Developers Meeting before OR10 on Mon, July 5, 2010

Logistics

• Location: UNED (the University), Facultad de Humanidades (Room B)
  • Google Map: Edificio de Humanidades - UNED
  • How to Get There: (copied from info provided in above Google Map):
    • From Moncloa Metro stop: 15/20 mins nice walk through the "Parque del Oeste"
    • Metro + bus :
      1. Metro "Moncloa" + Bus: 160, 161 o "A" (all Stops at Avenida de Séneca)
      2. Metro "Príncipe Pío " + Bus 46 (Stop: Senda del Rey - Obispo Trejo, in front of "Facultad de Derecho")

• Time: Meeting is scheduled from 12pm-6pm

• Other Items of Interest:
  • Fedora Committers will be meeting in same building (they are meeting 10am-6pm). No joint meeting planned on Mon, but we will have opportunities to introduce ourselves/socialize both at lunch and at the bar/tapas in evening.
  • Lunch will be provided (1:30pm-2:30pm)
  • We will break to bar & tapas around 6pm
  • Supposedly there are coffee / snacks to be had nearby (and we will take breaks to re-energizer)

• NOTE: This meeting directly conflicts with PASIG. PASIG is 8:30am-5pm on Monday

General Agenda

12-1pm : Intros, Overview & Brief Updates

• Intros & Overview – our meeting "ground rules" ** Changes to agenda? We should not be bound by this proposed agenda – we could always turn it into an 'unconference' of sorts, or reorganize topics.
• Past reflection on 1.6.x releases:
  • What have we learned?
  • What went well, what could we improve?

1:30-2:30pm

• Lunch is provided

2:30-3:15pm : Discussion of Current/Ongoing Projects

• GSoC project updates (Mark Diggory)
• 2.0 integration work (Mark Diggory)
  • Overview of DSpace Addon Module Design and Spring
  • Discussion of Storage Services and Storage Providers as an addition to 1.7
  • Discussion on Solr Integration and Unified Search/Browse as an addition to 1.7
• AIP Backup and Restore / DuraCloud integration project (DuraSpace & MIT & IUPUI) (Tim Donohue)
• 1.7 planning (what's in it?)

~3:15pm - Break

3:30-4:45pm : Discussion of RoadMap for Future

• Fedora integration proposal (Tim Donohue)
• Longer-Term Questions:
  • Source / binary distributions (what should/shouldn't be included?)
  • Should all webapps/UIs remain out of the box? (JSPUI, XMLUI, OAI, SWORD, LNI, SOLR)
  • How do we make upgrades/installs/configuration easier?

~4:45pm - Break

5-6pm : Discussion of Integrations / Communications with others

• SWORD updates (Stuart Lewis? or Richard Jones?)
• Community Advisory Team & DSpace Global Outreach Committee (Val Hollister & Tim Donohue)

~6pm

• Head to bar (with Fedora Committers) for Beer/Sangria/Tapas
  • For the people at the University: you can walk (20 mins) through a nice park (althought it may be too hot at 18.00) or you can take the bus 46, right in front of the Facultad de Derecho, in the same street where is the building you are having your meeting. You end up in Moncloa. Take the street Princesa down, on your right (the same side where the bus left you) and walk to the next corner, after passing through and arch (from the bus stop only 1 min. walking). The bar is "Casa Manolo", and has good tapas and a lot of beer.

The url and the location map: http://11870.com/pro/manolo-1934-bar-restaurante

In any case, the area of Moncloa/Arguelles is full of bars.

Attendance / Availability

• Tim Donohue – available all day.
• Stuart Lewis – available all day.
• Claudia Jürgen - not attending the OR10
• Robin Taylor - not arriving until mid-afternoon.
• Andrea Bollini - arriving at 3pm
• Peter Dietz – available all day.
• Keith Gilbertson – available all day.
• Graham Triggs - available all day.
• Richard Jones - will try to make it to the meeting at some point in the afternoon
• Richard Rodgers – available all day.
• Mark Diggory – available all day.
• Ben Bosman – available in afternoon.
• Brad McLean – will join in afternoon.
• Val Hollister – will join for most/all of day.

Meeting Notes

• 1.6 Discussion - Reflecting on how release went
  • Did we branch too soon – no activity in Trunk besides 1.7 (Stuart)
    • No decision on what we wanted in 1.7? May not have been fault of branching too soon. (Mark D)
  • 1.6 is now "put to bed" – likely no more 1.6.x releases (unless something major comes up)
  • Issue: We didn't test the Handle Server – issue now seems fixed, as we have a valid handle server on http://demo.dspace.org
  • Issue: Too much of a bottleneck for Jeff Trimble as the Documentation Guru. The model is right (to have one or more folks lead the Documentation effort), and we'd like to expand.
    • We should be more proactive to find people like Jeff. We need to enable more people to help with documentation (moving main docs to wiki should help)
  • Demo.dspace.org is 'great' – need to develop even better test practices.
    • Is there a way to encourage testathon interaction (need more feedback?). First one for 1.6.0 was great – been a bit downhill since then. Do repo managers feel it is "too hard"? Do we need rebranding? What's the positive "reward" – how to you incentivize them? (It works well when people are excited about it). A top ten testers – or a drawing based on number of bugs you found? TODO: Talk to Val – does Global Outreach have ideas on incentives??
    • Automated testing from GSoC
      • Integration testing
      • Performance testing is a need (small / medium / large reference datasets – how to build these / IP issues? TODO: Can Repository Managers help us find data to test with? Talk to Global Outreach committee.)

• DSpace Documentation post-1.6
  • Moving main documentation to Confluence wiki (hoping to get more non-techies involved in helping with docs)
  • Snapshot versions a problem for Fedora folks? Talk to them about mgmt of docs in wiki?
    • Export docs to HTML and PDF
  • "Hard Stop" to move just to Confluence based Docs (move back to SVN)
  • How do we make sure all patches have documentation before commits?
    • Move original wiki page (for prototypes) off to docs area and clean up

• Handle Server plugin – difficult to do migrations (have to sync up handle IP change with migration) – blind assumption that it's always running on same server as DSpace (Richard Rodgers)
  • • Should be replaced with a Service call (HTTP REST?) – any processing/slowdown here, likely not?
    • Related to REST GSoC project?
  • Graham Triggs has rewritten code entirely – one handle service that responds for all prefixes
  • Decoupling of handle server is necessary – all in agreement
  • CNRI working on new release of Handle Server 7.0
    • Anything we need to know about on our end
    • Decoupling for a future hosting service (DuraSpace hosting handle server for small sites?)

• GSoC projects
  • No transition from GSoC to codebase
  • Trying a team strategy – are we being successful or not?
    • Projects may need to get off the ground a bit before they can successfully collaborate...
  • JIRA not being used afterall (too much work for students?) – group meeting amongst students difficult to schedule (logistics / attendance difficult)
  • Varying levels in participation amongst the mentors / varying styles in mentoring
  • Overviews of projects
  • Are there ways to create face-to-face opportunities (Open Repositories, DSpace User Group Mtgs, have students present??). Should the money coming back to DuraSpace be funneled back to the Students (fund one to go to a conference)? Traditionally, most GSoC projects give the money to the mentors? That could be an option to get more mentors involved?
  • How to move projects into Trunk easily? Take them as they come.

• TODO: Full meeting on Discussion of Commit Rights & Committer Governance models. Should we still be called "committers", or give out commit rights more liberally?

• Modules
  • Dependencies from Trunk directly to Modules directory
  • Need to make this less "muddy"
  • TODO: Full meeting on Discussion of how we should split these "service boundaries" – so how do we want to manage trunk & modules.

• Discussions of 1.7 release – see notes at DSpace Release 1.7.0 Notes

• TODO: Where do we all stand on asynchronous module releases?? (Mark Diggory)
  • Fedora folks are running into this same problem – how can we work with them or learn from them.

• TODO?: Design competition for the next DSpace Theme! (Stuart Lewis idea)
  • How do we organize such a design competition / DSpace Global Outreach Committee?

• Feedback on "Community Advisory Team" – add more less techie people to Committers Group? Would rather pull them into committers group? How can we all better understand how DGOC works, and how Committers work. How can we merge these, or change the structure (Apache PMC??).