I've pursued this area for more than a year and I think using Ruby, Rails, and Ruport could be fabulous for managing and sharing XBRL, and also help measurably to spur wider adoption.
I have a feeling if five of us spent a weekend hacking on this, our work could have a HUGE impact on how financial reporting takes place not only in the US but also internationally. And I bet it would make us decent money. Maybe we could meet in New York or Boston?
The hurdle I've never managed to overcome is the complex way in which one particular aspect of the XBRL specification is implemented, related to the Discoverable Taxonomy Set*. I've tried a lot of different tools, technologies, and approaches (and at least one company has open sourced their java code for this and the SEC has promised to make their code open source.) I'm still searching for an elegant solution to this. We need a DTS Ruby library - a lot would happen quickly if we had that.
The more I've worked with Ruby, Rails and the excellent tools like hpricot and Ruport, it seems ever clearer to me that this would be a superb technology / process approach. I'd be interested in any attempts others have made.
Here are a few key links / references:
http://www.xbrl.org/technical/SGS-PWD-2005-05-17.htm
http://www.xbrl.org/Specification/XBRL-RECOMMENDATION-2003-12-31+Corrected-Errata-2005-04-25.htm- Brian
* Note: My major hurdle is grokking / implementing the Discoverable Taxonomy Set = "
A DTS is a collection of taxonomy schemas
and linkbases. The bounds of a DTS are such that the DTS includes all
taxonomy schemas and linkbases that can be discovered by following links or
references in the taxonomy schemas and linkbases included in the DTS. At
least one taxonomy schema in a DTS must import the xbrl-instance-2003-12-31.xsd
schema. See Section 3 (of the spec) for details on the discovery process."
--
- Brian