ADASS XXI BoF : Scientific Workflows in Astronomy

Presentations

• Introduction and Workflows@IVOA, André Schaaff
• Wf4Ever, Juande Santander Vela
• IceCore, Sami Maisala
• UVES workflow, Pascal Ballester
• Different Workflow levels in a project, Christian Surace
• OWL, Francesco Pierfederici

Discussion points

• How to boost the creation / sharing of Wfs knowledge in our community ?
  • People could be reluctant to fill with care the information required as part of the workflow preservation (data used, services, references, etc.) or to make the effort to ensure that it is reusable by others
• Wf4Ever is a European project with a limited time life
  • How to preserve Wfs on the long term ? What kind of things (user datasets, methodology, tools for backward compatibility, Licenses??, etc.)
• What can we learnt / what have we learned from other communities ?
• Publishing Wfs : review methods (peer review, volunteer, etc.)
• What parameters would you use to rate a Wf ? Or which ones would you like a rating system to provide you, so that you can select based on their quality / rating ?

Questions and Comments

• Eric N. (Fermi National Accelerator Laboratory): R has facility for reproducible research
  • Data linked with method in same document
• Alberto A. (CfA, ADS): big difference between high throughput, and group of people working on different smaller steps;
  • In Wf4Ever we are eating our own dog food
• Petr S.: we are moving in spirals, mainframes, PCs, clouds
  • Services providers then have to bore the load of providing those computing facilities
• Doug M. (SAO): Workflows as mean of structured communication between users and facilities
• Rating workflows:
  • rate according to how can it hold delta changes in data
• Adriaan R. (ASTRON): there is no way astronomers trust someone other than themselves. *having ways to assess authorship might allay some concerns.