Comments inline Sent from my iPad
On 2010-04-07, at 2:56 PM, Steve Jones <[email protected]> wrote: And yet there still is plenty of reason to be frustrated. Crappy REST Apis abound, data still isn't any more interoperable with JSON, and HTML5 is deep in a mire... Yup. Which I think stems from the original objections of the REST community to having a more formal interface description language. That will change, IMO. Not holding my breath, of course. And I think a huge difference is that, no they're not building them the way they always have. issuing SQL via VB or Powerbuilder was predominant... the use of URI and HTTP are big, interoperable differences. Even for those trying to being that old model back for the Web (like Microsoft with OData) Is that really a big difference? Sure the technology has changed but SQL was the "equivalent" back then. Sure HTTP/URI is more open but the net effect is still client/server just using a different (but better) technical protocol. Architecturally its the same even if in implementation its different. Architecturally it is an evolution of client/server. The big difference is on the data elements and in interface uniformity. That's a pretty big deal for interoperability. I'm curious what vision this is.... that everything would be magical and automatically interoperate without work? or the current reality of Mashups and crappy REST APIs that are reasonably easy to build and consume? This was the "vision" that MIME and "dynamic" REST based interfaces would enable the "automatic" consumption of services. Its me grumbling (yet again) that the technical myth of dynamic interfaces is rubbish while more structure published approaches (such as the Apple APIs) seem to actually work despite their technical "limitations". Firstly, I don't think dynamic interfaces are rubbish. I'll admit the jury is still out. Second, Apple's APIs are for the UI, not networked information exchange. The challenge is conceptual change not technical change. They are reflexive, in that they both influence each other. I'll give you that Client/Server is still Client/Server even on the Web. But the infrastructure people build client/server apps today allows them to consume content from across server implementations, trust domains, and geographies -- something you could never do twenty years ago. I worked on programmes where we exchanged information between systems across geographies and domains but in a very restricted domain. I'm not disagreeing but just saying that its the conceptual model that matters more than the technical one in actually delivering change and we haven't created a decent conceptual model to really replace client/server. I think one of the troubles is getting the terminology right. I have a feeling that what you call client/server is much broader than what I have in mind. I continue to work with and seek extensions to REST because I think the data elements in an architecture have huge impact on scale and interop, more than just the topology of roles of the components themselves (c/s, p2p, etc). The web is a client/server topology with a p2p data model. What we do continue to lack are ways to make data content itself more interoperable, as the Semantic Web learning curve is pretty steep if you just want a little semantics. Though I see some positive signs there too, even if it will take a while... I think the Semantic Web stuff is a bit rubbish to be honest. I don't think the maths behind it stack up and again I think that explicit ontologies are better than dynamic ones. Having worked with RDF and OWL regularly for 2 years, I believe the maths behind the SemWeb aren't new, in fact they're the most solid aspect of those technologies, though they have the highest learning curve. The Web part is what is new. SemWeb is all about explicit ontologies. I don't quite know what a dynamic ontology is... certainly it's not in widely deployed practice to my knowledge, so I'd gladly call it rubbish. ;) Cheers Stu __________________________________________________________________ The new Internet Explorer® 8 - Faster, safer, easier. Optimized for Yahoo! Get it Now for Free! at http://downloads.yahoo.com/ca/internetexplorer/
