Provenance meeting with Norman 20/02/08

Minutes

• Open Provenance Model (OPM)
  • Questions were raised about how Open it actually was and the ability to extend - Paolo will contact Luc Moreau with questions regarding this.
  • Seems to be too coarse grained and constrained for our needs - artifacts do not identify whether they are collections or not, no ability to add additional annotations to assist lineage.
  • However, provides a good baseline for interoperability with other workflow systems - our Provenance Service should be built to support import and export of OPM provenance.

• BioNanny^? - possible consumer of Provenance for service reliability metrics. In the initial phase BioNanny^? will not make use of provenance but could become a future requirement we need to be aware of - Ian is to investigate this further.

• We need 1 model, which must support many uses:

*X denotes human user, requiring UI

• We need to think about UI requirements now, as this will influence the model.

• Sharing is affected by the design of EMO's, for which we are awaiting further details and documentation from Jits.

Mode of Operation

• Consultation / Requirements doc on each of the Tasks.
• Post consultation, consolidate to model, interface proposal and architecture.

Plan and timeline:

• T5 is a cross cutting concern that affects all requirements, and is also partially covered by the EMO document

• D6 is a final consolidated document bringing together all the requirements into a final design.

• From D6 we prioritise and build....

Actions

• Paolo - contact Luc Moreau on how we can contribute/influence the OPM.
• Paolo - finialise 1st draft proposal for T4 (lineage model)
• Ian - investigate and document 1st draft proposal for T2 (service reliability).
• Stuart - investigate and document 1st draft proposal for T3 (debugging).
• Khalid - distribute Result Presentation proposal.