMHO. I think we should fix a language for data query now for developers to use and also work on languages for including schema or...
Seaborne, Andy
andy_seaborne@...
Sep 6, 2001 3:15 pm
36
Jan, I think we would need the transaction guarantees you mention to give consistent interpretation. Otherwise model-specific identifiers might change: e.g....
Seaborne, Andy
andy_seaborne@...
Sep 6, 2001 4:17 pm
37
... I don't care much but :: (concatenation) is different from : as protocol or namespace. There is no meaning to the prefix fragment - it is not a URI for a...
Seaborne, Andy
andy_seaborne@...
Sep 6, 2001 4:31 pm
38
... Well I can see the value in the KISS approach but I think personally I will want RDFS/DAML-lite support (path expressions, concrete types, subProperty...
Dave Reynolds
der@...
Sep 6, 2001 5:37 pm
39
Inkling is an implementation of SquishQL in java, using the jdbc interface for rdf database plugginability. It uses p,s,o beacuse it was based on Guha's RDF...
Libby Miller
libby.miller@...
Sep 7, 2001 12:22 pm
40
Here's another style question for SquishQL: should URI/URIrefs be quoted in some way or should they be unquoted? A related issue is how should lists be...
Seaborne, Andy
andy_seaborne@...
Sep 11, 2001 10:37 am
41
... A brief (but hopefully compelling) argument against using the (otherwise attractive) <> chars for quoting: We will want to be able to ship SquishQL queries...
Dan Brickley
daniel.brickley@...
Sep 11, 2001 10:48 am
42
... True: it would be nice to ship queries in XML. Are you suggesting always-unquoted-URIs or allow unquoted-URIs-where possible. My dislike of the...
Seaborne, Andy
andy_seaborne@...
Sep 11, 2001 11:06 am
43
I'm anti-quoting for simplicity reasons, but we need to do this in a principled way since ) and ( are allowed in uris. I would suggest that it's probable that...
Libby Miller
libby.miller@...
Sep 11, 2001 11:17 am
44
... Agreed; the readability of something that's for machine consumption is a red-herring, imho. jan (who hates using imho) PS. That should be, "who hates...
Jan Grant
jan.grant@...
Sep 11, 2001 12:37 pm
45
... But a lot of modern-day scripting happens between angle-brackets. For eg., the various content management systems (cold fusion etc), or Java Server Pages...
Dan Brickley
daniel.brickley@...
Sep 11, 2001 1:26 pm
46
... Yes, and we can only hope that the world will recover. Or that the promised tools will appear. On the other hand, McCarthy was using M-expressions for ages...
Jan Grant
jan.grant@...
Sep 11, 2001 2:12 pm
47
Where are we here? The issue is around: - readability - embedding in XML, especially scripting 1/ To me, readability suggest that whitespace delimiting would...
Seaborne, Andy
andy_seaborne@...
Sep 12, 2001 2:41 pm
48
Another start-up with a new magic database paradigm. "Sentences is the first commercial implementation of the Associative Model of Data, the first database...
Dingley, Andy
andy_dingley@...
Oct 1, 2001 12:17 pm
49
Hi, I am another HP Labs researcher. I am working on matchmaking of services and automated negotiation. While I am not directly part of the Semantic Web team,...
Having problems with message search?
Fill out this form to ensure your group is one of the first to be migrated to the new message search system.
