- From: Seaborne, Andy <andy.seaborne@hp.com>
- Date: Mon, 10 Aug 2009 14:21:48 +0000
- To: Jacek Kopecky <jacek.kopecky@sti2.at>
- CC: SPARQL Working Group <public-rdf-dawg@w3.org>
> -----Original Message----- > From: Jacek Kopecky [mailto:jacek.kopecky@sti2.at] > Sent: 10 August 2009 15:05 > To: Seaborne, Andy > Cc: Lee Feigenbaum; SPARQL Working Group > Subject: RE: Discovering service descriptions > > Hi all, > > On Mon, 2009-08-10 at 13:47 +0000, Seaborne, Andy wrote: > > > Option 1 - HTTP-Header > > > > OK. > > > > Extra round trip in some situations. But "ASK {}" is usually cheap > > and it opens the connection so next time it's ready. > > > > Variation: location is another SPARQL endpoint loaded with the > > description ready to be queried. Descriptions might get quite large > > when including lots of detail about the data available. > > Variation on variation: the original SPARQL endpoint should have the > location as a named graph, ready to be queried. The location header > would give us access to the file or to the graph name. > > > Variant (a bit of a cross between 2 and 3): add a new request parameter to > ask for the service description (including content negotiation). Sort of > like (2) but moves the handling firmly into the service code. > > > > http://example/service?servicedescription > > > > Not keen but can live with. Option 7 seems better. > > Many Web Services stacks do that, so it's an already well-known pattern. > Its standardization in SPARQL might strengthen it. > > Also, this option seems to be part of the protocol, and as such it might > be coordinated with the issue about how to do update through the > protocol. > > > > From what I've heard in conversations so far, I think that 2 - HTTP > > > OPTIONS, 3 - Well-known location, and 5 - Query all lack strong support > > > from anyone. Please correct me if I'm wrong. > > > > I can live with option 2. > > I would like to see option 2 investigated, because if successful, this > pattern would IMHO give life to the underused OPTIONS method. > > > > And another option: none of the above (0?) > > All the options above for discovery of the service description hinge on > the client either a) knowing the service endpoint, or b) having access > to a SPARQL engine other than through a network protocol: > > a) how does the client learn about the service endpoint? This is the way > the client might learn about the service description as well; in fact, > the client may even just learn the service description URI and the > description will point to the service endpoint. > > b) if the endpoint is accessible through a local API, the API can easily > have a method like getServiceDescription() or queryServiceDescription(). > I'm not sure if the group is standardizing an API, is it? > > In other words, the feature may be seen as a bit backwards - first the > client knows about a service and then it wants to discover the service > description? > > > Has this been considered and tabled already? > > Jacek I agree that this is just one way round where the endpoint is know and the description is to be obtained. I can envisage learning about endpoint from elsewhere. A rather useful case would be an aggregation of know services so an application can go and find a suitable endpoint for the data it needs. I'd hope this because an important usage as it's about client code discovering places to go rather than being hardcoded. I don't think it requires any standardization at this point in time. I'd see it as emerging, like search engines have emerged. Andy
Received on Monday, 10 August 2009 14:22:24 UTC