CKAN companies with computation capability?

As I know, CKAN becomes one of successful open-data portal infrastructures in past two years. Many public and private data managers are hosting their data sources with it. As a partial effect, many valuable datasets are distributed in different CKAN domains (like isolated islands in data ocean in my view). People may be frustrated when they are looking for some data, or they just want the results generated by some **private** data that is not suitable for public access.

I just wonder if we can extend CKAN with some computation capability that facilitates the uniform interfaces and data types provided by CKAN ecosystem. In the future, we will have another **open computing portal** besides the CKAN-based data portals. When the user comes to these computing portals, their demands can be routed to different data islands and perform local computation (private computing portal), then the results are returned to the user. This process is flexible and elastic. Private computing portals can build personalized security controls and local computing infrastructure, but the public computing interfaces should keep tight with the entire ecosystem.


For a primary target, we can build a prototype of this computing portals with CKAN. Some basic features, as I imagine, include a plugin-based computing framework and a flexible communication protocol between these public and local computing portals. I hope this idea to deserve further development in open communities. Those who are interested in this can send me message and start a new project to make it born to this unideal world.

The issue names no files, tests, or entry points. Start by narrowing the proposed CKAN computing-portal prototype into a defined scope, plugin framework, and communication protocol; completion criteria are not specified in the issue.

Written by the indexing model from the issue text.