Why would anybody want to use iRODS? It’s just a glorified file transfer protocol (FTP) system with a database, isn’t it? This paper answers the question. The authors point out that “there is a pressing need for the digital curation and preservation communities to implement strategies for ensuring long-term access,” and the iRODS rule engine provides a framework for such strategies.

The iRODS replication rules can be used to ensure that data is actually stored at multiple sites. But there is also a need to ensure that data stored today can be read by whatever applications may be available ten years from now. One approach to accomplishing this has been to save, along with the data, the current application software and its associated execution environment--this is not always easy to do.

An alternative, presented in this paper, is to use iRODS rules to convert digital objects at inception time to formats that are preferred at that time. For instance, objects that use proprietary document formats can be converted to an open Extensible Markup Language (XML)-based format. To illustrate this, a set of rules is presented for conditionally invoking micro services that convert Microsoft Word and Excel formats to OpenDocument and comma separated values (CSV) formats. The iRODS rule syntax explanations given here are clearer than the ones I’ve seen elsewhere.

However, this, in itself, is not sufficient. A requirement for periodic actions to be taken throughout an object’s life cycle exists, to ensure that the format in which it has been saved has not become obsolete. An example is given for conversion of files in TIFF format to JPEG 2000 formats.

Using iRODS 1.0, the authors implement a system that explores the concepts presented in this paper, using data from the Arts and Humanities Data Service (AHDS); they are now also looking at a technology-neutral approach that will insulate designers from the iRODS syntax.

I recommend this paper to anyone who uses or may be able to use iRODS. I look forward to seeing further papers on the author’s work.