An application server which exposes resources in a tree.
It's structured as a tree of Nodes and files (as leafs) in a domain (the root), nodes carries properties (like name, ACL, owner, and many more...) and can contain Nodes or files. Each Node is a couchDB document, as well as a object instance in your vm.
A node can be exposed over HTTP, node's instance and method can then be queried via JSON/REST (first), XML-RPC, SOAP (later), or even WEBDAV (and that would be lovely)...
Writing an application for KaraCos will consist of subclassing node or domain object and writing functions.
-
Mastering javascript, css, and html should be enough to develop webapps, even for server-side code.
-
nodejs rocks!
-
Google's V8 vm rocks!
-
The @bergie 's post finished to convince me about pertinence of javascript, event for server-side scripting.
-
As we want karacos to be able to run on a desktop, a minimal dependencies deployement would be fair.
-
As nedis exists, and is a drop replacement for redis, it will allows to have a single nodejs instance for a single user instance.
Once features of the karacos-nodejs impl will reach sufficient level, the KaraCos python impl will stall and certainly die or eventually find a new maintainer...