Working Groups

This page will be used to track meetings, outcomes, documentation and updates related to the iRODS Consortium Working Groups. These groups meet to discuss specific topics related to iRODS technologies, generate standards and reference implementations related to the given topics. Products generated by these working groups will be available via github and linked here for reference.

Mime-Type Working Group Update: December 8, 2015

The first meeting of the mime-type working group met on November 23rd where Nirav Merchant, Sriram Srinivasan and Tony Edgin presented the iPlant Collaborative’s implementation of their mime-type and info-type services. The presentation may be found here.

Notes from the presentation:

  • File Type versus Info Type
    • container versus content: e.g. many csv files with differing content types
    • cannot rely on file extension to confer content type
  • File Types serviced by Apache Tika
  • Info Type serviced by locally developed InfoTyper service
  • InfoTyper – web service on top of Heuristomancer
    • based on Clojure and Instaparse
    • implemented as a series of EBNF matches – least specific to more specific matching
  • Info Type used on front ends for informing choice of editor
    • not used for computation, semantic reasoning used for compute resources
  • Headed to community supported project, grammars posted on github

The current consensus is that Tika provides a sufficient service for discovering the File Type, while the Info Type is an open question, and domain specific, with various implementations in production.

Call for Participation

We would like to invite the iRODS Consortium Membership and community to participate in one or more Working Groups the Consortium wishes to form. The iRODS Consortium Technology Working Group has identified several important topics that have broad impact, both on grant-funded projects in the academic arena, as well as on new and existing users from the community. These working groups would work to gather use cases, examine existing implementations, and synthesize standards and best practices around which the iRODS Consortium may provide packaged reference implementations.

The four identified topics include:

  1. Mime Types – see gihub issue 2096
  2. Persistent Identifiers – see github issue 2443
  3. Services related to the EMB such as:
    1. Auditing Framework
    2. Notifications

Once participants have been identified, we expect each working group to meet for hour-long web conferences about every two weeks for a span of approximately 3 months. We will begin by hosting presentations from our collaborators to examine existing production implementations. Ultimately, our desired outcomes are documentation and decisions on: message formats, schemas, serialization technology, topics and associated services, expected control flow from clients through iRODS to and from these services, and potential new API calls from iRODS to support new functionality.

If you are interested in playing a role in these discussions, please email referencing any or all working groups you would like to join.