...
Minutes
Strategic Plan
Report outs and communication plan https://docs.google.com/spreadsheets/d/1YH3uNxFXPgrRlOv3Nh8z6I3bWCVyC2k6jY21DFjzLvI/edit#gid=0
...
RM - we need feedback from beyond the Islandora and Samvera communities. I want to know that if there a barrier since we got rid of the flash interface.
Technical Roadmap
...
Release API spec
AW - Looking to release API spec. The candidate went out 6 months ago. Can we as a group recommend it for release. From the API editors perspective we feel it is done. We now need agreement to release. The charter says we need approval from the LG.
Ben - we also wanted to have 2 full implementations before doing that.
AW - yes we have some requirements/recommendations. I haven’t done a thorough scan but I need to figure out what I made up aspirationally and what the charter dictates. We won’t get 2 implementations of the spec at this point. It’s a confluence of I don’t think those were hard requirements. It’s basically stable - to changes or churn.
Ben if that isn’t a hard target that’s fine but it would be good to see it in full implementation.
Andrew, representing the API editors, making request to release
Consulting wiki and charter: candidate rec goes out, 3 month period of review and input, and that the recommendation (1.0) comes out if accepted by Fedora Leadership group as sufficient quality to become a recommendation
Ben: Would like to have two fully implemented implementations before we do this?
Andrew: yes, some was test suite and implementations – is this in the charter/hard requirements or is it realistic if that we will get two implementations? The spec is stable, there is no churn in what is going on with the spec.
Ben: if it isn’t a hard target it is okay, but would be good to have some impls that fully work that recognize the spec is fully valid.
Stefano – aligning lakesuperior to Fedora specs – in roadmap stage. Could become a full implementation of spec at some point, no timeline. By hamburg possibly?
Mark: Islandora 8 a fully functional consumer implementation (end of May) – Jared a committer, Danny on editor team. If we’d found anything wrong with the spec by now, we would know.
Ben: release 1.0, work on Fedora 5 and Drastic, if discover there are needs for change, then would go to 1.1
Ben – motion to accept
Approval in the room. David will send it out for a vote on the Leadership group email.
Technical Roadmap for 6.0
Design summary – in response to what the perceived areas are lacking.
Rosy- can you talk a little more about the query service – native query service on Fedora to ask a handful of questions – what objects updated within a certain time frame?
Question: does this design adequately respond to the concerns/needs of Fedora moving forward? Need consensus on a direction.
Mark: how does this summary translate into a roadmap and a sprint plan?
Andrew: essentially: fedora API, no modeshape, ocfl comes in, and the addition of a query service, and notes about migration, more robust fixity. Does this sound reasonable? And if it sounds reasonable, how to reconcile with conversations about the interplay between the API and OCFL.
Rosy: Memory that whether or not 3 was part of the migration plan. Concern is that 3 needs to be able to migrate to 6. Didn’t see this in the initial documentation.
David: need to be made more explicit, but it is there about migrations. Transforming objects on disk rather than import export.
Rosy: concern about those not in Islandora/Samvera – making sure that people understand they don’t need to use RDF.
David: more info has been added about how you can use Fedora without RDF.
Patrick: sharing this design doc with his team, he is feeling positive about the data.
Mark: versioning on summary and is it part of 6 plan? (Andrew: yes)
David: everything in the API is coming along.
Ben: mapping between LDP and OCFL – still some challenging issues. Still the case?
Andrew: convo hasn’t advanced since leaving Virginia Beach.
Ben: what about the questions between OCFL and LDP?
Andrew: fear that LDP was driving too much how we are implementing OCFL, persistence.
Ben: question: if there are still outsanding challenges, are they the type that could still derail the implementation (if we approve the design doc today), or that we can make them work?
Andrew: there is the possibility that there are some ‘sticky stickies’. Agreeing on these higher level aspects, we need to commit in participating in validating the software as it is being produced. We need pilots, partnerships, engagement in the process throughout the development process. Either now, or the very soon future, can we establish some pilot partnerships to validate the development towards Fedora 6?
Karen: Or can we pay for pilots? (Some institutions are short staffed or developers have other tasks)
Tim: want to empower Andrew in role to have someone to work with on this development. Could we use funds to hire someone to identify reasonable datasets and do performance testing.
Andrew: likes the idea of hiring testers. Less confidence that hired testers would meet specific needs.
Tim: consultant, do site visits, get content?
Patrick: timing? Could be a good time to test if it is in the fall when they will have some capacity.
Andrew: timing – key question. Great core set of developers. But they are very stretched. It is hard to figure out the timeframe with the current trickle approach to development.
Maurice: on-ramp for developers. How much do you need to know to test? Is it just testing, or do you need to parse what is wrong? Maurice has good shop with expertise in ingest, but no one with Fedora development.
Andrew: junior folks – it is harder to ramp up on Fedora. With a skilled team they can be productive in a short amount of time.
Robin: tricky situation, working on selling the idea to work on this.
Erin: Q: is the goal testing specific use cases, or the partnership have an outcome that is more than just testing.
Andrew: would like it to be more than just testing. “ergonomics” of the project. Preferences between Fedora urls and ids and how it translates into the pear tree path. Feedback beyond just core developer team.
Patrick: Get code to test, give feedback.
Tim: could we engage with a consulting firm adjacent to our culture to get help.
Rosy: can she pay consultants to do the migration – but would be a couple hundred thousand dollars.
Mark: Lyrasis – heard ‘deploying staff’ on particular projects. Could organizational change interfere with our time frames?
Erin – junior staff could get trained but not an immediate thing – building a deep bench but not possible right away.
David – we have a number of ways we could make this happen. On some level, working with institutions who have use cases. We don’t want to build a bridge to nowhere. We need adoption. Commitment to say we’ll use it.
Sayeed – in that spirit, if, JHU has time to help, it will come in July. That is the good as it will get.
Andrew: aaron birkland’s ocfl client is a critical part of Fedora 6.
Tim: Lyrasis funds – put on projects (catalyst funds). Commitment to Fedora 6 would engender some trust? (Erin: they put out the call in January)
Karen – Lyrasis has a stake in Islandora
Erin – could be a collaboration between Fedora and the board to work on Fedora 6. Would need to justify this. Some avenues if we can create a strong business case and what the investment is.
Mark: also other interests – Fedora 6 on OCFL – strengthens preservation interests.
Erin – look at multiple components in the workflow Can discuss scope.
Evviva – Avalon – not hosted through Lyrasis, not where they are headed.
Este – would like to discuss the idea of a Lyrasis angle in regards to the Islandora Collaboration Group and use cases.
David – does anyone reject to the design plan?
Tim – angst is with testing, not with design plan. We need to involve the community. Needs to hear a commitment to do the work that Andrew needs to get this done.
Maurice – Path to adoption seems to lie with Fedora 3 users.
David – we have Fedora 3 folks who want to test/adopt
Este – still follow the mapping OCFL to LDP
David: Will put a vote out to the list for the design plan
Maurice: didn’t finish the activity streams – call for participation in the groups. Ready to go on some other streams now that we have the Product Position done.
David: clean up product position and put out a call for folks to join the groupsStefano - it would be nice to have. I just started aligning X to the Fedora spec.
Actions
- David Wilcox Rosalyn Metz Robin Lindley Ruggaber review product position doc, incorporate/resolve comments, and share with the rest of the LG.
- David Wilcox Ask for vote on approval to release Fedora API Spec 1.0
- David Wilcox Ask for vote to approve high-level Fedora 6.0 design
- David Wilcox Put out a call for more Vision and Strategy group members