Some years ago I had the opportunity of participating in a Document Workflow System that stored (semi-structured) documents as XML in an Oracle DataBase (using CLOBs).
The idea behind the project was very interesting but the implementation was pretty lousy and -the worse of it- inefficient: since documents where not structured, queries were not prepared but dynamically generated and executed.
Few years latter there was a new re-architecture but the data model was imposed and based on Oracle and vertical storage. Not bad but required a lot of tables, joins and insertion painful since inserting a document required multiple insertions and the number of records on each tables huge.
I still think about the goodness of the project and mostly in cloud environments.
So, I’ve started working on the system using Apache CouchDB as document storage and take advantage of the already existing infrastructure.
This blog is about the technologies used (CouchDB for Data persistency) and Kendo UI (for UI / UX) and about the project itself.
Eventually, I will publish the code and maybe you can start using it for defining Document Workflows and Business Processes.
Interested on the project please contact me!