The CoreGRID WP3 meeting in Pisa, Italy was not the success I was hoping it to be (at least for me since I am the only one disagreeing with the direction… the other partners seemed reasonably happy).
It seems to me (I guess I am always a sceptic) that yet again we are re-inventing distributed computing concepts for the “good of the Grid”. No one accepted my point that Internet-scale applications are based on protocol-integration and not abstractions like objects or coarse components with bloated interfaces, associated behavioural semantics, and a requirement for these to be globally understood. I think the Web has a lot to teach us but we are not listening. The discussion seemed to focus on the need to define high-level component models that are composable and expose the kind of functionality that “Grids need”!!!
We should keep the architecture simple and focus on composable protocols. The work on semantics will give us common understanding of data/information (I hope), while behavioural semantics will be captured through protocols. We’ll be applying policies and contacts in a declarative manner (e.g. the implementation of WS-SecurityPolicy on WSE). Large-grained, semantics-rich, all-encapsulating components should remain a thing of the past.
Let’s see what the Grid and P2P Architecture workshop will be like. BTW… just arrived in Crete, Greece and staying at a very cool hotel. Highly recommended (free Internet broadband access and a really really nice hotel 🙂
There’s a unique energy that comes with starting something new — a blend of excitement,…
As I continued work on BrainExpanded and its MCP service, I came to realize that…
Just over a month ago, I published "Playing with graphs and Neo4j". Back then, it…
After my initial implementation of some BrainExpanded-related ideas on top of dgraph using its GraphQL…
Say hello to the Graph Model Domain Specific Language (GMDSL), created with the help of…
As I wrote in previous posts, the manual recording of memories for BrainExpanded is just…