09 March 2015
- Dan Coughlin
- Jon Dunn
- Will Cowan
- Randall Floyd
- Karen Cariani
- Mike Muraszko
Notes from 3/9/2015 Sufia roadmap call with Mike Giarlo and Dan Coughlin from Penn State University, Jon Dunn, Will Cowan, Randall Floyd from Indiana University, Karen Cariani, Mike Muraszko from WGBH.
Mike G- The roadmap for Sufia is hard to predict because it’s development is based around what it’s users are asking for. Fedora4 in Sufia 6, really close to being officially released. Enabled support for large files up to 7 or 8GB.
Jon- IU might want to use Fedora4’s federation for large files but currently that feature is read-only, not write.
Mike G- University of Alberta is using Sufia as a base for Hydra North.
Portland Common Data Model, supported by Sufia. Supports different kinds of instantiations (One object master file, mezzanine, and proxy all having the same metadata).
PCDM basic support in May sprint
Ordering and nesting support late May / Early June
Other 2015 / Early 2016 features, mediated deposit, user interface, need for Admin section. There is a need for an admin section vs user section at institutions like Stanford.
University of Washington want batch ingest, but we want to get their requirements so that many folks can make use of that feature.
Karen- What about the ability to add metadata only records?
Mike G- Had a test instance that we were able to add metadata via spreadsheet with a field that had a value the repository was able to connect with.
Jon- Who decides what gets into Sufia?
Mike G- Someone with a Hydra CLA submits a PR, 1 or 2 people review it, any major changes could be discussed on Hydra Tech calls, like changes to the generic file or data model.
Karen- How do we contribute developer time?
Dan- Look at the doc in Hydra Community, sign up there.
Developers at LCDX will introduce new rules for those working on sprints, best practices.
Will- The May sprint will be dealing with PCDM?
Mike G- Yes, multiple files with single metadata record.
Will- File specific metadata is supported but we may not want that right off the bat.
PCDM- Handles FITS data output, stored as XML or map to RDF eventually, can support both.
Jon- How can we help with PCDM?
Mike G- After you get staff, have a follow up conversation with us and we can carve out IU/WGBH specific features for PCDM.
Mike G- Share your user stories, coordinate with Dan.
Dan- I’ll post a Google doc with our current user stories.
Mike G- Talk to me about putting in issues you want to see being covered, there are some PCDM docs on the wiki.