Attendees: Achim, Dan, Jared, Jon, Steve, Wendy; Jay (guest)

• Lawrence sprint: important outcomes
  Report; progress on Codebook View. Kelly is working on summary document.
• Future plans for Codebook View with Lawrence sprint as background
  Work on Codebook View is far enough to proceed with the planned development release of this view in September.
  A very subset of Codebook 2.5 is realized. The majority understanding is that this subset with all related files of DDI 4 will be enough for understanding the purpose of this Functional View by the DDI Codebook community.
• Continuation of the discussion on Dagstuhl workshops.
  List of topics to discuss:
  1. Possible tasks of Moving Forward work
  2. Open topics of Codebook View
  3. Infrastructure Vision
  4. One or two workshops on Moving Forward

  The two Dagstuhl weeks should be used for Moving Forward work.
  Tha majority understanding is that the Codebook View is not a topic for these workshop, see future plans for Codebook View above.
  First week's focus on technical topics: Structured Documentation, Production Framework, Test Cases, OWL/RDF-S, Program Libraries
  Second week's focus on content topics: Data Description, Data Capture, Modeling, Use Cases, Strategic Documents.
  A description of targeted results is available as draft.
  Mostly existing work should be used, refined, and integrated. Start of new work should be minimized. The overall goal is to combine the good parts of the work of the last years in order to achieve two working functional views (basic Data Description and basic Data Capture), a working class library, the resulting bindings and a working production framework.
  The overall goal is to achieve some visible DDI 4 results from the Moving Forward work.
  A preliminary table on participants by topics was agreed. Possible participants will be asked for availability.
  Planning will continue in next AG meeting on the outcome.
  

Attendees: Achim, Dan, Larry, Steve

• Discussion of Dagstuhl workshops: possible topics.

Attendees: Larry, Wendy, Steve, Jared ?

• Clarification of Lawrence Sprint topics.

DDI Moving Forward Advisory Group meeting

12 April 2017

Attendees: Larry, Wendy, Steve, Jared

• Reviewed Sprint invitations.

• Discussed codebook review activities during the sprint, including the need for Sanda to bring a laptop with XML editor installed.

DDI Moving Forward Advisory Group meeting

29 March 2017

Attendees: Larry, Dan, Wendy, Jon, Jared

• Reviewed Sprint invitations. Discussed adding a few other people from the process side (transition to Cloud). These invitees could attend for just the first few days if need be. Jared will distribute invitations.

• Wendy reported on Modeling Team plans to organize a set of virtual meetings this summer to address DDI and GSIM relationships and mapping.

DDI Moving Forward Advisory Group meeting

1 March 2017

Attendees: Larry, Dan, Wendy, Jon, Jared

• Sprint to be held after the IASSIST conference, May 29 - June 2.
  • Reviewed and finalized list of invitees.
  • Wendy will draft goals and deliverables for sprint and send to Jared. Wendy will also create page on wiki.
  • Jared will draft invitation and send to group for review.
  • Once OK, Jared will send invitations.
• Discussed the need to move lion to a new server, as well as the possible options.

DDI Moving Forward Advisory Group meeting

15 February 2017

Attendees: Larry, Wendy, Jared

• Sprint to be held after the IASSIST conference, May 29 - June 2.
  • Topics to discuss
    • Evaluate codebook view to facilitate deposit in an archive.
    • Test with known examples.
    • Identify gaps.
    • Documentation.

    • Include qualitative assessment as part of this review.

  • Tentative list of invitees:
    
    • Norway (Nesstar perspective) - Ornulf
    • World Bank (IHSN/Nesstar perspective) - Oliver
    • ICPSR (archive perspective)
    • CESSDA member (archive perspective) - Hossein Abroshan
    • Canada Nesstar user (Jane or colleague?)
    • GESIS (Wolfgang or colleague?)
    • Steve
    • Achim
    • Modeling group (Oliver, Wendy, Dan G., Dan S., Larry, Jay, Flavio, Jon, Johan, Jeremy)
    • Sprintmaster (Kelly?)
      
      
  Next meeting is on March 1, 2017 (2pm EST).

DDI Moving Forward Advisory Group meeting

18 January 2017

Attendees: Larry, Jared, Wendy, Dan, Achim

• Task description for Drupal programmer and Semantic Web consultant
  • Drupal programming work
    • The Drupal programming work is described in detail with time estimates in three documents on the page of the Cologne working meeting ("Task Specification").
    • Wendy pulls the content together into one document, adds a high-level introduction, and the total time estimates.
    • Jared and Wendy add information on communication way/frequency, contact people.
    • Jared clarifies if this work can be done in house, and in the positive case, if the process of organizing a bid would be necessary.
  • Semantic Web consultant
    • Achim writes together with Wendy and Jon a high-level description mentioning the tasks, the length of the work, the communication way/frequency, contact and reporting people. This document references the existing document on "RDF Work Specification".
    • Jared will forward this to the procurement group at UMich to prepare a bid.
• Michelle pulled back from all DDI responsibilities after changing jobs. Jared clarifies how a new vice-chair of the Scientific Board will be elected.
• Achim will probably drop out for at least 5 weeks beginning in middle of February due to a surgery and related rehabilitation.
• Topics for future meetings
  • Dagstuhl 2016 papers: status, next steps
    • Data manifesto
    • Vision paper
  • Consolidation of DDI-Views, next steps
  • Dagstuhl workshops: further collection of ideas
  
  Next meeting is on Feb 1, 2017.

DDI Moving Forward Advisory Group meeting

21 December 2016

Attendees: Larry, Jared, Wendy, Steve, Achim

• Results from working meeting in Cologne
  The successful meeting clarified open issues for the production framework and described related tasks. The report summaries the different work. The meeting page provides all related documents and information
  Report: https://ddi-alliance.atlassian.net/wiki/download/attachments/63504391/KoelnSprintReport.docx?version=1&modificationDate=1482275816627&cacheVersion=1&api=v2
  Meeting page: https://ddi-alliance.atlassian.net/wiki/pages/viewpage.action?pageId=63504391
  • Task descriptions for Drupal programmer and Semantic Web consultant are done.
    Next steps:
    • review of documents by TC, MT, and AG until Jan 11
    • High-level introduction paragraph will be added
    • Documents will be sent to Executive Board for approval
    • Jared will clarify the process how of organization of getting external help
    • Jared will ask if and how Drupal programming work can be done at ICSPR
• Chairs of AG
  • The availability of the chairs in the next months will be hopefully clearer in the next meeting.
• Consolidation of DDI-Views, integration of work of different groups: plan development
  • There is a general agreement that consolidation and integration work will have high priority in 2017.
    Plan:
    • Achim and Jared will have meetings with each MF working group to review the current state of work
    • A regular meeting of the chairs of all MF working groups including the modeling groups will be organized. The meeting should be also attended by the TC chairs.
    • Based on the outcome of these meetings, AG will make a plan on how to resolve open issues. This plan should be done in close consultation with the Modeling Group, because the model integration will be done by this group.
• Next meeting is on Jan 18, 2017.

DDI Moving Forward Advisory Group meeting

9 November 2016

Attendees: Dan, Larry, Jared, Jon, Wendy, Michelle

• Long term - next year - what products are we hoping to get released - this will help to determine where we need to focus our resources and prioritize our
  • Providing support for development work and project management support
  • Business specs for the production framework - the priorities may fall out from here.
  • Outputs for this coming year:
    • Codebook - which pulls in content pieces
    • RDF binding
  
  
  • RDF - up until now we lacked someone to spearhead this but now we have someone who is interested, has the time and has the knowledge.
    
    • Specify the RDF binding so we can use in the production process
    • definition and implementation that ties into the business needs - particular binding and how to implement it.
    • Would need to clearly specify what we need and desire as outcomes.
    • We would need the ability to say yes or no to the outcomes that Eric would specify - a review committee
    • Use EDDI to produce a clearer requirements for RDF specification
    • Reqts of RDF, types of constraints, → develop the specs for the RDF binding → review of the spec
    • Consensus to get Eric involved → need to work out the details of his involvement.
  • XMI - development ties into the requirement of a Developer
  • Priorities need to stay aligned with the DDI community!
  • RDF/Eric and XMI/developer are top prioirties over a Project manager at the moment
    
    
  • Project management:
    
    • With new tools in place we have a better handle on the overall project management environment
    • Still need to address some of the issues that were identified with the Paper overview of the current project management.
    • we can work through these along with more leadership from the Exec
      
      

• Possible support for Eric for RDF work?
  • Eric propsoed to work with the RDF bidning work
  • interested in working with us based on Daghstul work
  • Need to determine what we -
    • specification of the RDF - what should the target looks like - need to work within constraints of the DDI
    • Achim will create a job description or a task description - how much time/effort/ which will lead to funding to allocate - this will be passed onto Modelling Team for review

• Sprint EDDI - Jared will send out invitations this year, but in the future, the Sprint organizers should do this, unless there is a requirement that the invitation comes from the Director
  • Current invitees are: Achim, Arofan, Jon, Oliver, Olof, Wendy, possibly Kerrin and Matteus
  • Goals/Focus of the sprint are:

    • Identification of all open issues

    • Migration from Bamboo to Jenkins (set-up of a cloud server would be necessary prior to the meeting)

    • Detailed description of open issues if resolution is not possible in the meeting. The description should be the basis for involving more people in this (possibly a paid developer)

    • Resolution of issues if possible in the meeting

    • Knowledge transfer from Olof regarding Drupal

• Logistics:
• Moving this meeting up one hour so that Steve can participate - so 2pm Eastern
• Michelle will send email out to AG members to determine issues/challenges with this meeting time change before overall decision to move ahead with the time shift

TODO:

Michelle will contact individually if work/writing/documents are needed to close out Dagstuhl documentation

DDI Moving Forward Advisory Group meeting

14 September 2016

Attendees: Larry, Wendy, Dan, Jared, Achim, Steve, Michelle

Dagstuhl preparation - week #1 attendance

Elise - via Dan - still up in the air

1. 1. Add Elise to the list - and set as a last minute cancellation
   2. Dan to provide address of organization and email to Achim for Dagstuhl

Currently have 19-21 attendees

Harold - not available - Wendy will keep up relations with Harold as we move along.

Dagstuhl preparation - Week #2 attendance

22 confirmed + Sam and Kate and Oliver = 24 (yes that's really 24 )

updates

1. 1. Chuck cannot make it
   2. Sam Spencer - Steve will send formal invite - unsure but keen - will confirm and may have internal support
   3. Matt Wooler - cannot come suggested Katie McNeill - is a UKdata option - Steve will formally invite
   4. Oliver - still waiting for confirmation
   5. Mari - cannot attend - recommended Katia - who can attend

Preparation calls:

Sept 20 - Week #1

Sept 22 - Week #2

1. 1. May be a follow-up CV call with Wendy

Jared will send out email with call attendees for the 2 Dagstuhl weeks

Goal of the call:

1. 1. Jared - overview of the DDI
   2. Wendy / Achim - experiences of last year and goal of the workshops / what do we want to achieve / outline for their presentation
   3. Documentation - need to send outWendy / Michelle to update participants on the Sprint pageMonday call - Achim, Wendy, Jared, Jon, Michelle - prep for Tuesday call
   • • How will we handel those that cannot make the Tuesday call
   1. Wednesday prep call for the Thursday prep call

2015 Documentation - will be posted on last year's Sprint -

we will need to link it to this year's Sprint

Updates -

modelling team - Sept 23rd deadline to get everything up to TC

1. 1. Data Description group going up to Modelling
   2. Working expediently and getting pushed it through
   3. To have in TC hands by the 23rd of September
   4. Fast review - report card on it - JIRA in place now
   5. Turn around in a week

Codebook group

1. 1. Developed a view and working on adding in Data Description and Methodology
   2. Found some items that the Modelling team need to address - description of security and access - will need to be addressed in a general way

Sprints at EDDI, NADDI, IASSIST?

AG should plan if we need Sprints

Jared - budget aspect and if require - how do we make them happen

If we need Sprints - make room reservation

EDDI - already has the room reservation - if we have meeting - focus on the technical framework -

1. 1. Will be needed
   2. responses back from Review that will need to be addressed
   3. already have tech issues that will require a Sprint

NADDI - make space available - Cancel without penalty - 2 rooms - Passover starts on April 11 - food that is not bread and grain

IASSIST - may be easier - do it the week afterwards - 2 rooms

DDI Moving Forward Advisory Group meeting

3 August 2016

Attendees: Achim, Dan, Jared, Jon, Larry, Steve, Wendy

Status of invitations for Dagstuhl 1st week
It looks like we’ll have enough external experts covering all important areas and groups (CDISC, W3C, GSIM, SDMX, Provenance, RDA). 6 external experts confirmed. We are still waiting for 4 confirmations of external experts. 11 DDI people confirmed. This makes a total of 21. One additional person from UKDS will be invited.

Preparation of Dagstuhl 1^st week: a group (Achim, Jon, Wendy, Steve, Michelle) was meeting twice and will continue to meet. A draft master document is available at: https://docs.google.com/document/d/11HQUpmk7laGkTIUnL0ngg40vSuFsI3lg4nAXRZCCy7M/edit?ts=5799c77e&pref=2&pli=1#.

Documents of Dagstuhl 2015 workshop
Larry will focus on an additional part of the summary document regarding the follow-up work in Moving Forward working groups which was triggered by discussions in Dagstuhl (i.e. design patterns, data description).
Jon will have a look at draft documents and selects documents or parts of them which are suitable for publication on the wiki (5-6 documents).

Status of invitations for Dagstuhl 2nd week
Confirmed people 16; maybe 6; total 21. Two or three additional people will be invited for the topic Funding Proposals.

Preparation of Dagstuhl 2^nd week
Topic groups will be informed why these topics are on the agenda. One person per group will be approached for framing the work while the week. A list of open issues - which can be resolved in Dagstuhl – should be produced. Purpose and list of issues will be distributed to all participants in advance of the meeting. Groups: Complex Data Capture (Barry), Complex Data Description (Steve), Structured Documentation (Jon), Controlled Vocabularies (Sanda), Funding Proposals (Steve, Bill). Steve and Achim will have a separate meeting with the representatives of these groups.

A person for the documenter role for both workshops is searched.

Preparation group at GESIS (Achim, Arofan, Michelle, Wendy). Agenda items?
Postponed.

DDI Moving Forward Advisory Group meeting

20 July 2016

Attendees: Achim, Dan, Steve, Wendy

 Status of Dagstuhl 1st week invitations -- Discussed who can and can’t attend, as well as who has not responded. 

Counting two external now as "No" who didn't respond for several weeks. Confirmed external: 4, external maybe: 1, confirmed DDI: 11, DDI maybe: 2, outstanding: 2, total: 20
3 more external people will be approached.

Status of Dagstuhl 2nd week invitations
Confirmed: 14, maybe: 1, outstanding: 9, total 24.
One more person will be approached for the topic “Funding proposals”.

Managing output documents in both Dagstuhl weeks --  Issue is currently open. Should be discussed in August again.

Status of next development releases – postponed.

DDI Moving Forward Advisory Group meeting

6 July 2016

Attendees: Michelle, Arofan, Steve, Wendy, Achim, Jared, Dan, Larry

 Status of first week invitations -- Discussed who can and can’t attend, as well as who has not responded.  Decided to invite one more external expert: Joshua Mandel (https://dbmi.hms.harvard.edu/person/research-scientists/joshua-mandel)

Question from potential workshop participant -- We received the following question from a potential first week workshop participant: "Issues of interoperability between multiple metadata specifications".  ... wondering whether the intended objective of this item is for instance to map some RDF DDI properties to some other standards and get some concrete specification (or even chunks of ontology) in the end, or only to introduce DDI and the other standards to each other and to start some preliminary exploration.”  Decided to respond that we would like to do both but we are not sure if concrete mapping is really possible.  

Preparing for the first week workshop -- We think it would be good to have a call/document beforehand so external experts are prepared before the workshop.  A small group will form to plan.  Wendy is interested in helping, as is Michelle.  Think Jon might be interested, as well.  Steve can contribute ideas.  Achim will set up the first meeting.

Additional people to invite -- Discussed possible others to invite to the first week workshop.

DDI Moving Forward Advisory Group meeting

22 June 2016

Attendees: Michelle, Jon, Steve, Wendy, Achim

Output of Dagstuhl 2015, documents for participants

Documents should be provided on Dagstuhl 2015 page in Confluence

Documents "Recommendations" and "Design Principles" are ready for distribution. "Recommendations" is already on page, Wendy will make a link to "Design Principles".

A document along the lines of the "Recommendations" would make sense, reporting on finalized or ongoing activities regarding individual topics.

The other documents are often more in a status of minutes than reviewed resulting documents. Even then, some of these documents can be provided. Jon will have a look at the related documents in Google Docs, will make suggestions which documents can be copied to the Dagstuhl 2015 page, and will write an introductory paragraph for this list of documents.

Most of the output documents were at the end of the working week in Dagstuhl in a rough state. We should learn from this experience and prepare this year's Dagstuhl 1st week workshop better to achieve better output documents. The preparation of the workshop and the more structure in the workshop could help all involved people to achieve this.

Status of invitations for Dagstuhl 1st week

Achim will check with Jared, who got invitations and reminders. Achim will follow-up (in addition to Jared) with people where answers are missing. Steve will contact Simon Cox.

Invitation list for Dagstuhl 2nd week

Wendy will contact Taina and Sanda regarding their availability and recommendations for other people for the group on Controlled Vocabularies in Dagstuhl.

The group on funding proposals should consist of people with experiences with funding agencies. The group was narrowed accordingly. It looks like people with EU funding experiences are missing. Conference calls with people with DwB and CESSDA involvement would make sense, especially people from FSD, UKDS, and NSD.

The invitations to people with 1st priority should be sent out. Achim will talk with Jared.

Preparation for Dagstuhl 1st week

postponed to next meeting

DDI Moving Forward Advisory Group meeting

8 June 2016

Attendees: Dan, Steve, Wendy, Jared, Achim

1) Change of time for upcoming meetings -- New time agreed upon for meetings: biweekly on Wednesdays 3pm EDT, 9pm GMT +2, 5am GMT +10 (Thursday).  Jared will send out calendar invite.

2) Chair and vice chair of AG -- Achim and Michelle were elected as the new Chair and Vice-Chair of the Scientific Board.  They will also lead the Moving Forward committee.

3) Adding Jon Johnson to the Moving Forward group -- Agreed he should be added to the group.

4) Review of the Knutholmen meeting, especially in the light of practicability, using the class library for the Codebook Functional View to create both products (binding specifications) and examples (instances, text descriptions).  What should have first priority regarding development releases? -- Plan is that agent view, data description, capture, and statistical classfication will go out in next release (July).  Based on that, the codebook functional view can be built and hopefully released in September.  Thought it would be better to limit number of small ones to pieces that people could be familiar with them.  The codebook would then pull it together.

5) Status of invitations for Dagstuhl 1st week -- Jared will contact Daniella Meeker since Philippe cannot attend.  Jared will send out reminders to non-respondents.  In his follow-up to Jeni, Jared will 'cc' Arofan.  Achim will send the follow-up to Eric.

6) Start of invitations for Dagstuhl 2nd week -- Achim will send an updated list to the group.  We can discuss by e-mail edits/modifications before sending out invitations.

7) Access control issue discussed by Codebook group -- We needed to discuss work plans for the three areas of Codebook not yet covered - Citations, Methodology and Access Control. The first two are addressed by existing groups, but Access Control is not within any current group's remit. We should discuss this topic at the next AG meeting.

DDI Moving Forward Advisory Group meeting

5 May 2016

Attendees: Dan, Steve, Wendy, Jared, Jon, Larry

1) DDI deliverables and timeline – Discussed deliverables and timeline for releasing a product for public use. At present, both final deliverables and timeline are not clear. The group needs to firm these up asap. Will devote time at Norway sprint and IASSIST to figure out.

2) Dagstuhl first week – Reviewed Achim's invitation letter. Looks good. Jared will send out invitations the week of 16 May. Tier one invitees will receive invitation first, with tier 2 invitees emailed as tier ones decline. Invitation letters for Simon Cox, Jenni Tennison, Philippe Rocca-Serra, and Melissa Haendel will indicate that funding can support them. For all others, the letter will indicate that limited funding is available.

Discussed importance of holding a Webinar in advance to inform them about DDI and what we expect from their participation. We should also firm up our own expected tangible deliverables by the time of the workshop.

3) Discussed Achim's "Further Thoughts on the Outcome" document. Suggested adding one more goal: "Know whether what we're doing is compatible with various standards that they have, and whether we have or could have value added for them."

DDI Moving Forward Advisory Group meeting

30 March 2016

Attendees: Dan, Michelle, Steve, Wendy, Achim, Jared, Jon, Larry

1) NADDI Sprint update -- Things seem to be good to go.  Have established the priorities for the week.  Also have come up with approaches of ways to move through a lot of material in a short period of time.  

2) Norway Sprint -- The deposit for the venue and accommodations has been made with the hotel.  Several people have already responded whether they can attend.  There was a question whether Sprintmeisters receive funding support from the Alliance.  They are encouraged to submit funding requests for costs their organizations cannot cover.  Jared will send reminders to all invitees about deadlines for responding and for requesting funding.

3) Dagstuhl --

Reviewed Dagstuhl descriptions that Achim distributed.  Once finalized, these will be posted to the Dagstuhl site.

Reviewed financials related to travel and meetings.  We have $31,000 budgeted for FY17.  With two weeks of meetings at Dagstuhl this year (~20 people per week), along with the Norway Sprint, conservative estimates put us $10-15K over budget.  There are funds potentially available to cover DDI 4.0 related travel, which were donated by the Australian Bureau of Statistics, but we need to clarify first if these are truly unrestricted funds.  Jared will contact Mary to confirm.  Jared will also circulate a summary document about travel and hosting funds to the group by the next meeting.

Discussed whom to invite to the first week of Dagstuhl.  Determined we need to determine: 1) What is the rationale of inviting each person, especially so we/they can know what they can contribute during the week?  2) How can we prepare attendees in advance?  3) What are the deliverables resulting from the meeting?  Determined to discuss with a smaller group (Jon, Michelle, Steve, and Jared expressed interest).  We’ll plan for a discussion on the week of 11 April.

6) Upcoming meeting times -- At next meeting, we should firm up a meeting time for upcoming calls.

DDI Moving Forward Advisory Group meeting

16 March 2016

Attendees: Achim, Steve, Wendy, Jared, Michelle

1. When to hold EDDI (if we hold it at all)?  

Decided to hold Sprint after EDDI (week of December 12th).  The Executive Board will explore if there can be a little more funding, if needed, to help people bridge the weekend between EDDI and the Sprint.

2. Hosting two social dinners at Dagstuhl?

This was a follow-up item from the last meeting.  Total expense for this past dinner was ~$1,100 USD, so two dinners would be around $2,200.  This includes dinner, alcohol and transportation to the dinner.  We’ve budgeted $1,500 USD for this expense, so two dinners would be about $700 over budget.  Because the cost, we will present to the Executive Board for their review and approval.

3. IASSIST

Reviewed location of sprint.  Jared will follow-up with meeting site about how much extra days will cost, as well as location and names of bus stops.

Jared will write draft invite and send to the group for review.  Achim will send Jared sample e-mail template Mary previously sent.  Jared will send out invitations to participants on the 21st.  We’ll ask participants to confirm on the 4th, with financial aid requests submitted by the 15th of April.

4. Dagstuhl

Reviewed participant list, especially priorities on whom to invite.  Discussed inviting people from health standards backgrounds.  Jared will ask George Alter for recommendations on whom to invite.  For controlled vocabularies, will invite Taina and Sanda.

Discussed what we have budgeted for these types of meetings and how many people who might need travel support.  Achim will provide a revised spreadsheet to include a column with estimates on who might need travel support.  Jared will review financial information and provide an estimate on how many people we can invite based on the current budget.

5. Timing of next meeting

For the next meeting, decided to move to 3pm EDT (same time used for this meeting).  

DDI Moving Forward Advisory Group meeting

2 March 2016

Attendees: Steve, Wendy, Achim, Jon, Larry, Jared, Dan, Michelle

Agenda

1. Updates on NADDI Sprint

A location is confirmed: Edmonton Health Academy building, which is next to the Lister Building.  Catering has been approved, so lunch and breaks should be covered.

Attendees include: Wendy, Dan Gillman, Jon, Larry, Arofan.

Cannot attend but would like to help virtually: Jared, Oliver Hopt, Barry.

Wendy will check with Jay to see if he can attend.

2. IASSIST Sprint

Location: Knutholmen proposed (a description was sent by e-mail to the group).  There was consensus that this looks like a good location at a reasonable price compared to Bergen. Jared will move forward with finalizing the arrangements.

Wendy, Achim, Michelle, and Steve are planning who to invite to attend and what will be discussed.

The main topic is the codebook view.  The second topic is methodology.  The goal of the week is to finish a working codebook view, and then have the codebook reviewed from the user’s perspective (i.e., people who are using the old codebook instances).  Those reviewing should prepare in advance to have sample codebooks, especially machine-readable ones.  Also, the codebook view group should provide in advance a list of items that should be in the codebook view.

Reviewed attendance list that Achim distributed.  Jon Johnson will not be able to attend the meeting in Norway, but can help virtually.

3. Dagstuhl

First week will be focused on exchange with external experts and a review of the current status of moving forward work.  

Second week is limited to people from the DDI community and should focus on topics of working groups -- i.e., a more focused working week with deliverables.  Documentation is an essential part of DDI, and reusable documentation will be stressed.  Moving forward people, along with training and marketing, should attend.  In addition, because controlled vocabularies are essential for using DDI 3.2 and since we don’t have enough controlled vocabularies, we should invite people from controlled vocabularies to attend.  Separately, there should be a working group focused on writing funding proposals.  Content wise, the work on data capture (more complex surveys and other data sources) and data description should be continued, with new foci defined.

Everyone agreed that the topics proposed for the second week sound good.  As next steps, Wendy, Achim, Michelle, and Steve will write up description of work and then propose to the group. Attendees will then be invited with an official letter.

4. Dinner at Dagstuhl

Tradition is to hold a social dinner at Dagstuhl, which is funded by the DDI Alliance.  This is a good chance for the Alliance to show support for the many contributions of the members over the past year. Can we hold two sponsored dinners?  Jared will check funding availability, as holding two dinners will double the cost.

5. EDDI Sprint

Should we meet at EDDI for a sprint?  Decided to make the decision at/after the IASSIST sprint.

There might be need for a separate meeting to discuss the production framework.

DDI Moving Forward Advisory Group meeting

 3 February 2016

Attendees: Dan GIllman, Larry, Achim, Jared

1) Steve is requesting that the group move the regular meeting time to 2000CET in future? (two hours earlier than current start time).  Attendees are OK with the move.  Will need to reach out to others if the new time is OK.

2) NADDI sprint (April 11-15) -- planning update

James (NADDI organizer) has secured hotel rooms for the same $99 rate as the NADDI week.  He is working on securing meeting space (main meeting space + one breakout room)

The Alliance has funds remaining in travel/meetings budget to support travel requests.  Dan Gillman will look into attending the NADDI sprint and will need funding to attend.

Jared sent out invitation to TIC and modelers (plus Barry Radler) to see if people can either attend in person or help remotely.  Responses so far: Wendy, Jon, Larry, and Arofan are planning to attend.  Dan Gillman (this may change -- see above), Barry, Oliver, and Jared cannot attend but would like to help virtually.  Discussed how it would be good to have Dan Smith or Jeremy Iverson attend.  Jeremy cannot attend due to other meetings; Dan Gillman will send Dan Smith an e-mail inviting him to attend.

We had talked before about finding a sprint moderator.  Given the small size of the group, don’t think we’ll need one.

Someone will need to prepare in advance for the sprint (agenda -- topics to discuss, what to achieve).  Jared will ask Wendy if she can coordinate preparations.

To dos:

Dan Gillman will check if he can attend NADDI sprint.  If he can, he will need travel support.

Dan Gillman will invite Dan Smith to attend NADDI sprint.

Jared will ask Wendy if she can coordinate sprint preparations.

3) IASSIST sprint planning (May 23-27)

Discussed main focus of the meeting.  Still developing, but discussed not having too big a list is a good idea.  It would be good to concentrate on two substantive areas and something more technical like the model or bindings.  

Main ideas include: 1) Codebook view -- ensure all the dependent things are really working and available.  2) Methodology -- we couldn’t focus on this in the last sprints.  Especially process in the interplay with other components (such as data collection, transformation).

Regarding logistical planning, we are still looking into location and space, as well as identifying a second moderator.

To dos:

Jared will check with Heidi at the NSD about seaside or fjord sprint accomodations.  Achim has done some research on locations and will send the information to Jared.  It’s possible the NSD or university have access to seminar houses or retreats.  We will need lodging for up to 20 people, with meeting space of one main room plus one breakout room.  Additionally, we’ll need Wi-Fi access and meals.

Figure out the second moderator. Achim has contacted Marcel and is awaiting a response.

4) Dagstuhl meeting/sprint

We have two weeks available at Dagstuhl (October 17-21 and 24-28).  Several options: 1) using entire two weeks for DDI development, with one week exploratory (working with other communities) and another week very focused on development.  2) first week as training (not basic, but very focused).  3) Using the week to review funding opportunities and develop proposals.

 5) To discuss at next meeting: Longer term plan for new views and working groups.

DDI Moving Forward Advisory Group meeting

20 January 2015

Attendees: Wendy, Dan, Jon, Jared, Larry, Michelle, Achim

Agenda

1) NADDI sprint planning

There was a feeling in the modeling meeting that there is a sprint needed, especially a consistency review of the modeling.  A lot was done at Copenhagen, but mostly production.  We have a number of things modeled before we used patterns.  We now need consistent rules of using patterns so that when we put something out for people to review -- using collections, relationships, and processing patterns -- we’re using them in a consistent way.

Decided to hold the sprint April 11-15 (NADDI is April 6-8). Number of people to attend estimated at 8-10.

To dos:

Jared to coordinate with James (NADDI organizer) about space (main meeting space + one breakout room) and hotel rooms (try to get the same $99 rate as the NADDI week).

Jared to check on DDI funding for travel requests.  

Jared will send out invitation to TIC and modelers (plus Barry Radler).  Will ask if people can either attend in person or help remotely.

Figure out who can be a moderator for the sprint.  The group will think about this and discuss at the next call.

2) IASSIST sprint planning

A remote location in one of the fjords around Bergen would be ideal. This would be in the sense of a retreat meeting like in Dagstuhl.

Main focus?  To be determined.  In terms of release schedule, it comes as we’re wrapping up and pulling together Q1 2016 release.  We had talked about using it to look at codebook, but also reviewing functional views.  That is, codebook functional views drive the content of this meeting.  For this, we’ll need production and content people.

Decided to hold the sprint May 23-27 (week before IASSIST).  Number of people to attend estimated at 20.

To dos:

Achim to look into space and hotel.  

Figure out who will moderate.  Michelle agreed to be one moderator.  Achim will ask Marcel to be the other moderator.

3) Update of the work plan for the next phase of Functional Views.

The main driver should be codebook functional view.  Almost everyone knows what a codebook is and the purpose.  If this is really working well, that’s a model for all the other functional views.  Then we can say to the community that we can really use the new DDI for these other functional views.  Therefore, the idea to release a limited codebook functional view by this year is a very good idea.

Regarding timing, can we release prior to Dagstuhl sprint (Q3 2016)?  If focus on this during IASSIST, there shouldn’t be a problem.

Wendy will update the release schedule page.

4) Dagstuhl meeting/sprint

We have two weeks available at Dagstuhl (October 17-21 and 24-28).  One option: using entire two weeks for DDI development, with one week exploratory (working with other communities) and another week very focused on development.  The other option: first week as training (not basic, but very focused).  Still exploring options.

Attendees: Larry Hoyle, Jared Lyle, Michelle Edwards, Dan Gillman, Jay Greenfield, Achim Wackerow, Wendy Thomas, Steve McEachern

1. Report on Copenhagen sprint (LINK)

Main outcome: we have a working production line on the build server.  Basically, we can press a button and generate on the basis of the model generate XML and RDF.  This is based in the cloud and not dependent on local hardware, but we can still use this process on local machines.  All related files are provided on the repository.  This is a good thing for testing purposes.

Production workflow:

During the sprint, we focused a lot on design decisions for production framework.  E.g., with respect to functional views, we did clarification and documented simplification that classes in functional views are complete classes with all properties and relationships. Documentation for the view can detail which properties and relationships should be used.  We understand now what we use to build them.  We have a draft of a piece from the technical committee re: standard pieces to put in views, at least those that resolve in instances that are managed or have existence over time.  So we now have standard information about what will be available in views.  

Question: we use a relationship called realizes as a way of implementing a collection.  Does that figure into a view?  Answer: it’s a different thing.  It’s more of a concrete use of an abstract class.  

RDF discussion from Michel Dumontier.  

Clarified some of these issues but also determined we need to talk with more RDF experts.  The main critique of Michele is that the current RDF model does not take most advantage of sophisticated querying and usage of semantic reasoner (tools that do automatic reasoning on the basis of RDF).  This means that the RDF instance should provide a good basis to use these tools.  Michel’s comment is that we don’t take full advantage of these tools.  Achim’s understanding is that we need to do just a few tweaks to take advantage, but needs to first discuss with Thomas.

UML and XMI.  Documented the used UML constructs this can lead to a DDI flavor of XMI.

Update regarding the design of the production workflow.  

Basically, there is more and more of a move in the direction that the XMI of the independent model is understood as the canonical version and including related class level documentation in the XMI.  This can then be used for multiple purposes downstream.  This makes it more independent than the Drupal input tool.  To achieve this, two minor programs have to be created - XSLT to extract reStructuredText and reStructuredText to HTML.  Need some programming capacity for that.

Binding rules

The first step was done.  The binding rules were extracted from the code and documented, both for XML schema and RDF.  This is then the basis to do the next step: formalize these rules in a formal language and use it as the configuration for a transformative program.  The question is which language can be used to formalize the rules.  The other question is the transforming program (currently XSLT is used, but might be easier to use java program).

Design decisions

We are tending more and more to the decision than XMI is the canonical form of the model.  Most design decisions were focused on the view.  We have been harvesting design decisions from earlier sprints and creating a living document of design rules (so things don’t get lost).  Also decided that any decisions done are followed up on in Jira tracking system.

Overall

All participants had the opinion that this was an efficient meeting.  Marcel moderated and did a very good job.  Still important to mention that not all expected results have been achieved and further work is necessary -- either to further refine product or to change production workflow.

2. Report on outcomes from Dagstuhl sprint (Attached, as still in draft)

This document pulled out agreements and decisions that came out of various discussions.  Next step is to push agreements and conclusions to various committees for implementation.  E.g., rewriting of basic design rules so you could actually measure against it for both the modeler’s and TIC review point.  That needs to go back to modeler’s and TIC for review and acceptance, as well as the AG.

Wendy will post to Confluence site in a week.  Please add comments by then.

Next step is to pass along to-dos to different groups (e.g., data capture, data description) and ask for feedback.

3. Proposed new work product: XMI as a Work Product of the Alliance (LINK)

The suggestion is to consider a form of the XMI (platform independent or platform specific) to be something that we want to have as an official work product.  Also, what kind of priority should it be?

As background, XMI is called an XML interchange format.  It’s defined by the same group (OMG) that defined UML.  It’s used to exchange UML models between different tools.  We’re currently using a specific flavor of this (Enterprise Architect’s ).  The Drupal system exports this type of XMI.  We are then transforming to the representations.  The thinking is there are more UML tools than Enterprise Architect..  It would be great to have an XMI flavor that can be consumed by all major UML tools and then people could use UML tools to look at the model or generate java libraries or generate database schemas for relational databases.  This means we wouldn’t need to do this as a task by the DDI Alliance, but the community could do this with the tools.

The work is to find a robust XMI flavor which can do this.  There is a so-called canonical XMI recommendation by the OMG.  We have to test this with different tools.  For example, Eclipse UML tools are nowadays the most popular tools.  

Seems like a good idea.  Can think of other uses -- e.g., someone could write code to investigate the model or do their own specific diagrams out of the model.  All this could be done if you have things that are machine harvestable.

A concern about one aspect of this: it might be possible that some tools might render XMI we produce in a funky way and cause some level of misinterpretation.  Response: the XMI remakes a distinction between the formal definition of the model and the diagrams (visual representation).  Currently we are concerned with just the formal definition.  If people import into UML tool, the formal definition cannot be misinterpreted.

4. Updating the work program and timeline for DDI 4.0 development

Will discuss next time (January).

5. Coordinating meeting times

Wendy will post existing meeting times for all groups.  We can then review a good time for this group to meet. One additional comment: for the planned overview of conference call times it would be very helpful if group chairs of all these different meetings can provide agendas two days in advance of meetings. This way people can prepare for the meeting and meetings can be more focused.

AG Meeting November 11, 2015

Present: Michelle Edwards, Jon Johnson, Jared Lyle, Wendy Thomas, Mary Vardigan, Joachim Wackerow

Update on Bitbucket

TC is finalizing the structure of the overall repository. This will reflect the production process. Specifications will be transferred from source forge soon. There will be development space as well – it works like Github. 

2. Update on Bamboo

It is a continuous integration platform that allows us to use the cloud for our work. The Alliance now has a license for this product. Franck Cotton can be a contact for this as he uses AMPED.

3. Recommendations from Dagstuhl 2015

Wendy and Jon pulled together some recommendations. They went through the multiple documents created at Dagstuhl and surfaced the recommendations that the groups discussed. 

We need a couple of iterations on this to have agreement and people from Dagstuhl should have a chance to look it over. We don't want to reopen the discussions that we had at Dagstuhl, though. Recommendations are by assent rather than consent. This is an initial overview of the main recommendations coming out of the meeting.

We should be very transparent on this and everyone from Dagstuhl should feel comfortable with this. The AG needs time to read this and comment on it. 

There should be a report on the workshop. What are the results and recommendations and reasoning behind this? The second step is which recommendations we want to realize and for what reason. Then we need to determine how to implement the recommendations we have accepted. 

We can call the original documents reference or background documents so that the expectation is that these are not finalized publications. We could go back to the authors and ask for summaries, but that would be difficult. 

We also need a decisions document. Depending on the area, those decision may need to be made by different groups – for example, one of the content areas or the process model or production. 

This process needs to be communicated to all people involved. 

An example is the Design Principles. This should be reviewed by the modelers, then the TC, and then if there are disagreements the AG will need to make decisions. And ultimately everyone working on the model should know the principles. 

A document about recommendations and decisions can be sent to the Dagstuhl Institute as results of the workshop.

We might send the recommendations to all Dagstuhl participants to ask about accuracy and whether anything is missing. The final recommendations can be distributed after that along with how decisions will be made. 

How can we involve people like the external experts in the DDI work? We could ask them for advice. 

We should all try to share our thoughts on the recommendations via Google Docs. December 9 is the next possible meeting time. 

October 28, 2015

Present: Dan Gillman, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan

The group discussed the recently concluded Dagstuhl workshop that was a combined Sprint and DDI4 Review.

Wendy and Jon have started a summary of the documents produced at Dagstuhl. Dan will be editing the document he wrote at Dagstuhl and will turn to this in the next two weeks.

There are recommendations in the papers that we need to pull out and then analyze and approve. Then we need to decide on what we will implement and change based on the recommendations. Wendy and Jon will give assignments of what to look at.

There was a discussion to analyze instances and determine which elements are being used. This could lead to defining views. Another approach to simplifying is to identify places where there is more than one way to do things.

We want to establish a working procedure that dictates how we are doing things. This is our opportunity to make changes that we carry forth. The Design Principles may have broader application than just DDI4.

Do we have consensus among ourselves about what the experts were telling us and is there anywhere we need to strongly change direction?

Style of RDF – where you are making assertions about things rather than about data. We may not be leveraging all of the capabilities of RDF. However, it is not clear what an ontology is in the RDF discussions. To make this work, you need to have both data and metadata in RDF.

Issue of domains. Are we trying to model the domain or be broader? In some discussions our domain is "data".

Simplifying the number of elements through an analysis of existing use cases

Using Ontology Design Patterns in our modeling – we need to understand this better but do need a prescribed pattern that has been modeled and then apply it whenever we need it.

Creating reference implementations and test suites for the new spec. All agreed with this.

Creating primers for the DDI specs so they are more accessible. All agreed with this.

Collaborating on grants with the external experts

AG Meeting October 7, 2015

Present: Arofan Gregory, Michelle Edwards, Wendy Thomas, Larry Hoyle, Dan Gillman, Jon Johnson, Achim Wackerow

The group discussed the agenda for the Sprint/Review week in Dagstuhl.

 No minutes provided

September 2, 2015

Present: Adam, Dan Gillman, Larry Hoyle, Jon Johnson, Jared Lyle, Steve McEachern, Mary Vardigan

The group reviewed the evolving agenda for the Dagstuhl workshop.

August 9, 2015

Present: Larry, Dan, Steve, Achim, Michelle, Jon, Wendy

Dagstuhl Sprint

We will need special roles for this meeting as the sprint is more complicated because of the review. Michelle will moderate the meeting, Marcel will help to coordinate, Jon will oversee the outputs and deliverables. We will need a facilitator's meeting every day to make sure everything is covered.

We should identify materials to read in advance. We should describe them and their state of maturity. 

We might divide the meeting up like this:

Day 1: Introductions, background, questions to all

--A short introduction about what is DDI with history, Codebook and Lifecycle, etc.

--What is Moving Forward and the high-level goals with the status of the work and the process as well as Moving Forward at this workshop and the goals with the review and the sprint (data description and methodology) as well as the architecture.

--Review criteria base on Design Principles with interoperability, etc.

--Short presentations (five to ten minutes) from external participants on related work that has connections with Moving Forward; architecture and model, how are they produced (automatic?), how RDF and OWL work with it, how are external vocabularies integrated? Data Description is an important area for Moving Forward so we would like presentations from other standards on this. These could be sent prior to the workshop.

Should all external presentations be one after the other or should there be time for discussion? There will be 7 or more of these. 

There will be two content groups – Methodology and Data Description. The external people should sit in on these groups also as they do their reviews. 

Three stages: (1) Presentations, (2) Review and capturing all the comments that go into draft documents, and (3) Continuation of work. 

The overall goals in inviting the external people to the sprint is to review the intensive effort going into a sustainable DDI4 so we know we are on the right path and are not building in isolation; exchange with the other standards, which may be used in parallel – most standards are "islands" now on their own; the OBO Foundry thought about high-level rules to harmonize ontologies and maybe we can learn from them; forge better connections to other standards and use these relationships in the future. Related to this, it is good to have others become more aware of DDI. 

If these are our goals, we need to integrate them throughout the week. We can have multiple breakout groups with known goals and outputs. We need to guard against scope creep also. We need very concrete comments from them and need to elicit how to solve common problems. Some people are research-oriented, some are consultants, some are technologists, etc. There will be a mix as usual that we have to manage. 

The Methodology Group will be looking at use cases. It started out more based on surveys but with now other data sources and more methodologies are coming into play. Methodology will draw on Process and Provenance models as well and will rely on interoperability. 

The Process model will be another point of intersection with others. This will exist in other standards as well. 

We should try to collect some questions to guide the review. The Production Framework, syntax representation, models – there are lot of detailed questions we can ask on this. A second area is how to integrate and make use of other standards and how can this be modeled in a model and realized in RDF OWL? The whole idea of the semantic web is to make use of other vocabularies. 

The Working Groups related to the review  – Modeling, TC, Data Description, Methodology, Process – can come up with questions. We can start a shared document on this to capture questions. Our questions can be the structure or outline of a resulting document. 

We need to understand the external people better also. They could point us toward a small number of their documents that encapsulate where they are coming from. We can't afford to spend three days getting to know each other. We will end up with a page with links to all of this. The external people should think of their own questions. 

One of the things that didn't work well last time was that the rest of the people not engaged in citation didn't learn a lot from the external people. 

We may be able to get some external funding for this group as it is similar to domain repositories. 

We will have the possibility of evening sessions again. Having something on Qualitative one evening would be good. 

Release Plan

We need to revisit this and see what could be ready for a January release. We will put an Under Revision note on the page. After Dagstuhl and EDDI we will have a more detailed picture of where we are and what the possibilities are. 

Advisory Group Meeting July 22, 2015

Present: Adam Brown, Michelle Edwards, Dan Gillman, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan, Joachim Wackerow

Dagstuhl Participants

The group discussed the status of invitations for the combined Dagstuhl Metadata Workshop (to review the model-based DDI) and Sprint, to take place in October. Overall there are 18 confirmed people and 7 free spaces, waiting for 2 responses, with 5 external people confirmed and 3 external people needing funding. Suggestions for new invitations were put forward. We need replacements for the health sciences, openEHR, Semantic Web, and CDISC slots and made a decision to invite five more people.

In terms of special roles, Michelle will be the main moderator of the Dagstuhl meeting and will participate in AG meetings now to prepare for this. Others like Marcel can pull groups together (especially the external participants with Achim and Arofan) and Jon can be a coordinator and work on resulting documents. Embedding people with the groups is a good idea because otherwise it is difficult to go between groups. In the plenary it is good to have one moderator.

In terms of preparation of a draft agenda, we need to get this out soon as we have external participants and may want some of them to present. A small group should prepare a draft agenda and put it out to the others in the group. We might also collect a list of questions for the external participants as well as which areas they should focus their reviews on, which materials they should get, etc. Wendy, Achim, Steve, and Michelle will work on this.

EDDI Sprint

This meeting should be the week before EDDI – November 23-27, 2015 – in Copenhagen. Eleven people have been identified to participate. We might also involve students. We should think about making a call for a person like this in the future. This year, the EDDI tutorials are the following Tuesday and the conference starts on Wednesday. There is a Semantic Web conference on December 8 and 9. This week of the sprint is Thanksgiving week in the US. Invitations should go out shortly for this meeting. Achim will take care of reserving the hotel and venue for the meeting.

AAPOR

AAPOR has a transparency pledge to sign on to a program to describe the data according to specific guidelines. This looks very close to the DDI elements and we should follow up on this. There is an AAPOR meeting in Austin in May 2016.

DDI Advisory Group Meeting July 8, 2015

Present: Dan Gillman, Larry Hoyle, Steve McEachern, Mary Vardigan 

Review of Dagstuhl participants

Identification relating to DDI 2.5 and DDI 4.0 

This item will be revisited when more people are on the call. With 2 you can use a UUID as the ID, so this may provide some kind of bridge. When you get a long string like that, it is hard to look through an XML instance. These IDs are not human-readable.

 Standardizing codelists

The real issue is does DDI want to get involved in standardizing code lists? There are some coordination efforts along with the UNECE Standards Committee on Modernization. If there were code lists available ahead of time, it would help in harmonization. There are lists of geography, education, occupations, etc. As this gets into cross-national studies, there is quite a bit of work that can go into this. You need to know how to allocate numbers across categories when you map.

Under SDMX is a very prescriptive setup – they have a technical standard that describes the infrastructure for how SDMX is managed and messages are sent and also the content-oriented guidelines, a set of resources, some of which are considered standard, that were used in the construction of messages for exchanging data. This includes cross-domain code lists that are in the SDMX registry. If you are transferring data and you want people to understand the meanings of codes in the various columns, you need a code list. Some lists are already managed and maintained as standards themselves. But there are many that are used and built ad hoc and could be standardized.

The question from the meeting that Dan attended today was does the Modernization committee want to take over the issue of standardizing code lists? The code lists are useful beyond SDMX.

For the DDI there are questions. The Alliance could say it wants to work with the UNECE on this standardization. It could be part of DDI4. Or it could be moved to the Scientific Committee that forms a subgroup to look at the issue.

If there were a place you could go where someone had already encoded a list in DDI and had given it an identifier so that it could be used by reference, it would make life easier for researchers (this could be done with DDI3 as well as 4).

The most important thing for interoperability is that you are pointing to the resource you are using and the terms and definitions are correct. Then there is no excuse for a misinterpretation.

The DDI community could go a long way towards pushing the notion that we want to use code lists that have been previously defined and held in a common place.

First we need to decide is this something worthwhile doing and then if it is, to what extent do we want to take it on and invest in it?

This is a great argument for reuse and the approach of DDI, especially DDI 3. But this could be used in 2.5 as well.

POST MEETING NOTE (wlt): I think the primary issue for DDI is that we want the controlled vocabularies expressed in something that can be easily used by DDI. We currently have a DDI Genericode format and are working on another DDI format. All of these formats are available for use by DDI-C 2.5, DDI 3.x and presumably by 4.x etc. I guess my question is UNECE focusing on structure or content? We clearly have interest in the provision of any standardized controlled vocabularies coming out of the statistical world. Our primary goal as a community is to have them jointly published in a format that is understandable by DDI for purposes of validation and general reuse.

AG Meeting June 24, 2015

Present:  Adam Brown, Dan Gillman, Wendy Thomas, Mary Vardigan, Joachim Wackerow

1. Simple Codebook

At the meeting of the Simple Codebook Subgroup today, the group uncovered some issues the AG may need to consider. The Codebook in DDI-4 will have to look different than DDI-2.5, because it will be rendered in UML and contain classes and attributes that differ from DDI-2 constructs, yet we will have to produce a map that allows for translation between the two. The Simple Codebook Subgroup has begun the work to produce this map. 

Given this, we have several issues that are AG and Scientific Board may have to address:

We think the right approach is to freeze DDI-2 (2.5 is the last version) and manage all further development of Codebook within the framework of DDI-4

Create a mapping from version 1 of the Codebook view in DDI-4 back and forth to DDI-2.5

Use the mapping to allow DDI-2.5 compatible tools to work with Codebook view in DDI-4

Developing countries, the World Bank, and IHSN (among others) might be materially affected by this migration to DDI-4. All efforts to make this transition as smooth as possible should be carefully planned out. In particular, we need to make sure all tools developed for DDI-2 continue to work against the Codebook view.

We don't want the infrastructure of tools to die just because we are managing things in a different way. However, with DDI4 you will need to manage your identifiers. DDI Codebook has been seen as easier to implement because there is no need to manage identifiers in the same way. But the proposal isn't forcing you to use the identifiers in DDI4 at all. A view doesn't force you to use any particular infrastructure in DDI4.

The same mapping issue exists for the DDI Lifecycle branch.

The compelling reason to freeze 2.5 is to maintain everything in one structure. As we move forward, we will be managing everything in the DDI4 model and anything we need to produce to be equivalent to 3.x or 2.x we will be able to produce. But will all development happen in the framework of 4? This is not yet a given.

If there is a complex structure around IDs in 4, you write out what you need and you don't include the other things.

The point of having the mapping is to specify attributes in 4 that are equivalent to, say, 2.5. To make sense of that, there is a binding of attributes. You are ignoring the class structure in 4 to map to 2.5.

This is simply a proposal and doesn't advocate for freezing 2.5 right now. The AG cannot decide this and will need broader input. The AG is about 4 and can't make decisions about 2.5 or 3.

DDI Codebook is always backwards-compatible but can have new things added.

This new direction will be raised with the TC and then the discussion can come back to this committee.

2. Dagstuhl participants

Achim Wackerow provided a spreadsheet with a list of potential participants. The idea is to invite experts to review the progress that DDI4 has made. This review would take place in addition to the sprint activities. We should have a specified set of slots for this with a set time to respond. With one from each special field, there would be 9 people. We could start to invite these 9 people if people agree. Having one third external experts in the meeting would be OK (8 people). The DDI people should be reviewed as well in terms of the topics to be discussed. Funding is also an issue. GESIS might support the travel of three people.

In terms of topics for Dagstuhl, Data Description is still a moving target and Methodology is critical as well and we have some momentum after Minneapolis. Review of the content in Drupal in terms of consistency is also a priority and documentation is also important. However, we shouldn't focus on the production framework at this meeting as people would be spread too thin. Instrument is another possibility. Achim will fill in the spreadsheet to determine who might be available for the specific topics we are interested in.

Between now and October, at least in theory a lot of the work on the production framework could be done offline. The issue is when people are available and when they have time. Production framework people can work at EDDI also.

This means that there will likely not be any releases till EDDI. We still have to get the bindings out. There have been 70 responses on Release 1, which was a good response. Many of the responses have to do with the documentation.

A group of eight or so people will work on the production framework at EDDI.

Action: Achim will send around the edited spreadsheet and then we can make invitations.

3. Timeline for releases

The timeline is currently listed on the main page of the Confluence site:

 Project Management

4. Project manager

Action: Mary will follow up with people who have indicated interest and will get a written statement from them and make sure they have approval from management.

DDI Advisory Group Meeting June 10, 2015

Present: Adam Brown, Dan Gillman, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan, Joachim Wackerow

1. Dagstuhl sprint/review
There has been a plan to have an external review of current work: the model-driven approach and production framework, content of various areas/packages. We would invite external experts to help us with this review. This would be an internal review of current work according to high-level goals and design principles, review of the goals and design principles themselves. The goal is that DDI can communicate in future with other standards. The standards can't be "islands" and will need to interact with others in a network of standards.

These experts are suggested:

Model-driven approach

·           Model-driven approach. Dan Gillman mentioned:

o   Cory Casanave (OMG, http://corycasanave.modeldriven.org/about/)

o   Elisa Kendall (OMG, http://thematix.com/elisa-kendall/)

·           HL7. Jay mentioned:

o   Charlie Mead (http://www.octoconsulting.com/about-us/news/2012/dr-charlie-mead-appointed-director-of-health-it-octo-consulting-group, tutorial on “The HL7 Enterprise Architecture Framework”: http://hssp.wikispaces.com/file/view/AtlantaTutorial-MEAD-SAEAF-Intro.pdf)

·           Other standards with model-driven approaches

o   UN/CEFACT (Arofan mentioned two people)

o   GML (https://www.seegrid.csiro.au/wiki/AppSchemas/HollowWorld, a former project at CSIRO)

GSIM

Exchange across standards

·           RDA – Research Data Alliance. Speaker at IASSIST:

o   Gary Berg-Cross (http://ontolog.cim3.net/cgi-bin/wiki.pl?GaryBergCross)

·           Open Biological and Biomedical Ontologies (http://www.obofoundry.org/)

o   Jay knows people from this group

Other standards

·           CDISC

o   Sam Hume (participated in workshop at Dagstuhl 2014, more content-oriented)

·           openEHR

·           SDMX

            CASRAI (research administration)

Linked Data community

Both contributed to Disco:

o   Benjamin Zapilko (GESIS)

o   Benedikt Kämpgen (FZI Research Center for Information Technology at the Karlsruhe Institute of Technology)

      Michel Dumontier (LOD for Life Sciences)

Ideally we could have 6-8 external experts but not more. This requires work from us to guide and prepare them and introduce them to our work and what they should do for us. We should have a detailed idea of what other content working groups would be possible and what the agenda for the working groups are. We should continue our development work while having the review. This would be similar to last year's sprint with enhanced data citation. It would be different regarding the working groups as it would be necessary to have a more defined agenda for the WGs. A little more structure than a normal sprint will be required. It will also need to be output-oriented with papers, rules, use cases, best practices, etc.  A formal critique would be good along with mapping across standards for common information and deciding where to interact. DDI could be the "glue" between standards. We will not be doing as much production work at this sprint.

DDI should be an implementation of GSIM in terms of methodology and specific process and it would be useful to show that DDI takes the GSIM ideas and expands them in specific areas. Maybe we should have specific content groups that are related to this effort at the sprint. We pay attention to the lifecycle but we don't have aspects of the lifecycle designed well yet.

We need a detailed picture of what we want to achieve and a document on this that we might send to possible invitees. We should have a list of people to talk to and which working groups would make sense in parallel and which DDI people to invite. People who have contacts should get in touch with potential participants and they will need to understand that the participants will not be paid. We should move quickly on this. Will we need a facilitator for this? Some people will need funding, others not. How many resources do we have? How much do we want to invest in this. Mary will start document and Achim will edit it. Achim will start a spreadsheet of possible participants. A discussion of participants will be on the agenda for the next meeting. This group should start thinking about people to consider.

3. New working groups

AG Meeting May 20, 2015

Present: Arofan Gregory, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan

Extending the Review Period for DDI4

A reviewer examining DDI4 asked that the review period be extended to June 14. However, we have frozen Drupal during the review and we have a sprint coming up for which we will need Drupal. So we will need to "unfreeze" Drupal now, but going forward, as a normal part of the review, we will publish the model and the comments will need to be on that package and not on the contents of Drupal. We need to be very clear about what the canonical model for review is.

Preparation for Sprint

Wendy has made a lot of good additions to the Confluence site to enable people at the sprint to work efficiently. Steve and Wendy and Anita have done a lot of planning in advance to determine how best to use people in the sprint according to their strengths and interests. The production team will work all week and the other teams will work on specific days with people going between groups.

We have the prescriptive and historical process models spec'ed out but not fleshed out fully. This would be a good piece of work for a few people to work on during the week. There should be a call with Jannik at some point on versioning in the modeling team. Wendy has a student who will be taking notes. There is a concern that we may not accomplish everything on the agenda.

We should circulate something for people arriving early to get together.

Naming DDI4

The Marketing Group discussed this and decided that calling it DDI (with no qualifiers) was the best strategy. This version is intended to be an umbrella for the standards to bring them together in a more integrated way so it is best not to give it a name that could be more confusing. We are moving to a strategy of task-driven views so this is a new approach.

Things we version may be views rather than anything further down in the model in the library. For marketing reasons we may want to package a set of views but we don't want to be forced to do versioning because of dependencies. People will cite the views and this becomes the basis of interoperability.

Advisory Group Meeting May 06, 2015

Present:  Adam Brown, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan

1.       Minneapolis Sprint – Agenda and Facilitator/Leader

Local arrangements are well organized. Wendy has started adding links to the site for the sprint. Because Monday is a federal holiday, there will be some notes about that.

The TC and the Modelers will have the right technical people there to work on the finalized bindings and Drupal items. We need to make sure time is set aside for that. We can then release the bindings in June/July. People need to focus on the style of the RDF and XML for the bindings and how well they reflect the model. The Q2 package will likely come later. We should put the sequence of releases on the next AG agenda.

Wendy has gotten a few people on the Methodology side and we need to be clear what we want from them.

We don't have a sprint leader yet, which is very important. The group identified a possible leader and will contact that person. Wendy and Steve will work with this person and liaise with the technical group. This small group will prepare the agenda.

Sprint notes can be taken in Confluence.

2.       Invitees for the Dagstuhl Sprint

This goes back to the release schedule and where we will be at that point. In general, the idea of having outside experts evaluate the DDI release and having people from other metadata standards interact is good. CSPA may figure into this as well. We will take this up in the next meeting when Achim can attend.

3.       Coordination with CSPA project

The idea of more formal collaboration between CSPA and DDI came up last week and in general there seems to be support for this. So far, Flavio has had to do all the coordination between the two groups. The CSPA Logical Information Model (LIM) will straddle the conceptual GSIM model and the implementation standards models (DDI and SDMX). The CSPA LIM group will work stage by stage on essentially a more detailed version of GSIM. They need a UML model. We need to understand their timeframes and try to collaborate in a proper partnership where DDI gets something out of it. The Executive Board needs to take a role in approaching the CSPA project about a partnership. That would be the first step. We should discuss this at the members meeting as well.

4. Licensing and Copyright

The TC is finalizing a document about this to reflect the changes in the process with DDI-LMD. There are options to review and the goal is to get this to the AG, the EB, and the SB. This will be on the next meeting agenda.

Advisory Group Meeting April 22, 2015

Present:  Adam Brown, Dan Gillman, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan, Joachim Wackerow

Release

Most everything is in place and the plan is to release tomorrow after the TC meeting. We haven't yet set up a Town Hall meeting but the idea is still in play.

Intellectual Property

There was a discussion of what the most appropriate license for DDI-LMD products is and the current status of the ongoing discussion. For the Development Releases, we are using the LGPL license. Previously the specification was separate from its documentation so we gave them separate licenses. Now the package is integrated. There are some limitations of the LGPL so the TC continues to look at other options. Each of the options has issues. This will come back to the AG and then on to the Executive Board.

 Simple Codebook

A point was raised in the Simple Codebook Team meeting that there are DDI-C elements in the Methodology section not yet covered in DDI-LMD. Some of the issues not covered are actually better described as being part of Fieldwork. Fieldwork is the implementation of the design and the post-collection activities. Collection Situation, Actions to Minimize Losses, and Response Rate are examples. Some of this relates to quality control. We need a working group to cover this area. We should note that we only have a Development Draft so far and we still have options. We don't want to lose these items, so we should create a bucket for this in DDI4. This is a big gap in our work program that we need to think about going forward. Some of this will be addressed in the upcoming Sprint.

 It was also pointed out that we should rethink the DDI-C approach to Methodology given more recent developments in this space like Total Survey Error. Some of the existing elements are dated and artifacts of their time and should probably be deprecated. There have been a lot of advancements since DDI Codebook was released.

The Survey Design and Implementation group has added content in this area that will be part of DDI 3.3.

One of the issues with DDI Codebook is that it is basically backward looking and is not generally used to drive process. Could there be a core description and a means of typing it – a high level broken out into its constituent parts? This could be built on. Codebook currently gives a very high-level description.

Should we still be supporting the "ugly" parts of Codebook? Can we come up with better terminology? We don't want to replicate content that shouldn't be replicated.

Another issue is that there are times when things don't map cleanly from Codebook to Lifecycle. These situations are difficult.

People using DDI Codebook may want to continue to see the same terms, but we could publish a mapping to handle this. Some people get around this by changing the terms in the interface rather than in the actual model.

We can try a mapping and come out with a summary for other groups doing this.

DDI Codebook has a "wrapper" and we need to think about that vis-a-vis DDI4 as well. We have views for different parts of the overall lifecycle, but how will people link together what they have done using different views? This is an important question that relates to interoperability. The kind of flexibility that we are providing makes this more difficult. What does it mean to conform to a standard and what does it mean to write out a conforming set of elements and be able to read something that someone else has written? Your definition of what you want to read may not comply with what someone else wants to write. We need to be careful about how each implementation tells the world what they have done. If you don't know who you will be sharing with, this becomes an n-squared problem.

This needs to be on the agenda for an upcoming sprint.

We should provide a migration path from DDI Codebook to a new Codebook View in the new DDI. This should be addressed by additional migration documents.

 Methodologists for Minneapolis Sprint

The group discussed people to invite to cover the Methodology area and whether we should have a telecon during the meeting to pull in some people who cannot travel to the sprint. We have people from Minnesota with expertise in harmonization and confidentiality. We need a good use case to work through. We need a way of automatically determining what some group is writing out to be able to understand what they are giving us. The first goal is to make sure that people really understand the issue.

There is the issue of passing metadata, and someone on confidentiality would be good here. But we do need someone from the design area as well. And what about design of the analysis, which should happen in advance?

Planning the Dagstuhl Sprint in October

At Dagstuhl, we should try to review the work we have done so far, using comments from the public, our own views looking at the high-level goals and design principles, and experts from other standards. The Sprint last year with data citation experts was a good model to use. We will get feedback on what we are going and will make contact with the other standards to make sure that overlapping standards are talking to each other.

Which kind of standards are suitable for this? ICPSR is becoming involved in the bioCADDIE effort to create a discovery index across health and life sciences data, so this should yield some ideas for contributors. The Research Data Alliance may be another good group to tap. Open biological and biomedical ontologies are interesting developments to create reference ontologies. This type of thinking is good. Standards with a model-driven approach would be good to invite – EBXML and GML are examples.

This review (involving five to ten new people) would take place in parallel with the development of new content.

Questions:

How might this affect getting coverage work done as this will limit the number of spaces for Dagstuhl?

At the point of Dagstuhl will we have enough done so that we can review something? If not, we could do this a year later, but that seems too late.

We have two weeks set aside at Dagstuhl and need to determine how best to use them. Achim will explore funding possibilities at GESIS. If the Alliance pays travel for some people, GESIS may be able to pay as well.

We need to write up something for the webpage on this.

Ongoing role of the AG

Are we providing the kind of value anticipated? An advantage of having this group is that it keeps the TC from being in the position of being accused of running the show. The AG has an important role as it is a mix of people with different perspectives. The focus on more high-level issues is critical. The TC needs to focus on getting things out and can't always focus on the higher level things. The AG is an important link in the communication chain because an update from the AG is a standing item on the agenda for the Executive Board.

Attendees: Wendy, Arofan, Steve, Dan, Larry (Mary and Achim had conflicts, Adam experienced connectivity problems)   1.       Status of Q1 release

More worried now than awhile ago. Documentation seems good but some of the modeling stuff is starting to cause problems. There is more consistency checking than we were aware. If Oliver and Achim can work out the XML Schema issues we should be OK. Primarily we need to get the model in shape and the bindings working. Absence of project management is not helping. We have parsing examples but there were issues around the packages. Waiting to hear back from Achim. We have no solutions for the RDF documents (merging the documentation in) but we have ideas. We have to answer how do we document in a way that is useful to this community. We need a clue as to how we intend to do it. ACTION: Arofan is checking with Franck Cotton regarding a useful structure for RDF documentation. We can also raise this question as part of the review.

We need a clear list of specific problems and whether or not they are show stoppers.

2.       Minneapolis Sprint arrangements

Monday (Memorial Day) will require special arrangments. These will be posted on the Sprint page on the confluence site

Link on IASSIST Conference Site - ACTION: Wendy will notify them to use the following link https://dditools.atlassian.net/wiki/display/DDI4/Minneapolis+Sprint%2C+May+2015

Provide local Food/Drink options for lunch and dinner. Anders Swendsrud (MPC staff) will be available for notes

3.       Future Sprints – as per Mary's email request

Possibility of a yearly TC meeting post Moving Forward Project On going meeting prior to IASSIST as part of the Alliance working process to support TC and working groups as needed

4.       Collaboration tools

Achim is merging documents from TC and what the AG received last meeting Access will be provided to all Sprint participants

5.       Feedback from consultation on release 3 priorities (see 2 pieces of feedback attached below). Do we agree?

INCLUDE: Citation, Methodology, Data Management, Study inception (from Vancouver Sprint)

PROCESS CONCERN: Coverage of 3.2 materials is required. Who is checking for this? Jeremy and Dan would be good for this. During the review ask them to focus on this. Wendy talk to them about this.

Still need clarity on simple vs. complex. These concerns should be addressed and documented

6.       NSI Information Architecture sprint

I was going to update on what I had heard regarding progress. Which isn't a lot other than that there is a lot of discussion around the use of fragments to support the CSPA services (nothing new, myself and Arofan have both discussed that we them before) and what it is they need to support their use cases.

It seems unlikely that anything they come up with will feed into directly into 4, as they seem to want solutions now which are likely to be based on 3.2. As ever we need to keep apprised of their issues and support them to ensure they don't get frustrated and start casting around elsewhere.

If we want to support HLG there is a job of work to be done. This is broader DDI and needs to be brought back to Mary. Arofan is sending paper around and this should be on our next agenda.

7.  IASSIST session has been accepted. Issues of concern noted in #5 should be touched on during this presentation

DDI Moving Forward Advisory Group Meeting
February 11, 2015

Minutes

Present: Dan Gillman, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan, Joachim Wackerow

I. Sprint participation

We are currently lacking in Methodology expertise for the sprint. We need to discuss some specific methodology that some people really know. Most people that are involved know the ideas at a more general level but no one is an expert in one area or another. We need to scope this and figure out how to organize this information. Can we achieve what we want to in Minneapolis with the other groups meeting as well? We will have to prioritize which things are in the upcoming release and which are not.

Wendy can provide a note-taker for the meeting. Having a report of the whole week is important as well. There was no report from London.

We should identify a core set of things that we want to accomplish at the meeting.

Methodology – Three days. This has been pushed to the background several times. We have the details on Sampling and Weighting from 3.2 and 3.3, but we need a specific use case to test against. Dan has enough expertise in coding that he could cobble this together. This is an exploration and framing whereas others are more on the production side. We can bring in remote users on this on Wednesday or Thursday of the sprint meeting.  Should we poll people who will be in Minneapolis? Anita, Barry, Dan Jay, Steve, Larry, Wendy are all potentially on this team.

Simple Codebook – Time dedicated to this remains to be defined. Simple Codebook should be completed at this meeting. DDI Codebook 2.5 can be realized as a view. Are they independent or dependent on each other? Ultimately we would like to be able to extract DDI2.5 from DDI4. We will need to be clear on the use cases behind each view. We need to do a mapping between 2.5 and the new model to identify what is missing (building on the work that Wolfgang and Jenny did at Dagstuhl). There is a mapping between 2.5 and 3 and a mapping between 3 and 4, so we may be able to do this now. Marcel, Michelle, plus one or two others could get a lot done.

Complex Data Description – Two days. Turning to Complex is important to have the functionality that is in 3.2. Seven people are in this group. Currently we are just focusing on rectangular survey data, a subset of all the structures we want to describe. We still need to iron out a data structure, but that should not be difficult. Non-rectangular files need to be introduced into the discussion.

Complex Instrument  –Mary will get in touch with Barry about what needs to be done.

II. Using Confluence/Jira/Bitbucket

Achim wrote up some recommendations about using Confluence:

Concept for the DDI Alliance Collaboration Platform

Joachim Wackerow, 2015-02-11, 1^st Draft

Introduction

The three systems of the Atlassian collaboration platform, the wiki (Confluence), the issue tracking system (Jira), and the version control system (Bitbucket) should be used in a way that the work of the DDI Alliance is supported in an efficient way. This comprehends all working groups, special projects, and the DDI community. Most of the groups work in a publicly transparent way. The systems should be used accordingly.

The goal is to have all working documents, issues, and files etc. in the systems. All items should be publicly available and detectable by search engines.

Note: It is currently not clear if the systems (open source license) could be used in a way that information is not publicly available.

The collaboration systems should be complemented by an email system with publicly visible email lists and searchable email archives.

Note: It is currently not clear if Atlassian supports email lists or if additional systems like Google groups should be used.

Possible Groups

Governance

Executive Board

Scientific Board

Executive Director’s Advisory Group

Technical Committee

Working Groups like

• • Controlled Vocabularies Working Group
  • Qualitative Data Model Working Group
  • RDF Vocabularies Working Group
  • Survey Design and Implementation Working Group
    • Tools Catalog GroupWeb Site Development GroupProjects likeMoving ForwardIndividual Member RepresentativesPossibly for each member?DDI Community
      • Special Interest Groups like
        • Developers Community
          • Software projects
      • DDI Community

Note: Headers in italic font are for ordering purposes. It is not clear if it makes sense to understand them as a group.

Details

Each group in this sense should have an own work space in each of the systems.

Each person should have an individual account and/or email subscription.

Each person can be member in multiple groups.

Each group could have a person who is responsible for administrative purposes like configuration of user accounts, organizing the systems in a consistent way. The other option is that this is done in a centralized way. This is dependent from the questions how simple or complicated the administration is and if human resources are available for centralized tasks.

All main entry points of all systems should be available on the main page ddi-alliance.atlassian.net to ensure easy access to the information.

The assumption is that some groups tend to do work mostly in a publicly non-visible way. This comprehends: Executive Board, Scientific Board and the Executive Director’s Advisory Group.

Comments

Having links between Confluence and the main site is important. Having archived email lists would be good if that is possible through Atlassian. It is not clear if there can be restricted areas in Confluence.

John Shepardson should also see this document before we send it out to the Scientific Committee.

How independent would the community pages be? All pages regarding a DDI community should make it clear that this is not the position of the DDI Alliance.

We need a template so that groups know how to configure their pages.

January 28/29, 2015

Access to Atlassian tools

We need a policy and framework for who should have access to the tools. The TC and Modelers should clearly have access. There can be project groups. DDI developers are candidates to use JIRA and the wiki, but they are not part of the Moving Forward project. TC and the Advisory Group could make a proposal. We need to determine how to use this for the Moving Forward work and the TC work and all the working groups. We want to include more people in the open collaboration platform. Ideally we could have a mix between the Confluence site and the user list. The site could record the user list content for searches and comments, for example. We could offer each of the DDI working groups access to the wiki for collaboration. The challenge is to have a current working environment that is part of a bigger picture. Mary, Kelly, and Wendy will talk about an overall framework.

We need a document on overall use and managing users on the tools.

2. Special Interest Groups

These groups should be given space in the wiki. Achim will write something up on this for approval by members. Would we need to change the Bylaws to create these SIGs? It should be in the purview of the Scientific Board to approve or constitute groups. The NSIs could be a SIG, a standing group. SIGs should be very flexible and should have a definition in the Bylaws. Adam and Achim will follow up with Ron on this.

3. Documents on the wiki

We should archive older files to have a history of the development of DDI4. There is a good space for this on the wiki currently.

4. Minneapolis sprint

We currently have nine people who have accepted for the sprint. We do not have a sprintmaster, however. Wendy may be able to find someone at Minneapolis who can do it. Arofan was looking for methodologists who might be interested in the sprint. MPC has offered their conference room (seats over 50 and can accommodate people working in both ends) plus another room. There is a lot of study space in the building that we can use as well. Mary will send Wendy a list of responsibilities for the sprint facilitator.

5. Versioning of DDI4 packages and functional views

Will we ever have a DDI 5 or is the form of DDI4 a model where each object is versioned and each functional view is versioned so we don't need to version DDI at a higher level? And how should we market it then? DDI4 is a working name but could be changed to be more descriptive (DDI4-ever?!). One of our goals in doing all this was to not do Big Bang releases. DDI 4 and 5 could be decoupled from the versioning. If we think about when we go into maintenance mode, we may want to release new things when they are ready. The functional view is just references to specific versions of objects. We will need to keep around older versions of views. We may have to deprecate older objects. Should every identifiable have a deprecated flag? This is administrative information. We need to figure out how to add "deprecated" and version number. This sounds like something the TC needs to discuss; could they prepare a briefing document for this group? They will need to maintain the objects. When do we need to resolve this – before Q1? Probably not but we need feedback on it.

January 14/15, 2015

Present: Adam Brown, Dan Gillman, Arofan Gregory, Larry Hoyle, Wendy Thomas, Mary Vardigan

1.       Update on progress

a.       Review status of release 1

The modelers and the TC are working in parallel to review Release 1. The TC has looked at the HTML documentation and made some changes. They are drafting up the instructions that come out with Release 1. They are also creating a simple page for the review with a link to Jira. Triage will happen in Jira.

The model is well in hand and the modelers are ready to review Release 1. However, there are five critical things that are barriers to progress at this point:

XML schema hasn't been produced yet

DocBook hasn't been produced yet

Object level diagrams need to be derived from Drupal

There are issues getting the documentation out of Drupal and into XMI so it can be embedded in schema

RDF outputs are needed

People in charge of all of these things have limited availability in January so this will happen in February.

This is a three-part process: finalizing the model, finalizing the production framework, and review and release.

In terms of documentation, we need to separate content from presentation.

b.      Progress on release 2

Simple Instrument and Simple Data Description are in good shape. Other components of the release have been started but are in process.

2.       Sprint planning/funding approach (Mary)

There is now a draft plan for how and when to invite people to the May Sprint. The topics for the May Sprint are now seen as components of Release 3: Complex Data Description, Enhanced Citation, Complex Instrument, Simple Codebook, and Methodology Views, although this may change as we get feedback on the other two releases. We should also get feedback from the community (Scientific Board) on the topics to be part of this, those that are priorities for the community. This can be part of the agenda for the Minneapolis meeting. Adam will send something out to the SB about this for people to weigh in with their priorities and comment.

3.       IASSIST sprint planning

a.       Discussion of participants

Participants that would be ideal to have at the May Sprint were identified in the areas of Complex Data Description, Enhanced Citations, Complex Instrument, Simple Codebook, and Methodology. Qualitative is not a separate group any longer but is part of Data Description. We could have parallel sessions happening in Minneapolis. There is not a hard and fast rule on the number of participants for the sprint as it depends on the location of the rooms in Minneapolis. We should also send something to a handful of agencies and the Modernization Committee on Standards to see if they have methodology experts to contribute.

4.       Project management update

At a recent meeting of the Executive Board, the Board expressed its views that the PM role sounded too large for a volunteer to manage. The job description was revised accordingly by Mary and Adam to focus more on communication. The AG needs to take a greater role in this plan as do the Team Leaders. They need to monitor and keep the progress going. Calling this a Project Coordinator seems reasonable. The next step is to get back to the EB on this and then send it out.

5.       HLG Information Architecture sprint (Ottawa Feb)

Following up on the HLG meeting that Arofan attended in November, there will be a Sprint in Ottawa that will focus on CSPA and information architecture, and DDI and SDMX will be important there. This is still in the planning stage, but DDI should be represented. Adam will contact Steven Vale about this so we can find out what the deliverables are and who should attend.

December 17/18, 2014

Present: Dan Gillman, Arofan Gregory, Larry Hoyle, Steve McEachern, Wendy Thomas, Joachim Wackerow

Recommendation Regarding DDI 4 Release Schedule

The TC has sent a recommendation for changes to the release plan. We are now talking about a Q1 release that would be before the end of March 2015 and a Q2 release in July 2015.

For the duration of this project, the review cycle will likely be on a quarterly basis but there will be no release in Quarter 4. However, we need to play it by ear to see what the workload from the first release is. Is a 60-day period good for comments? We can do two to three releases a year. To do this formally, we might consider having a 60-day comment period for each draft release. There is a letter ballot sent out to the entire Alliance and every voting organization is asked to respond. You can approve with comments or not approve so you get a sense of how the community views the development of the standard, how much consensus there is, and who is objecting and what they are objecting about.

We have the notion of views and the whole library – what is the approval process? Some parts will stay stable but other views will be added. We have versioned objects in the library – those objects never leave. This is clean but reusing the objects or making versioned views is not as clean. We need a policy around this and will need to revisit it.

We have the content of the first two releases clear, but beyond that we need opinions about what the demand is for new content. In the second release, we now have Classification, more extended Discovery, SimpleDataDescription, Simple Instrument, Historical/Prescriptive Process.

We need a sandbox area for the model in Drupal where views are optimized for development and not for delivery – this is a good way to work on things and model them informally.

Getting Feedback

We should step back to look at what content we have and where we should go next. The NSIs may have thoughts on this. The process should be more intentional and based on the needs of the community. Developing a roadmap and contributions to the roadmap might be helpful. We have, for example, prioritized classification management as a use case and the real world need for this may be low demand. On the other hand, the people who have been working on this model do represent the user community and have incorporated needs of their communities, so this has happened to a large degree.

To get feedback from the community for each release, we put the release out first for review as a Development Draft. We will have a specified timeframe for comments (60 days was suggested) and a template for comments. We need a public-facing websites with disposition of all the comments. Process: Disposing comments is most efficient in a small team and this would be the TC – then a larger group would review that and come out with a public presentation of comments. We may need a telecon to discuss these things. We have to be careful if we are inviting people to discuss particular issues because we still need comments as a whole.

Then we say this is what we have on deck for release next, and we invite comments on this via email and provide feedback in JIRA on priorities. We are optimizing a way to find specific people and institutions who have valuable input. For example, ask INSEE for their opinion on Instrument.

Special Interest Groups

Should the Alliance incorporate this concept into the organization?

Methodology Working Group

We don't want to spread ourselves too thin but need to make progress on this. We could get a large number of methodology people at the IASSIST sprint and this should be on the agenda. From the modeling perspective, the Methodology group should not convene until February because a lot of what they need is being worked on now. Can we identify core people now but not start the work yet? We need to understand what it is that methodology means and what we are talking about. Steve suggests a first meeting of the group for a scoping discussion. This should be in late January or early February. Dan, Wendy, Barry, and Steve will identify people.

OWL/XSD-XML

Achim indicated that OWL should be part of the first release and that we also need "a document on the design of the XSD and the mapping rules from UML to XSD, which  should be available for the release. This is currently missing. Both, this document and the resulting XSDs, are important to have a good basis for broader discussion and comment on the style of the XSD/XML."

There are some issues with the OWL and RDF but nothing is a show-stopper so we should move ahead. In January we should get a virtual group together on this. There will be some issues, but we need to get more input from the community.

We don't yet have a schema that parses and this needs to be resolved. This is doable in Quarter 1 2015, though. Arofan will interact with Oliver on this and open this up to others as well once the immediate problems are solved. Any design decisions should have more input.

Both OWL and XML Schema should be part of the Q1 release.

There are people who can do virtual work on RDF who were not at the sprint in London, and there are others at GESIS who can contribute. How do we organize this? Achim will start work the middle of January and can lead this effort by email and phone. We need more perspectives to make decisions. Mapping to other existing vocabularies in the Semantic Web is another issue. We could highlight this as an open issue in the review. There is a policy issue: Would we like to use other vocabularies, to what extent, and do we want our own objects to be independent from other vocabularies? We decided not to have foreign things in the model, but now we have to deal with the representation. We know what the options are in the RDF technology and we want the experts to tell us how to implement the mapping.

After the release we should have this discussion in a broad way. In the first release we will have a plain OWL representation without other vocabularies. We should ask for comments on the fact that we only have the DDI namespace and are not using other vocabularies.

AG Meeting 2014 12 10

Present: Larry Hoyle, Wendy Thomas, Mary Vardigan, Joachim Wackerow

Debrief from the London Sprint and discussion of actions arising

Achim and Arofan will produce a report on the Sprint. We need to get people who were not at the Sprint to understand what was done.

Overall the Sprint was productive and everyone was engaged. The RDF work still needs to be done and will have to be done virtually. Another issue is that availability to work on the Drupal side in January and February is limited and other people's time is at a premium. Getting feedback from the production framework people to make sure that what comes out is what is needed is critical.

This is a review process and will be iterated a lot. How much of the XMI to XML and RDF limits production? Quite a bit. We need to get the transformations done to put out a release.

By mid-January most people should be ready to start work again.

Modeling Team Decisions

The modeling team has made a lot of decisions and will have a communications document by the end of the week. This needs to be pushed to everyone who is doing modeling. We might add something like Project Principles or Design Principles and Conventions to the Confluence wiki. There is a concern about the legacy information on the wiki also. We will be sprinting for a long time and should be able to point people to the right principles and conventions that are up to date. Older information should be archived or deleted.

Use of Atlassian tools and communicating with the community about their use

We now have an open source license for the Atlassian tools. How widely do we want to open the tools? On the modeling team page, we implemented a subpage "Ask the Modeling Team a Question." When it comes to Jira, it's a more controlled environment. In the long run the collaboration tools should be open for any purpose for the community but we need separate spaces in the system. TC needs a controlled issue system. It would be nice for users to enter issues in another space. To do all these things, we need a manager with good rules to organize the access rights, etc.

Achim will write a proposal to Mary and Adam about the special interest groups like the Developers. Access at different levels is related to this.

We should put out an announcement about the new tools with the URLs and pointers to group pages.

Review of items for the first release package

(1) Conceptual (Library Package)

(2) Representations (Library Package)

(3) Primitives (Library Package)

(4) Agents (Library Package)

(5) AnnotationCoverage (may become Discovery Library Package for release) - AnnotationCoverage has since been changed back to Discovery

(6) ComplexDataTypes (Library Package)

(7) Process Core (Library Package)

(8) Identification (Library Package)

(9) Agents (Functional View)

We said we would put out the Discovery view but are also now putting out the Agents view. We want to cover all of Disco in the Discovery view but right now 90% of what is in Disco is not there.

We now have the capacity to produce tables that map DDI 4 to DDI 3.2 and GSIM. The model in Drupal is now more GSIM-like than earlier.

The biggest gap is figuring out how the Schema documentation will work. OWL will probably not be included in the first release.

The style of XML is what people have to use. The Agent View will illustrate the style of XML. Right now the Discovery view and package are the same (same with Agents). Ultimately the views will cross packages and we want to show this.

The framing of this release is very important. This release is preliminary and is a test of all the systems. We need to be clear about that so that expectations are not too high.

We will plan to add some elements from, e.g., Conceptual to show people that the Discovery View is different from Package. This will need to be added to the priorities list for the modeling team. This will be added to the total release package.

Re citation, if we can get in the role and degree of contribution compatible with the "Credit" vocabulary, this will be good and the dynamic set of attributes can come later. Getting this into the first release will be good.

Project Manager

There is a scaled down job description for this that involves coordination of virtual teams, communication, and wiki maintenance. This is going to the Executive Board for approval and if accepted will be circulated widely to get a volunteer.

Arofan suggested a possible candidate who was involved with OASIS.

Future sprint planning (dates and locations, topics, participants)

The next sprint will take place in Minneapolis in May 2015. The focus will be a mix of content modeling and technical work. Modeling people, content people, and framework people will be necessary for this meeting. We will need breakout meetings, probably four rooms.

The pieces of the next release are Simple Instrument (Capture), Simple Data Description, and Classifications. Is this realistic to release in June with the modeling complete by March? The datum has been added to the SDD discussion but we want to put this out as envisioned at Dagstuhl with the variable cascade. It should be made clear that this will change.

There is a provenance and process design model also being worked on. This might be finished earlier in the delivery chain and could be added to the second release package.

Classifications is very close to being finished. Codebook will take more time as will Methodology. We should probably juggle some things around in the current Release Plan. TIC will come back with a proposal for revamping the release plan and this will be discussed by the AG. At the first meeting in January we can talk in this group about the revised plan and the participants in the next sprint.

AG Meeting 2014 11 19

Present: Adam Brown, Dan Gillman, Larry Hoyle, Therese Lalor, Steve McEachern, Wendy Thomas, Mary Vardigan, Joachim Wackerow

Confirming the Outputs of London Sprint

Jira has big chunks of work called “epics” which are broken up into smaller achievable pieces that can be worked on by day at the sprint. In Confluence, they appear here: https://dditools.atlassian.net/wiki/display/DDI4/London+Sprint

Complete tools work needed for January release

Put together the draft January release for review by the TC

Complete modeling for packages to be released in January

Determine how the production framework is managed after the project

Therese is adding in the smaller pieces that need to be produced under each epic.

All work should be done using the collaboration tools.

DDI 4 Advisory Group Meeting
November 10, 2014

Adam Brown, Dan Gillman, Larry Hoyle, Therese Lalor, Wendy Thomas, Mary Vardigan

London Sprint

This will be organized as a “regular” sprint. Should the facilitators’ meetings continue to take place given that the group is smaller? The benefit is that the group has a consistent view of the progress being made.

Therese would like to get the full group involved in planning the work program on the first day and then see where the group goes from there. The people attending are very clued in and should be involved in the decision-making. A compromise is to do this first and then have periodic check-ins during the week to assess progress.

We are still missing RDF expertise at the London sprint. Achim is following up with Franck on whether he knows any available experts who might be able to attend.

Sprint Funding Arrangements

We need more advanced notice of sprints so we can decide who is essential and who should be funded.

One issue that arises in terms of planning far in advance is that it requires a tight analysis of the plan moving forward. When we get to IASSIST, will we be working on modeling or content issues, or both? Right now we don’t really nkow.

At the beginning of March we can put out a call for funding requests. We need to think about how the feedback on Release 1 figures into this.

Feedback on Release 1 and Beyond

We need to work on how to get feedback, what that feedback looks like, and who to request it from.

Mary will contact John Shepherdson to get his views on issue tracking. Jira can handle both issue tracking for existing specs as well as development issue tracking.

Project Management

We need to scope the project manager role to be manageable so we are not asking too much. A suggestion was made that we focus the project management more on the coordination across teams and less on the planning and have the AG take on the planning. The project manager can concentrate on prodding the content teams.

Dividing the operational aspect from the subject matter or technical aspect makes sense. The project manager doesn’t need a deep technical understanding of all the requirements. But the manager has to be involved in the planning to understand the structure and process to be able to support it.

There may be levels to this. There is the practical piece of gathering minutes of meetings, etc., especially between sprints when we need to keep momentum going and keep things flowing.

In the absence of a project manager, it’s a good idea to have someone to keep an eye on those meetings that are happening. We should draft a description with this scaled down set of duties and then publicize it to, e.g., the IASSIST community, the statistical community, and the data management community. Anyone with project management experience should be eligible.

Mary will write up a description and circulate it to the group.

Release Plan

The Executive Board reviewed the new Release Plan and approved it. Kelly will put it on the wiki as the next step. We can include this in Adam’s monthly report that goes out to everyone, including the NSO community. An update could be given at the HLG workshop as well. The IASSIST list is also a good group to send this to.

Kelly will update the communication plan to cover this and will put it on the wiki.

Adam and Mary will work on communications for the HLG workshop.

Long-term Planning

Steve had put forth some ideas about where we are and where we need to be going, looking forward to the first part of 2015. We need to look at the complex cases and what our approach is there. We need to bring to the fore those groups we want to be active.

This should be the main focus of our next meeting.

Next Meeting

The AG will meet every two weeks starting in November. We will do a Doodle poll for coming weeks to see which day and time works best for all.

November 3-4, 2014

Present: Adam Brown, Kelly Chatain, Dan Gillman, Arofan Gregory, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan, Joachim Wackerow

Update from Dagstuhl

Kelly sent out a summary report from Dagstuhl, which described the accomplishments of the sprint and also listed action items.

For the Data Citation group, the most important next step is to determine how to handle and model controlled vocabularies in DDI as there are competing proposals. There is a conceptual model for this in GSIM but we still lack the implementation level, although XKOS has aspects of this.

In terms of Simple Instrument (now known as Data Capture), the content modeling is close to being complete and can be sent to the modelers.

Codebook is in a holding pattern until the views it depends on are modeled.

For Data Description, the picture is more complicated. An agreed-upon model came out of Dagstuhl and is about ready to go, but since Dagstuhl there has been an extended discussion about modeling at the Datum level. The group discussed putting out what we have versus trying to remodel based on the ongoing discussions. Data Description is slated to be part of  Release Package 2 in April 2015. It was decided to go with what we have in hand, which is the traditional view of a dataset with variables, but to continue this valuable discussion to see how far we can get.

The Classification group made some gains at Dagstuhl with some remodeling done. The group needs to incorporate some notes and communicate with those not at Dagstuhl.

In terms of integration of views, there is still a lot of work to be done. Discovery needs additional work. The Process model is pretty close to being ready. The Data Capture piece is the first implementation of the process model and it seems robust.

The Modeling Team has a fair amount of work on its plate and it was suggested that Therese join those meetings to facilitate getting the work done.

There were a couple of side discussions at Dagstuhl that need to be captured: the CDISC-DDI mapping and the URN resolution work.

Looking ahead to London sprint

The group reviewed the list of people planning to attend the London sprint during the last week of November. Currently the list numbers about 11 people but two more are possibilities.

The goals for the sprint are to finalize the production process and to package drafts for a first release in February. Some repackaging of material on the Drupal site is also necessary.

In London, it would be good to hold facilitators meetings twice a day as in Dagstuhl with Therese, Achim, Arofan, Wendy, and Jon participating.

In terms of production, a finite amount of work needs to be done on the XML Schema representations, and there will be people at the sprint with the right skills to accomplish this. However, there will be no one in London to work on the RDF representations. Some of this work may be able to be done ahead of the sprint. Also, some possible participants with good RDF skills who are in the London area have been identified and will be contacted.

NADDI sprint and alternatives

The group discussed whether it will be possible to bring together a good set of participants for a sprint after NADDI in Madison. Alternatives are a virtual sprint and finding other meetings where DDI people may be in attendance (RDA was suggested).

Feedback process

Once DDI packages are out for review, we need to have procedures for taking in and disposing comments. For transparency there should be a log of comments and a way to formally publish their disposition. It is possible that the new Jira and Confluence systems will be able to document the feedback in a way that is efficient and transparent, but we need to look into this. Having templates for feedback and automating input into the tracking system would be ideal.

Project management arrangements

Kelly is no longer able to participate as Project Manager, although she will continue to maintain the wiki for now and will also contribute to the redesign of the DDI website.  The group thanked her for her substantial contributions.

It was decided to prepare a Project Manager job description and post it to not only the DDI community but to the wider IASSIST list.

Next meeting

Monday, November 10, at 2pm EST

September 29, 2014

Present: Kelly Chatain, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan, Joachim Wackerow

Communications plan

Kelly provided a communications plan with columns for information to communicate, stakeholders/audience, method, responsible party, and frequency. Providing monthly updates to everyone on the status of DDI 4 was seen as a good idea by the group. The Scientific Board should also be notified about the project regularly and when approvals are needed. Some internal communications among modelers is included in the plan, but all communications should be seen as public since they might be made available to others.

Release plan

Kelly provided a new release plan with the first release occurring in February 2015. The releases should be considered betas or drafts for public review and a preface written to indicate that this is the plan right now but that unanticipated events may necessitate changes. The AG should take a week to review and let the group know if they have changes. Then this new plan can be communicated with the preface. The TC will write guidance for each release about how it should be used and its purpose.

Communicating targeted invitation process for Dagstuhl

Mary provided a draft message. It was decided that this message, with edits, should be the first of the status updates that will occur monthly and as such it should come from Adam on behalf of the AG. Mary will edit the message and send it to Adam.

London sprint participants.

There is a need for more expertise in linked open data. Achim will contact possible participants in this space.

September 15, 2014

Present: Adam Brown, Dan Gillman, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan

Dagstuhl Attendees

There is space for one more attendee. It was decided to invite Jenny Linnerud.

London Sprint Attendees

Several names had been suggested for this meeting, which will be a technical sprint. Mary will send out invitations so that people attending EDDI can make plans. She will also update the spreadsheet of invitees.

In general, we should have some subject experts at the London sprint as well. Once we have acceptances for EDDI, there may be other subject matter experts we can tap.

Sprint Funding – Need Topics

We need to make sure we are clear about the sprint topics and the competencies required to produce the deliverables so that the Executive Board can decide whom to fund having the information they need. Topics are:

Refining the renderings to XML and RDF

Refining the production framework

Finalizing packages for publishing

More work on XSLT

Technical modeling

There is a modelers meeting on Wednesday morning, so Wendy will bring up these topics.

DDI 3.2 Beta including Sampling Information

TIC brought this to the Scientific Board because people need the sampling information but don't want to wait for DDI 4. The dilemma is whether TIC should put all of their efforts into DDI 4 or pursue the beta at the same time. Dan converted the XML to UML in Toronto, so it is ready to go into DDI 4. The beta would be backwards-compatible with 3.2. We will get this issue out to the Scientific Board for their views.

Release Plan

We have agreed to postpone Release #1 till January/February 2015. The Executive Board wants assurance that future deadlines will not slip if we publish Release #1 in January/February 2015. It is difficult to know since there are technical and content factors. Given that things are still evolving, we may need to extend the timeline to be more realistic. We may want to build in more time between releases (an extra month or so?). The process for getting feedback and integrating it hasn't yet been worked out. Alternatively, should we go back to the NSIs and the HLG  for more resources? They have contributed additional people to attend sprints. We should discuss this with the EB again. in the meantime, Kelly will work on a more realistic release plan.

Profiles

The NSIs want profiles that are currently in 3.1 converted to 3.2 and would like the community to work on this together. It isn't clear what the priority should be for this work. DDI involvement would add to the consistency of the profiles, but we have a lot on our plates right now. Adam will try to get more information on this – especially, whether this project is the generic profiles or something more specific. The UNECE committee on standards is working on this also. We need to be aware that agencies find GSIM and DDI daunting and they may be waiting for DDI 4 to to commit to adopting DDI.

August 25, 2014

Present: Dan Gillman, Arofan Gregory, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan, Joachim Wackerow

Chair Adam Brown was not able to make this meeting.

1. Upcoming Dagstuhl Sprint

Participants

So far 23 participants had been invited, with two open slots (Dagstuhl attendance is capped at 25). It was decided to issue two more invitations with the goal of ensuring good representation of NSOs and then to start a prioritized waiting list in case other slots opened up.

Action: Mary to issue invitations

Action: Achim to add a waiting list to the master list of invitees

Agenda

There will be a special emphasis at Dagstuhl on enhancing data citation information in DDI. This work is funded by a grant from NSF and will support the attendance of 10 people at Dagstuhl. This group includes some individuals external to the DDI effort who will need to learn more about DDI, so we need to think about the agenda for the week and the kinds of information to provide in advance. We may want to have a call with new people before Dagstuhl to get them familiar with the standard.

We may also want to hold a teleconference during Dagstuhl with people who have created a taxonomy for the research contributor role. These individuals were not able to attend in person.

Drupal support

We will likely need Drupal support during the week and should have access to Drupal experts.

Action: Larry and Mary will assemble a list of the kinds of sessions and information that new participants will need, and then Achim, Arofan, and Wendy can determine how to integrate this information into the week.

Action: Mary will ascertain the availability of the taxonomy experts to be on a call the week of Dagstuhl.

Action: Mary will issue an invitation to Drupal experts to be available for assistance during the week of Dagstuhl.

2. Release Plan

A rationale for changing the release plan was laid out: the content and tools needed for the first release, scheduled for the end of September 2014, are in rough status, and there would be no benefit to publish with the goal of getting meaningful user comments. It was also mentioned that “release plan” has the connotation of finished products, which is not the case with our iterative approach. It may make more sense to call these “pre-releases.” Also, given that the model-based DDI is an R and D effort, we may also want to indicate that dates are subject to change.

The view was also expressed that we need a comprehensive document/policy that discusses and positions the different DDI development lines and versions (DDI Codebook, DDI Lifecycle, DDI RDF vocabularies, and the model-based specification that is being worked on now) in relation to one another. This would include future plans for maintaining and updating the different versions. This document should be available before the first DDI 4 release so that users can make informed decisions about which products to use.

Action: Arofan will create a first draft of this document for discussion.

Related to this, a suggestion has been made that the detailed sampling information assembled by the Survey Design and Implementation working group be added to the existing DDI Lifecycle, perhaps as a beta version (the sampling piece has been finished for quite a while but has not yet been made available). Not everyone agrees with this approach as this could take away resources (and content) from the modeling work. The Technical Committee is writing up this issue and will be contacting the Scientific Board for guidance. It was also suggested that the Marketing and Partnership Group which is soon to be convened might weigh in on the issue of how to market the different DDI product lines.

A suggestion was made that, given the difficulty of having a good product by the end of September, we should aim for an internal pre-release package in November with the goal of validating the package and procedures at the EDDI Sprint, which will be technical in focus. The group agreed with this, but there was not clear consensus on a formal release date for Pre-release #1 to the public. The group was divided on whether this should come out in mid-December just after EDDI so that we have one package out in this calendar year, or if we should delay the first pre-release until the end of January 2015. It was decided that people should take a week to mull this over and discuss it via email and then make a decision.

Action: Mary to send out email about this and initiate a discussion.

August 18, 2014

Present: Adam Brown (Chair), Kelly Chatain, Dan Gillman, Arofan Gregory, Larry Hoyle, Steve McEachern, Wendy Thomas, Mary Vardigan, Achim Wackerow

1. Confirmation of Dagstuhl invitees

Participants agreed to the tentative set of invitees developed by Achim, Wendy, and Mary. This sprint is complicated since up to 10 people will be working on adding enhanced citation information into DDI as part of an NSF-funded project.

Action: Mary will send out invitations to recommended invitees on Tuesday, September 18 with a deadline of responding by Monday. At that point we will move to the second tier of invitees when needed.

2. Relationship between DDI items and other standards

It may not be up to the AG to make these technical decisions but the group can contribute some guidance.

The consensus seems to be that DDI shouldn’t be pulling other standard’s elements into our model and that only DDI-modeled elements should be included.

There should be a way to express the relationship in a formal way if there is equivalence so a program can follow this.

Having a single document to describe this would be good. For the next meeting, the document will be updated and people should then read the document and comment on it.

3. Release plan and communication

This was postponed until the next meeting.

4. Next meeting

Monday, August 25

5. Sprint around IASSIST 2015

The recommendation is that we do have a sprint the week before IASSIST. This will be the week of May 25, 2015.

DDI Moving Forward Advisory Group Meeting
July 29, 2014

Minutes

Present: Adam Brown (Chair), Kelly Chatain, Dan Gillman, Larry Hoyle, Therese Lalor, Steve McEachern, Wendy Thomas, Mary Vardigan, Achim Wackerow

Discussion of Chair Role

This was the first meeting of the Advisory Group (AG). As there had been three nominations from within the AG for Adam Brown as Chair, it was decided that he should serve in that role.

Status /Duration of Moving Forward Project

Teams

Therese Lalor, DDI 4 Project Manager through July 31, provided an update on the status of the DDI 4 project. She indicated that there were five teams currently working:

1. Simple Data Description Team (No chair yet)

2. Simple Instrument Team (Barry Radler)

3. Classifications Team (just starting up with Flavio Rizzolo as modeler but no chair)

4. Tools Support Team (this includes the production run, Drupal, OWL, XMI, etc.) (Arofan Gregory)

5. Modeling Team (Arofan Gregory)

The Simple Codebook Team, chaired by Wolfgang Zenk-Moeltgen, is also meeting with a hiatus until later in August.

Release Dates

Therese indicated that the project is aiming for a September date for Release 1, which will  contain the Agent, Process, and Conceptual components of the DDI 4 Library and the Discovery Functional View.

It was pointed out that while the above components are under review separately, there has not yet been an integrated modeling review to ensure internal consistency. Moreover, there are still several outstanding issues regarding the production process, and the release plan in general seems ambitious. Quality is important and we shouldn’t rush the process just to meet the timeline. For these reasons, it was proposed that we carefully craft a description of the September release when we communicate about it to the public. The description should indicate that this is a provisional release put out for comment and feedback. It is not expected to be perfect in terms of content and we want to get reactions to the product suite that we put out.

Action Item: Mary will draft a description of the provisional release.

Content Backlog

Therese led the group through a presentation about the content or product backlog, which is the driver of progress according to the Agile development paradigm. The backlog is a high-level list of requirements and user stories; anyone can contribute to user stories and add new ones. Other ideas can also be proposed.

The idea is to chunk up the backlog into marketable products and to make many smaller releases, iterating over them as feedback is received. The project focuses on the backlog and only works on that. The Product Owner (in this case the DDI Moving Forward Advisory Group) and the Project Manager decide together on the release schedule.

Release Plan

The AG needs to confirm the products to be released. In particular, the AG needs to review Methodology and Comparison/Harmonization to see what is in scope and what can be done in the time allotted.

Release 1 is scheduled for September, Release 2 for December, and Release 3 for March 2015. Right now, Release 3 items are scheduled to be worked on at Dagstuhl in October. It was pointed out again that this is an aggressive timeline and we should be thoughtful in how we communicate about it as it may be subject to change and there will be iterations. Quality and sustainability of the DDI process should be the key considerations.

As an organization, we will need to determine after March if there is a formal follow-up project to complete DDI 4 or if it moves into the existing framework for updating the specification. This will need to be taken up with the Executive and Scientific Boards as there are resource issues.

The history of the March 2015 deadline was that there was a strong commitment to support GSIM with an implementation model, and a project span of 18 months for this activity was seen as optimal. While this may have been ambitious, it was seen in a positive light by potential implementers and thus was taken forward.

It may be that we need a Phase 2 of the project. We should aim to complete the September and December releases to the best of our ability and then assess where we are to see if the timeline needs to change.

A question was posed regarding what it would take and what needs to be done to meet the September release date, given the concerns about the integrated review not having taken place yet. The test production run done in Toronto resulted in the identification of bugs and gaps in the process, some of which are being worked on right now. There are issues with the generation of OWL, XML Schema, and documentation that need to be addressed. The idea was to do a test run every two weeks until all issues were resolved.

Dagstuhl Sprint

The group discussed the topics for Dagstuhl (Release 3 items):

Simple Codebook

Comparison/Harmonization

Methodology

Complex Instrument

Complex Data Description

Content modeling would be the focus for the sprint at Dagstuhl, while the EDDI sprint (December in London) would be dedicated to technical modeling. We need the right people to attend both.

Enhanced Data Citation Project

Larry noted that he had just received word that the University of Kansas-University of Michigan proposal to NSF to add enhanced data citation information to DDI was successful, and the grant can support seven people to attend the Dagstuhl sprint in October. These people will be content experts with knowledge of data citation. Larry will create a new user story about this topic and share the proposal with the AG. This means that we should possibly reprioritize items for the Dagstuhl sprint.

Next Meeting

The group will meet again next week to discuss the Dagstuhl sprint further. There was a request to discuss other issues at the sprint, including hypothesis building, qualitative data, and DDI for new data types. Methodology and/or Study Inception may cover hypothesis-building. This was deferred to the next meeting. Another topic for the next meeting is the EDDI sprint, tentatively scheduled for the week before EDDI in London.

Attendees: Adam, Achim, Dan, Steve, Wendy, Larry, Arofan, Thérèse

The group went around the table with each member discussing what they thought the priorities for the Dagstuhl Sprint should be.

There was strong agreement that the following views should be on the agenda:

Meeting Minutes 2014-2017 (the work funded by NSF)

Finalising Meeting Minutes 2014-2017 and working on Meeting Minutes 2014-2017

Finalising Meeting Minutes 2014-2017 and working in Meeting Minutes 2014-2017

Meeting Minutes 2014-2017

There were a number of other views mentioned. It was decided that the following ones would be included in the sprint agenda:

Meeting Minutes 2014-2017 (should include data records)

Additional note from Achim, 22/8/2014 (also added to User Story):

This view will support description by data records for qualitative segments and other objects like new data sources. This is basically a general description method for semantically not well known or well defined objects (at design time of DDI 4). This topic includes a focus on the draft model on qualitative data in terms of identifying objects which are already covered by the quantitative model, which can be described by data records, and which are necessary in a dedicated qualitative package.

Meeting Minutes 2014-2017 but actually this is some combination of Meeting Minutes 2014-2017, Meeting Minutes 2014-2017, Meeting Minutes 2014-2017 and hypothesis testing  (which is not currently in the content backlog)

Steve volunteered to update the content backlog entries to reflect the new understanding of the Qualitative and Methodology views.

There was some discussion about identifying participants for the Sprint. I think Larry said that NSF will fund 7 people (who are data citation people) to attend. [on reflection after the meeting, Thérèse notes that there maybe be some other funding considerations to take into account.] An initial list will be prepared by Arofan, Achim and Wendy and will be discussed with Mary next week.

The AG will next meet on Monday 18th / Tuesday 19th at the same time.

DDI Moving Forward Advisory Group meeting

14 September 2016

Attendees: Larry, Wendy, Dan, Jared, Achim, Steve, Michelle

Dagstuhl preparation - week #1 attendance

Elise - via Dan - still up in the air

1. 1. Add Elise to the list - and set as a last minute cancellation
   2. Dan to provide address of organization and email to Achim for Dagstuhl

Currently have 19-21 attendees

Harold - not available - Wendy will keep up relations with Harold as we move along.

Dagstuhl preparation - Week #2 attendance

22 confirmed + Sam and Kate and Oliver = 24 (yes that's really 24 )

updates

1. 1. Chuck cannot make it
   2. Sam Spencer - Steve will send formal invite - unsure but keen - will confirm and may have internal support
   3. Matt Wooler - cannot come suggested Katie McNeill - is a UKdata option - Steve will formally invite
   4. Oliver - still waiting for confirmation
   5. Mari - cannot attend - recommended Katia - who can attend

Preparation calls:

Sept 20 - Week #1

Sept 22 - Week #2

1. 1. May be a follow-up CV call with Wendy

Jared will send out email with call attendees for the 2 Dagstuhl weeks

Goal of the call:

1. 1. Jared - overview of the DDI
   2. Wendy / Achim - experiences of last year and goal of the workshops / what do we want to achieve / outline for their presentation
   3. Documentation - need to send outWendy / Michelle to update participants on the Sprint pageMonday call - Achim, Wendy, Jared, Jon, Michelle - prep for Tuesday call
   • • How will we handel those that cannot make the Tuesday call
   1. Wednesday prep call for the Thursday prep call

2015 Documentation - will be posted on last year's Sprint -

we will need to link it to this year's Sprint

Updates -

modelling team - Sept 23rd deadline to get everything up to TC

1. 1. Data Description group going up to Modelling
   2. Working expediently and getting pushed it through
   3. To have in TC hands by the 23rd of September
   4. Fast review - report card on it - JIRA in place now
   5. Turn around in a week

Codebook group

1. 1. Developed a view and working on adding in Data Description and Methodology
   2. Found some items that the Modelling team need to address - description of security and access - will need to be addressed in a general way

Sprints at EDDI, NADDI, IASSIST?

AG should plan if we need Sprints

Jared - budget aspect and if require - how do we make them happen

If we need Sprints - make room reservation

EDDI - already has the room reservation - if we have meeting - focus on the technical framework -

1. 1. Will be needed
   2. responses back from Review that will need to be addressed
   3. already have tech issues that will require a Sprint

NADDI - make space available - Cancel without penalty - 2 rooms - Passover starts on April 11 - food that is not bread and grain

IASSIST - may be easier - do it the week afterwards - 2 rooms