Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 142 Next »

 20170216

ATTENDEES: Wendy, Larry, Dan S., Jeremy, Johan
Review of DDILIFE issues associated with article by Larry and Achim posted on 2934 including: 2935, 2934, 2933, 2932, 2931

Worked on 2934 to clarify exactly what was being addressed and how to make this machine actionable information that could be used by statistical packages

 20170209

ATTENDEES: Wendy, Jon, Dan S., Jeremy, Larry
The following have been entered in Bitbucket and need review:
3515
3494
3517
3514
DDI Lifecycle issues related to Variables were discussed:
3512 - wlt will write up edits
2935 - agreement, Dan S. will write up for review prior to entry
2934 - Discussed - will return to next week
3522 - agreement, needs to be entered

 20170202

ATTENDEES: Wendy, Jon, Dan S., Larry


Q2 review wrap up:
What needs to be completed prior to report
Do we do a preliminary report (by end of February) with a follow up at annual meeting (some issues may not be resolved easily)
Working with other teams where issues were assigned
What should be included in report?

Discussion

  • We can say at the end of Feb/March that this is the current status
  • There are generic issues that are still being worked on
  • If people are particularly interested in particular things they can follow up on issue tracker
  • GSIM and RDF broader issues
  • They would be resolved by next release???
  • Respond to individual reporters
  • Evaluation of the review system in TC
  • Follow-up with other groups:
  • We had 85 issues raised and were shoved off to various places
  • Describe how we're going to manage over the next few months
  • Follow up with each group getting a short paragraph from each group about status of these issues: Codebook requirements, things dependent upon other things, etc.


DDI-Life:
Reviewed for inclusion in 3.3 and basic comments
*=DDI3.3 priority, F=Future
*3532 - Measurement Unit
*3533 - Measure Definition
*3534 - Include documentation on DDI - CV availability
*3535 - INSEE use of Conditional text uses this as a means of dealing with language variations
F3536 - Incorporation of DDI-Views Process Pattern
F3537 - Rest Protocol
*3538 - Field for unavailability remarks /reason for data not being available
*R3539 - Controlled Vocabulary for Data Access
*3540 - Detailed Information about Individuals
Got to this point: members asked to review and identify any that should not be DDI3.3 Priority
3541 - Variable Quality
3542 - Detailed Information about Literature
3543 - Field for Products
3544 - Field for File Size
3545 - Add Measurement and Measurement Construct to Instruments
3548 - Would like to capture relationship "isReplicateOf" and opposite

 20170119

ATTENDEES: Wendy, Dan S., Larry, Jon
Dan S. won't be available next week
D4W2-65
D4Q2-90
D4Q2-91
D4Q2-85

Note: XKOS review extended through 2017-02-15

 20170112

ATTENDEES: Wendy, Johan, Larry, Jeremy, Jon, Guillaume, Dan S.

Q2:

Reviewed triage decisions, approving approaches

Lifecycle issues 3533 and up
Reviewed
3533
3534
3535
3536
3537
Send out for comments on priority for 3.3
3538
3539
3540
3541
3542
3543
3544
3545
Timelines:
Create post Q2 closure to avoid confusion and post to DDI lists (TC will monitor):
Instead of having a review specific have a DDI4-TriageReview issue tracker.
Send out note:
Will start looking at these the 19th so if you still have items to get in do it by then.

 20170105
ATTENDEES: Wendy, Dan S., Larry, Jeremy
AGENDA:
  • Reinstatement of DNS work
  • RDF Vocabulary Public Review (XKOS, DISCO)
  • Extension of Q2 review period (through Jan 15)
  • 3.3 work
Reinstatement of DNS work:
  • registry.ddialliance.org
  • arpa domain name - applying for a URN namespace
  • Other stuff has been up and running at the point of the agency resolution
  • Do we need some additional instruction about registering an endpoint
  • DNS service records are not something people use everyday
  • The biggest question is what should the format be supported (if 3.1 is not supported going forward is this a problem)
  • Should just be Agency, ID, Version
  • Should be simplified to support the Agency, ID, Version content and not go into the detail
  • There seems to be a lot of extra stuff such as late bound information
  • Examples the type of item in URN should be removed (a 3.1 particular)
  • Should focus on current and future versions that will all use the same URN format because that works with
  • Very specific about which characters are allowed where and how agencies, versions, and ids can be defined - change to NCName to allow us flexibility in the future
  • It should be opaque to the outside
  • None of this is automatic - this is something a computer programmer will have to write
  • Just makes it officially recognized - this is what the ddi namespace means
  • The reason for doing this is to keep someone else from register ddi as a namespace
  • Draft up a statement of work and pass around for review, followed by approval from Jared and send to Achim
3.3 work
  • Updating issues to be in sync
  • Put on next weeks agenda
Next: new issues in DDI-LIFE, Q2 triage/results
 20161222
ATTENDEES: Wendy, Larry
AGENDA
  • Reinstatement of DNS work - Yes
  • RDF Vocabulary Public Review (XKOS, DISCO) - XKOS by EOY and DISCO when Achim says its ready
  • Extension of Q2 review period (through Jan 15) - YES
  • 3.3 work - delay until next meeting
I spoke with Achim at EDDI and apparently the work he was doing regarding filing for a DNS. This will be the basis of getting approval for the DDI URN and having a related DNS entry for arpa.urn.ddi (for the DNS-based resolution of DDI agency ids).

Here is the statement from the 2014-06-02 DDI Alliance Meeting of the Scientific Board:
Joachim Wacker reported that he had been working on the issue of resolving DDI URNs, which involves communicationwith the Domain Name System (DNS), a hierarchical distriubted naming system for computers, serices, or an  resource connected to the Internet or a private newwork. The Alliance needs to interface with DNS to resolve DDI URNs now that 3.2 has been published and the identification mechanism is stable. Achim will be developing a Request for Comment document for the Internet Assigned Numbers Authority (IANA) and then will apply for the DDI URN and DNS entry, which will help to resolve DDI URNs int he future. URNs will be resolved in a two-stage process: the DDI agency ID gets resolved by DNS and the requesting software receives an answer as to what kinds of DDI services are available; then the local system is queried and URNs are resolved.
Funds have already been allocated for this filing. The most recent document I seem to have is from 2009 and is attached. However, I may have something more recent on my home computer.
The point of the agenda item is that Achim would like a formal go-ahead before he picks up work on this again. As you see this document is from 2009 and would require updating to support all versions of the DDI URN.

Here is the latest version:
The required work would comprehend the adaption to the current DDI URN format and to some underlying RFCs which changed in the meantime changed. 
The formal definition of a DDI URN is required if we want to use them broadly. This RFC draft (see above) defines the DDI URN before the great identifier discussion. Furthermore it defines a distributed way - based on DNS - to resolve DDI URNs in a two step process to local services and to physical locations.
This work and the definition of a DDI REST protocol could build the basis of a distributed DDI-based infrastructure.
One question is which versions of DDI URNs should be described by this document? I assume 3.2 (and 3.* >3.2), 2.5 (and 2.* >2.5), and DDI-Views (no maintainable anymore).
Re Disco:
Open issues are here: https://github.com/linked-statistics/disco-spec/issues?q=is%3Aopen+is%3Aissue
 
Some issues regarding the model should be probably closed before the review, like 220-228. I can try to get this resolved with Thomas and Ben by email in mid of January. All others I can try to resolve in January or they will be open while the review. They are not critical. Some cleanup work has to be done on the website as well. I can take care for this. I’ll be in office again starting from Jan 16.
 
We have to deal with some time restrictions. Thomas is available only on the weekends for calls. Ben is busy until end of January with lectures. I assume both would be available in February for answering possible questions by email regarding new issues coming up while the review.
 
This could result in a plan of finalizing Disco for review in January and starting the review for example on Jan 30 or Feb 6 (more relaxed).
 
Re DDI URN:
See also my email from Monday.
I see following main questions:
-        Do want to go for a DDI URN? I’m still convinced that this is the right approach.
-        If yes, which versions of DDI should be addressed by the RFC for the DDI URN and which not. I assume 3.2 (and 3.* >3.2), 2.5 (and 2.* >2.5), and DDI-Views (no maintainable anymore). Can DDI 3.1 be omitted?
-        Should we plan a REST protocol? I think this is the next appropriate step to provide a basis for distribute DDI resources (identified by the DDI URN, resolved by DNS and local services, and queried by DDI REST).
Currently we just should express a strong interest in this; then we can go ahead with making a plan.
 20161117
ATTENDEES: Wendy, Larry, Dan S., Jeremy, Chantal, Flavio, Jon
New member: Chantal Ripp
From StatsCan working with Data Liberation Initiative using DDI-Codebook with some transition work to DDI-L and DDI4
Issues remaining possible for 3.3 (see spreadsheet)
Next meetings: 11-24 Thanksgiving Day US (no meeting); 12-1 WT not available (Jon); No meeting on 12-8 EDDI; No meeting 12-15 Sprint
 20161110
ATTENDEES: Wendy, Jeremy, Johan, Dan S., Larry
D4Q2:
D4Q2-35, 34, 33, 15 verified completion and closed
filed DMT 107 regarding consistency in usage of URN, URL, and uri in conjunction with xs:anyURI
DDI 3.3:
Set prioritites for meeting through December
  • Go through to see  which are resolved but not entered and correct status - send a list
  • *In/Out and Binding documentation should be earlier rather than later to feed into DMT issue on this
Technical tasks from Mantis
  • Is there a need for a "technical" task? No - reassign types
  • Triage what is there in terms of importance, relevance
 20161103

ATTENDEES: Wendy, Jon, Dan S., Jeremy, Johan

AGENDA:

  • Review process update
  • TC assigned issues


Review process update

  • Went over the general process and Triage results
  • Reviewed report Dashboards created by Jon
  • Discussed means of tracking issues assigned to other groups
    • completion
    • action taken
    • review of handling - was it really dealt with or just tabled
  • ACTION: Jon and Wendy will look at options for better tracking of group actions

TC Assigned Issues

  • Resolved the following issues in D4Q2: 5, 12, 15, 33, 34, 35, 44
  • Fixes were made and once reviewed will be closed


 20160929

ATTENDEES: Wendy, Jon, Jeremy, Dan S., Larry

Q2 Prep
Workflow is missing from Lion
  • Couple of other things different in Lion from diagram
  • Workflow realizes process and page 7 Using Process Pattern. Add in.
  • Workflow sequence contains WorkflowSteps
  • Process contains ProcessStep - make note difference.

TC review of Q2 release contents

  • Look at scorecard and design rules to see if there is something we need to add
New workflow for Q2 JIRA
  • We should be able to follow post passing on to another working group so track within other groups and if they have been resolved.
  • Link to issues as filed in other group
  • Track / review linked use tags or labels as to which group it went to
  • How does MT follow what other groups are doing? We could put in an issue in DMT with all the links
  • If TC assigns an issue to a group we should be looking at how they are following up if only for reporting purposes and compliance with design rules
  • Wendy will look at JIRA for possibilities, revise and/or provide implementation of Jon's workflow draft

ACTIONS

  • watch for materials for comment over the next week


Dan and Jeremy unavailable next 2 weeks for calls

 20160922

ATTENDEES: Larry, Dan S, Jeremy, Jon

Apologies: Johan, Wendy


AGENDA

JIRA Workflow

Jon didnt do - will do so for next week


DDILIFE-3511

Function and documentation of in/out parameters - for NEXT week


DDILIFE-3533

Need Wendy's views on implications on this. As it appears that it would mean connecting the NCube structures / references into the variable cascade.

It was though that the NCube structures were thought about solely to manage instance variables and this could mean extending in, 3.3. was not meant to be leading to major changes, so a judgement is needed here.


DDILIFE-3532

Adding MeasurementUnit to RepresentedVariable makes sense for ValueRepresentation (not for CategoryScheme - though Larry raised whether this might be a use case).

Agreed to look at other elements in RepresentedVariable to make sure that other elements should not also be included


Other elements on VariableRepresentation are:    


VariableRole?,

r:WeightVariableReference*,

StandardWeightReference?,

ImputationReference?,

ConcatenatedValue?,

r:ProcessingInstructionReference?,

(r:ValueRepresentation | r:ValueRepresentationReference)?,

MissingValuesReference?,

r:ContentDateOffset?,

r:AggregationMethod?

Also, whether there is any interaction as a consequence of DDILIFE-3512 and 3503


DDILIFE-3529

For discussion next week

 20160915

ATTENDEES: Wendy, Olof, Larry, Jon, Johan, Jeremy, Dan S.

AGENDA:

For Q2 release:
  • Report card
  • Determine what the workflow should be for Q2 JIRA issues - Jon will provide recommendation for review next week
  • Review the list of publication packages and make sure its accurate. Send to Jon and update postings.
Recommendations for Dagstuhl Week 2 Agenda:
  • CV quick presentation on where they are at and where things are going in the future
  • CV group tied in more with TC in terms of their priorities and production cycles
  • CV group tied in more with Training and Marketing
  • TC - we have people actively working on DDI and would it be productive to have an hour or so on what areas we should move into (is there low hanging fruit we could take advantage of) - goes for MF and for looking at Lifecycle and even Codebook
  • CV - there was a whole discussion of how they express CV's (Genericode, DDI Custom Metadata, Existing DDI 3 code for CV's, XKOS, others)
  • Quality statements is a big issue EU - controlled vocabularies utility in this area
  • CV - we should move them to bitbucket - as part of their introduction to use of Confluence and JIRA
To Do Items:
  • Add enter to JIRA  project workflow used by DDI-Life and DDI-Codebooks - Wendy
  • Is there a way to get a database dump from Lion? Yes Johan can arrange to have by next week.
FOR NEXT WEEK:
Scheme for review JIRA tracker
Binding and IN/OUT parameters
Review of any unagreed LIFECYCLE issues
Hold time in your schedule to review package from MT during the week of 24-30 September.
 20160908
ATTENDEES: Wendy, Jon, Dan S., Dan G., Jeremy, Larry, Guillaume
Agenda:
  • Issue 3529 - identification in Reference
  • Issue 3530 - generic grids
  • Issue 3531 - Publication of individual items
  • DDI Management document

DDI Management Document:

Approved for sending to Director's Advisory Group

DDILIFE 3529 - table for additional material on implications

DDILIFE 3530 - Generic grids is tabled to post-3.3; other related issues need comment and updates - will happen over the next two weeks

DDILIFE 3531 - Resolved to move isPublished to AbstractVersionable - additional information regarding additional publication levels was determined to be out of scope. See comment on issue.

Update provided on Q2 2016 Review - anticipated receipt of package 2016-09-23, we will need to move quickly to get review out in one week

 20160901
ATTENDEES: Wendy, Jon, Dan S., Dan G., Jeremy, Larry, Guillaume
Agenda:
  • Issue 3529 - identification in Reference
  • Issue 3530 - generic grids
  • Issue 3531 - Publication of individual items
  • DDI Management document
3529
  • Issue gets back to expressing ID content in a reference where the scope of uniqueness is the Maintainable
  • See JIRA for notes
  • STATUS: examples of XML for different cases are needed. We don't want to open the whole identification can of worms again. Place on Agenda for 2016-09-08
3530
  • See JIRA for notes
  • The discussion is good and should be continued but will it get into 3.3?
  • Useful to inform DDI 4 discussion and could be raised at Dagstuhl
  • Review specific grid issue in JIRA and identify those that have recommended resolution and those that are critical. Get these into 3.3
  • Do not hold up 3.3 on the generic grid option
  • Begin entry of resolved issues into schema
  • Agenda item for 2016-09-08 include list of grid related issues and their resolution or tabling
3531
  • Tabled for 2016-09-08
  • Please review and comment on prior to meeting
Data Management document
  • A couple of items were noted as needing highlighting
  • Wendy will revise and reorganize document as a proposal to the Director's Advisory Committee
  • Need comments over the next week
  • Final comments on 2016-09-08 then delivery to Director's Advisory Group
 20160818

ATTENDEES: Wendy, Jon, Jeremy, Dan S., Larry

Reviewed process for managing the Q2 review process. No changes or additions. Will revise during process as needed.

Discusses overall work flows and work activities of the DDI Alliance in terms of interface with TC activities and responsibilities. Discussion based on conversations with individual members over the past month. Summary of discussion will be written up and distributed to TC members for editing and further action. Topics covered:

  • Use of JIRA as a means of input between groups
  • Better discipline in use of JIRA (number of items stuck in "in process" mode is disquieting)
  • Clearer more concise goals in working groups with timelines
  • Clarifying management roles between and within groups - expectations of group leaders and members
  • Balance between concrete outputs in the short term and support for more conceptual discussions/work
 20160811
ATTENDEES: Wendy, Jon, Dan S., Jeremy, Guillaume, Flavio
3.3 QuestionGrids
examples have been placed on 353
  • Proposal was that be one or two basic grids and then others built by hand
  • Roster or rows based on codelist - can have another kind of grid
  • Allowing addition of simple label on grid - or new simplified grid alternative
  • Guidebook to using QuestionGrids will be available in English - INSEE
    • found 3 types which they describe: roster type, one-dimension grid, two dimension grid, benchmark different representation of grid where modeling is fundamentally different
  • A simplified grid would allow people to place labels where needed
  • For next week:
    • Sketch or example of proposals
    • Suggestions for what we want to provide in terms of other grids (more generic and more straight-forward)
After next two meeting: take a view of what can we get into 3.3
Review process in Q2:
Alter basic approach to:
  • Have shorter meetings with fewer people to triage and prioritize items for discussion
  • Clearer guideline for when something gets tossed back to MT
  • Good approach and need to clearly define the roles of the groups
  • TC needs to clearly define the issue for MT not provide specific solution
  • Will depend on the amount of feedback we get from Q2 and what the content of comments are
  • We have more than just the model and the content, there is process, documentation
  • We were also responding to each person who entered a comment - We still have a responsibility for the feedback to the commenters
Role of TC:
  • Triage function
    • What is it about
    • What are the issues touched on
    • Who should be correcting this (TC, MT or other)
  • Communication with commenters
  • Overall evaluation
ACTION: Draft up a definition of group roles and specific activities of TC in review (wlt)
 20160804
ATTENDEES: Achim, Dan S., Wendy, Jon, Hilde O, Guillame D, Larry, Jeremy
3518 - Question item linked to Represented variable
Concept and universe incorrectly placed in represented variable
Not a universe, it a Unit Type. this was corrected in Issue 3512
When a question is reused, the instance variable will be linked to the represented variable and hence 'self harmonised'.
DECSION:
a) add the link to Question Item 0..n these also may need to be linked to different Represented variables (structured mixed response domain)
b) review how to add to Question Grid given that it collects multiple responses and need a clear link between a subset of responses and a specific RepresentedVariable
 
Build a structure that clarifies the path between response set/subset and specific RepresentedVariable
 
 
3122 - 497(Mantis): Need a means of indicating if question translation was done for data collection or not
Translations need to be maintained independently of the source.
Use a controlled vocabulary in TypeofQuestionGroup
3526 - Other, please specify" in QuestionGrid
a. In progress
b. Look at examples - requirements and based on that explore additional grid like formats that have a more open use.
Create new issue to draw up a 'Demand for a Generic Grid structure'
 20160721

ATTENDEES:

Achim, Dan S., Jeremy, Wendy (Hilde paper)
Questions in 3.3
  • Why are only some representations
  • [Instance]Variables and where they come from (question source, variable source, generation instruction, on the fly binding) - simplify and clarify
  • Questions should reference the RepresentedVariable to clarify the flow of information even when specific bindings are not being done
  • You need to be able to do this from multiple response questions - should these forms of bindings be associated with representation
  • Address how data is actually captured (ABS scenario where each item in a code list is a binary response)
  • In some cases we need references to the variable and some to specific representation or value in representation.
  • Need to differentiate support for documentation and for production/automated processing
  • Documentation: If you are doing X use A, if you are doing Y use B, etc.
  • Separating out types of Question Grids and their uses
  • Embedding different formats in StructuredText in DDI - problems of the XHTML in the string, problem of users of Latec, Markdown, etc. The thinking was to be able to control it, but the issues is that people have difficulties with that or would like other formats. How do you embed XHTML in RDF? They have in theory some of these constructs. There is something in OWL but its not often used so what is the value of this? Some people like to have the set way of separating content in lists, headers etc. There might be a solution of removing the XHTML within the standard. Caused problems with autoclass generation.
TO DO:
  • Think about the multiple response issue and how to solve it
  • Pull out the JIRA issues; whats been resolved; what needs to be reevaluate
  • Additonal managed representations options
 20160630

ATTENDEES: Wendy, Jon, Achim, Dan S., Jeremy, Dan G., Flavio

Update on RDF Vocabularies

  • Reviewed list of information needed to get RDF vocabularies out for review
  • Links will be to DDI Alliance site pages www.ddialliance.org/Sepcification/RDF/Discovery    ../XKOS
  • Get current XKOS version from Franck as well as link to implementation
  • DISCO - Missy is an implementation
  • Verify Franck availability during review period.  Franck and Dan G. will do XKOS follow-up
  • Thomas Hartmann and Benjamin Zapilko will do DISCO follow-up
  • Review will be for 8 weeks with reminder at mid-point to get comments in
  • Double check to see who the announcement comes from  (TC and/or Director)

Update on Q2 work

  • Patterns issues:
    • if we go back to the previous way should not cause a major problem
    • How we use the pattern and how its realized - we need to be minimally happy with it
  • Statistical Classification needs to be redone, Data Description, Format is more preliminary (put out with caveat of only CSV and Fixed), Agent
  • Could take 3 weeks instead of 2 - Can we get some of this through as a flow for the purposes of documentation?
  • List of documents that need review for accuracy - make a google doc to work on with sections
  • A thing we could call out is whether the approach to patterns make sense

Preparation for discussion on Questions in 3.3

  • How question grids are currently used
  • The user defined grids (generic grids) - arrange questions and structures that they want to (people want to use it for everything)
  • Describing more complex grids - how is this contrary to multi-mode
  • Separation of reusable question / its use; ties up with the clarification of the process pattern (3.3 vs 4)
  • Separation of display and content
  • goal is to get 3.3 decision with a mind to 4
Meetings in July
  • July 7 - RDF update if needed
  • Other meetings will take place only if we have topics - Q2 work primarily


 20160623

ATTENDEES: Wendy, Larry, Flavio, Jeremy, Jon

Reviewed documents on Cardinality Rules and revisions to Identifiable and Annotated Identifiable

Finalized documents and sent to MT

 20160616
ATTENDEES: Wendy, Achim, Dan S., Jon, Larry, Jeremy
Recommendation: change Annotation to ComplexDataType
When you are annotating there are two options
ONE: all content as first order properties
versionResponsibility
versionDate
...
title
creator
...
TWO: retain Annotation as a ComplexDataType
versionResponsibility
versionDate
...
annotation
      title
      creator
       ...
  
Discussion:
  • Use of Annotation as other than annotation of the metadata class; i.e. citationOfDataFile  Datatype=Annotation
  • Probably would like to retain Annotated as a ComplexDataType

Discussed (no real decision):
BibliographicCitation returned to Annotation to support a means of providing a string for a published citation by the provider
Recommendation that properties from current AnnotatedIdentifiable be part of Identifiable
What should be AnnotatedIdentifiable?
  • Things that have intellectual content and need to be discoverable in their own right should be AnnotatedIdentifiable
  • Example: A CodeList would need annotation as would the categories that are being coded are concepts with intellectual content, however the "Code" attached to it only intellectual meaning within the context of the CodeList and if it needs identification should just be Identifiable.
Looked at the questions Why should a class be identifiable? Why should a class NOT be identifiable?
Took notes on original document. Will reframe with recommendations for changes to model and decision model for making a class identifiable or annotatedIdentifiable.
 20160609

ATTENDEES: Wendy, Jon, Dan S., Jeremy, Dan G., Johan, Achim

AGENDA:

  • Summer schedule
  • DDILIFE-3527 Ornulf's paper
  • Update from Sprint and IASSIST

SCHEDULE:

Vacations:

  • Jon - July 2-24
  • Johan - all of July
  • Dan S - June 23
  • Achim - August
  • Jeremy - Sept 29
  • Dan G - August sometime

Activities and due dates:

  • June 9 - 15 Finalize RDF for review
  • June 16 - 22 RDF Vocabulary page set up
  • June 23 - 29 RDF Vocabulary public review
  • June 30 - July 6 Q2 set up pages/JIRA; Q2 request issues
  • July 7 - 13 3.3 question structure resolution; Q2 pre-announce
  • July 14 - 20
  • July 21 - 27 Q2 development release
  • July 28 - Aug 3
  • Aug 4 - 10 Codebook request issues
  • Aug 11 - 17 3.3 schemas
  • Aug 18 - 24 3.2 documentation; 3.3 documentation
  • Aug 25 - 31 3.3 review
  • Sep 1 - 7
  • Sep 8 - 14 Codebook set up pages/JIRA;
  • Sep 15 - 21 Codebook pre-announce
  • Sep 22 - 27
  • Sep 28 - Oct 5 Codebook development release

DDILIFE-3527

  • Ornulf paper - read for next week
  • Most of his concerns are now met - how things are brought into codebook
  • DDI 4 identification
  • Point for instructional material
    • When people are modeling they use the annotated identifiable as a base when they could be either a ComplexDataType or even just an Annotated base or add a relation to Annotated
    • Can TC come up with guidelines for when something is a ComplexDataType, Identifiable, AnnotatedIdentifiable, or independent use of has Annotation
    • We may also want to enhance the identification paper to include the guidelines

UPDATE:

Reviewed Larry's spreadsheet and made a detailed assessment of mapping into model
Jay and Dan did work on Methodology with a scheme on how to relate the designs at various levels of specificity to describe study level (high) to detailed level
Glossary need - particularly in terms of use of the class (can we harvest off terminology from documentation of the use of a class in a view to create a cross-user-base terminology)
DDI-R from Poster Session not currently in repository of R so now we have a number of them but they are all in development stage; which approaches people are taking and what are the limitations
IPython preceded by a tool Jupyter (presentation session) - how could this fit into our idea of documenting code with general/generation instruction
CESSDA projects with lots of presentations with work in progress - There is a group developing some more abstract rules on the technical framework (Johan) and categorization of tools (ranking software regarding usability, maturity, etc.)  (Hossein Abroshan hossein.abroshan@cessda.net) https://wiki.earthdata.nasa.gov/pages/viewpage.action?pageId=49446977
CESSDA Metadata Management Project is working on a DDI profile for CESSDA purposes. Contact Mari.

Would it be useful to have a page where we could track things we'd like to keep track of; there is work on a more comprehensive listing of papers, presentations, etc. that we may want to link into. Bring this idea into the group working on this (Achim)
ACTION:
Achim will draft points for a note to Mari from TC regarding CESSDA DDI profile work and Wendy will send from TC.



 20160519

ATTENDEES: Wendy, Larry, Dan S., Jeremy, Jon

DDILIFE 3525/3526:

Discussion centered on both the specific issues and a more general review of Questions and Question Grids in 3.3. In particular we are trying to limit backward incompatible changes, but at the same time want these to work at least for the known/general use cases. Dan S. is doing a review of how and where particularly grids are used so we can define a set of use cases we need to meet. These will drive our review of the Question and Question Grid structures.

We are clear about the problems presented and will include these in the general review needs.

ANNUAL REPORT:

Identified items to be included in annual report. Wendy will draft and send around for final comments or additions.  Included were items focused on last year and those that would be focused on next year.

Noted meeting with Sanda and Taina in Bergen. Any Controlled Vocabulary issues that others would like raised should be sent to Wendy before 31 May.


 20160512

ATTENDEES: Wendy, Larry, Dan S., Jeremy

DDILIFE 3525 tabled until next week

TC review of MT materials for review

Since that review we come up with the idea of platform specific ADD TO list on Status Report:

 https://ddi-alliance.atlassian.net/wiki/display/DDI4/DDI+4+Q1+Development+Draft+-+Comment+and+Review?preview=/491690/2064406/Status%20report.docx

  • PSM for each XML and RDF (others when available)
  • Binding rules for each XML and RDF (others when available)
  • Is everything that is in conceptual model being carried forward (PIM to PSM lossless)?
  • Is everything that is in Drupal being carried forward to the PIM?
  • Is each thing in the Drupal Class form being carried into PIM?  Where is the code for creating the Drupal (PHP behind the Drupal)  Visability of tools and scripts - a link and some instructions on how to obtain it  Is this another delivery object?  Could obtain comments on production line?   Not the generation of XMI itself - Check with Olof for location of PIM creation code 

Added to Capability Metrics:

  • Current model ability to cover existing DDI-C, DDI-L, and GSIM classes  
  • Responsiveness to new areas of interest and coverage of those areas

These should be commented on in the status report

REVIEW PAGE:

  • Have people on different content model teams make a list of key things: why things were still done in a specific way; issues still under discussion; issues that resulted in multple possible approaches
  • add this to the Q2 "how to review page"
  • REMOVE Special Issues section and move up the list of related documents (fewer as more of this may be in documentation)

TC 2 week turnaround once delivered from MT to do the Status Report

If anything is frozen we need to be explicit about that - need decision from MT about Freezing/period for Freezing part or all of Drupal

TO DO List:

  • Inform MT and AG about required turnaround time for publishing Development Review package
  • Get decision from MT regarding Drupal freeze
  • Send email to all groups modeling packages that are in the Q2 review (ask MT regarding previously published packages)
 20160505

ATTENDEES: Wendy, Larry, Dan S., Jeremy, Jon

DDILIFE-3525

General comment: we may want to do a users survey and see what kinds of grids are supported then either create a one-grid to rule them all or a set of grids that address simple standards cases and then more specified complex types. This would be good prep for DDI4 discussion. Dan will survey what's being supported.

DDI has always tried to the logical form of the questionnaire to be multi-model and some of this seems to be large screen and paper. A separation of the logical and display when we choose to go into display.

ACTION: Dan will do the survey and Wendy will review the specific issue for next week

Review and update of Task List on TC page

  • Adjusted dates for development reviews (DDI4), RDF vocabularies and 3.3
  • WLT - review Design Principles document against original MF principles list for completeness, bring any outstanding issues to TC next week
  • Set process to transfer an Agency ID - Talk to Jared on writing up a procedure on this
  • Jon will have draft of 3.2 high level document update ready for TC review soon - Once reviewed this can be an expedited sub-minor version change; follow-up would be to work with Training Group on developing content for Use Cases/User Guide

Next weeks agenda:

  • DDILIFE-3525
  • Any work resulting from review work above
  • Process for reviewing material from Modelling Team for publication


 20160428

ATTENDEES: Wendy, Achim, Jeremy, Dan S., Jon, Johan, Larry

Design Principles:

document under discussion 

  • added Conceptualization  to 3rd principle under Capability
  • Discussed possible metrics and further steps

Documentation:

  • Readability score
  • Acronym usage
  • Structure of documentation to reflect principles -- easy to mark as having or not
  • Quality is empirical so there needs to be a measure of use

Design:

  • proportion of classes that carry forward
  • change information
  • time between updates in relationship to the number of changes
  • metrics on JIRA issues
  • is there a way to use labels in JIRA to identify an issue with a design principle
  • is there a way to tie up usage of DDI to responsiveness principle?
  • complexity - how deeply nested are classes in the model - gratuitous complexity
  • compatibility with previous versions

Capability:

  • ?

ACTIONS:

  • Give to MT but also present to Scientific Board for support and further metric development
  • Compare design principles to design principles to see if they conform
  • What are the consequences of non-compliance
  • Compare these against the earlier design principles

Functional Views documents - See DMT-59, DMT-34

PHDD comparison and Public Review: discussion

  • Does it make sense to put PHDD into public review and possible eventual publication?
  • Data Description group will review content.
  • Core features to describe a CSV should be done in common ways
  • Fixed record length data
  • A case with multiple records
  • Rich set of validation features and to deal with messy CSV data
  • Put out for public review, raise the issue and ask for comment
  • There is quite an overlap so a bit inclinded not to go out for public review - lh
  • There are some differences such as multi-record cases, tables for egyptian spread sheets
  • Overall W3C is richer
  • PHDD and Disco are intermediate specs
  • Is there a means of interacting with W3C to expand spec - their scope was strictly CSV files; we've missed their comment period
  • Maybe just focus on DDI4
  • Would it be better to amend the extend the existing W3C
  • If they are using the RDF recommendation they could just use the W3C tools
  • We could go into public review to collect comments and get in touch with W3C community and hold the decision to publish
  • Who would use PHDD in the interim? -- Possible users would only be in the DDI community who are using legacy data and RDF
  • Raises the question of how we absorb other standards - this is a major usecase for use of other specifications in DDI4 (on the model level and on the RDF level)
  • We have two or three solution paths on the table but they have been parked
  • Timing: we need to be in a position that at Dagstuhl we have a firm decision to take forward. Can we have a side meeting in Norway to frame the open questions and issues.
  • W3C approach: 2 independent implementations that are interoperable - subsets of pieces of releases
  • Public review: strategic use of review to get in touch with W3C people, need to frame the issue in a paragraph for the public review
  • When we look at the two they have a lot of overlap one would lean towards the W3C because of buy in. You'd end up dealing with the other pieces as extension of W3C.
  • Probably just take XKOS and DISCO into review.
  • Review of PHDD would take up time.
 20160421

ATTENDEES: Wendy, Johan, Larry, Jeremy

AGENDA:

  • DDILIFE-3521 added comment and recommended solution  Agreed to solution
  • Getting response on 3521, 3513, and background report on 3524 to wrap up this work  
      • 3521, 3513 agreed on  
      • 3524 still needs a few weeks - goal to get all this done before IASSIST
  • Is there a decision yet on publication of PHDD vocabulary?  (which alternate vocabularies are seen as negating the need for PHDD?)  
  • Agenda items between now and IASSIST (28 April, 5, 12, 19 May)
    • Check to see if done:
      • Review and re-disseminate the documentation regarding approval of DDI Agency ID submission (2016-01)
      • Update of Dublin Core files in 2.5 (early in 2016)
    • Design Principles document from Dagstuhl 2015 - review of document and measurable assessment of materials from Modelling Team (2016) -- sent out put on next weeks agenda
    • Review of How make a View document  
    • Annotation - how far do you go with that
 20163031

ATTENDEES:  Wendy, Michael Iannaccone, Achim, Jon, Dan G., Dan S., Larry, Flavio, Jeremy, Kelly

AGENDA: URL issues; DDILIFE 3511, 3516, 3513, 3521; RDF review

URL issues (with Michael Iannaccone from ICPSR)

Michael: Two things need to be resolved Is the current server compatible or be made compatible If the URL's can be made compatible with Druple

Michael will investigate Needs a set of all files we need to serve so we can ignore all of these paths He needs to test

Achim and Dan will provide URL for current files to Michael

The service they are on right now is $50/year and has been a reliable plateform, but it may not provide a lot of support An option at $32 a month would provide the support needed Possibly hosting on ICPSR server rather than U of Michigan - needs to look into

RDF Review:

XKOS is ready to go some delays with DISCO Achim will let us know when there is a firm date

DDILIFE Issues

3521 - ask Johan to respond

3511 - need response from Guillaume, Hilda, and Achim

3516 - only for PhysicalInstance

3513 - Time looks good people need to look at Geography

3524 - Dan and Jeremy need to report

Larry added comment to DMT-21 regarding duplication of Unit in Population as noted in today's discussion.

NOTE: Getting response on 3521, 3513, and background report on 3524 should wrap up issues work on 3.3. Jon and Wendy making major advancements on the documentation so we are making steady progress.

Who's going to be at NADDI?

Dan S.,  Jeremy, Wendy, Dan G., Larry, Arofan


 20160310

ATTENDEES: Wendy, Achim, Dan S., Jeremy, Larry, Jon, Guillaume

AGENDA:

Q1-2015 Review Summary

  • OK to release

URL Naming Scheme

DDI 3.3 issues

  • Need to implement: 3507, 3514, 3515, 3518, 3522
  • Decision made / need to implment: 3512, 3516, 3517, 3519, 3523

Remaining status:

  • 3510 Dan is waiting for remaining items to be resolved prior to providing a schema for review
  • 3511 Need background work - wlt
  • 3513 Need background work - wlt
  • 3521 Need background work - wlt
  • 3524 Table for 1-2 weeks as this is a topic of discussion in some upcoming work for Dan and Jeremy. They will provide background information following their work.

NEXT WEEK:

  • URL Naming Scheme
  • Remaining DDI 3.3 issues
 20160303

ATTENDEES: Wendy, Larry, Johan, Achim, Dan S., Jeremy, Jon

Final Review of DDI RDF Vocabularies: (summary document)

  • Minimum 1 month public review
  • Draft versions will be provided where they already are
  • Only few final issues to be dealt with
  • Issues is PHDD
  • There are other recent vocabularies which seem to cover same content
  • If so recommend withdrawal of PHDD unless it goes beyond, if we have additonal features we can get in contact with W3C
  • Proposed timeline would work well with 3.3 and Q2-2016
  • Continue to use GitHub for review and then move following review and publication
  • Ben and Thomas would be available during review period and May for review and resolution of DISCO issues. Franck would deal with XKOS
  • Issue of the Easter holiday. Add statement in review statement that if you plan to review and have a difficulty in meeting this time frame please contact....
  • Thomas will leave GESIS at end of March and finalize PhD in March/April but is still willing to resolve issues. He will be unavailable after this period.
  • Need to resolve the PHDD issue.

TC deadlines:

  • Text for the release notice for Jared to send out. WT draft up from previous notices.
  • Achim is on all other lists and can send it out in his name.
  • Achim will send out pre-announcement on the RDF Vocabularies list and talk to Franck.

Controlled Vocabularies - Technical structures:

  • How to use CV's in RDF?
  • Proposed SKOS format and transformed Genericode to SKOS
  • Some are hiearchical and SKOS can represent these as well as multilingual definitions
  • Genericode is using a workaround to express hierarchies (dot delimiter); also for handling multi-lingual defintions
  • Base Genericode tools won't understand this
  • Current SKOS representation is RDF-XML
  • This could be used as a standard technical format for future DDI CV's

Proposal:

  • Should have RDF-XML as an additional format
  • In future the SKOS format is the canonical format

Discussion:

  • Need to align this with Custom Controlled Vocabulary package content
  • There are several nice tools for visualizing
  • How to produce these controlled vocabularies
  • Need a production tool to produce SKOS RDF-XML
  • Someone needs to do a sanity check to make sure the custom metadata doesn't have features that are not being captured in SKOS
  • Compataiblity of SKOS controlled vocabularies and XSKOS

ACTIONS:

  • Custom metadata review and compatibility with SKOS
  • Compatibility with XKOS
  • Input tool
 20160225

ATTENDEES: Wendy, Larry, Jon, Achim, Jeremy, Dan S., Johan

AGENDA:
Last set of DMFQA issues - Structure of ComplexDataType Name, Definition and Description usage, movement of Definition from Member to Concept, Inheritance of ConceptualVariable from Concept


ComplexDataTypeName:
Multiple names 0..n
Each name needs a context (optional)


name
content = xs:string
context = EnumeratedValue whatever

  • This is much clearer than using LocalID
  • Place name on Concept and then where it is needed.

ACTION:

  • A group of people are concerned with having a definition and/or description
  • Put definition on Concept and make sure that the others that need it actually get it.
  • DMT-56 has to really think about what it really means to have all these things being concepts as well as linking to concept (expand on this in the final report summary). Where in the Variable cascade it is more meaningful that it Measures a concept. See DMFQA-50, 25, 7 Breaking the inheritance chain between ConceptualVariable and Concept.
  • Review all of them
  • Distinction between definition and description. There are the people filling in the metadata and those who do the searches.

One way to test if this is acceptable is to go with what has been proposed in the table as agreeable and highlight it in the next review. This is the best solution we have and is it workable or unworkable and then go from there. Make sure this get clear in the review instructions. Especially from the search direction. Issue 41 suggested that only one or the other which was rejected and that there needed to be clear documentation and more disciplined use. Write up the two parts clearly.

When a description is used without a definition is the description documentation still appropriate?
 

 20160218

ATTENDEES: Wendy, Larry, Dan S., Jeremy, Flavio, Achim

DMFQA-66,71 (attachments on 71)

Closed and sent to MT as DMT-59. General information provided for instructions to Business Modelers and users. Issues of common content PARKED in the MT.

DMFQA-28,41,53, 38 (attachments on 41

Resolved for recommendation 1 in renaming of Label and more clearly defined role of Name, definition, and description.

ACTION:

  • Review use of definition if moved from Member to Concept, consider where classes receive their definition by a relationship to a Concept.
  • Review structure of ComplexDataType Name as an extension of International String
 20160211

ATTENDEES: Wendy, Jon, Larry, Johan, Jeremy, Dan S., Achim

DMFQA-71, 66:

Walked through and edited document Functional Views.docx on DMFQA-71

ACTION: review lion content in Utility and edit to reflect changes

SPRINTS

IASSIST: Marcel and Michelle will both be available to act as SprintMeister for IASSIST Sprint.

NADDI: Wendy and Jon are organizing. Please send any TC specific requirements for discussion or outputs to the list.

Q1 Review document:

Send comments to list and/or edits to document. Still need to resolve 2 sets of items before releasing (71,66  and Name/Label/Description, etc. issue)

NEXT WEEKS AGENDA: Please add to calendar


 20160204
 

ATTENDEES: Wendy, Achim, Larry, Jay, Jon

AGENDA

  • DDI-LIFE 3522, 3523, 3525
  • NADDI Sprint

DDI-LIFECYCLE 3522

Agreed proposal from Wendy

addition of the following object to DateTimeRepresentationBase and ManagagedDateTimeRepresentation:

DateRangeLimits (0..1) type="Range"
Enter the Minimum and Maximum values of the valid date range in the format provided in DateFieldFormat.

DDI-LIFECYCLE 3523

Deferred discussion, as verbal input from Wendy required

DDI-LIFECYCLE 3524

Deferred discussion, general view was that this is possibly only needs documentation / examples, Seek views of Jeremy / Dan to see if this is a real world problem elsewhere.


Discussion on NADDI Sprint

There are enough people to make the Sprint work, focusing on a review of content and preparation for Q2 Release

Wendy and Jon will work on setup etc

Likely attendance is: Wendy, Arofan, Larry, Jon, Dan G (and possibly Jay)

Remote attendees: Barry, Oliver, Achim (possibly Olof)


Items for Next Week

a.       Deferred items from last week

b.      DMFQA – 71  Views: should they have a basic set of classes?

c.       DMFQA – 66  Functional Views: Usefulness of Functional Views

d.      IAssist Sprint Content 

 20160128

ATTENDEES: Wendy, Achim, Jon, Dan S., Jeremy, Larry  

AGENDA: Remaining Q1 Issues:

Views and content

71 - Views: should they have a basic set of classes?

66 - Views: Usefulness of Functional Views

  • [Wendy will get out a revision of a proposed solution later today for discussion next week]
  • Once we've nailed that down, how does it gets injected?
  • Still don't know why the views need the content If you want management of archival material in the class you need a "study or codebook" object which points to a class of the metadata that describes it.
  • Continue to work on this one
  • Make sure requirements are clear

ClassificationItem, Code and CodeList

52 - Unclear difference between ClassificationItem and Code

48 - So many ways to create a CodeList

  • [should we just pass this on to MT? it may not get resolved until we see the bindings and can test it out.]
  • move to MT as noted

Ability to provide a relationship to a Concept

50 - Review all classes that extend Concept for appropriateness

  • [today]
  • move to MT as noted

Name/Label, Definition/Description Issue

53 - Insufficient documentation for Designation

38 - Chapter 10, Designation, Relationhsip, P147

43 - InstanceVariable, RepresentedVariable, and ConceptualVariable should have semantic "name" property

41 - Redundant properties: definition and description

  • [today]
  • Used Label differently than was used in Statistical Packages. If we are unclear we need to adopt a clear document. the original point of the issue was if we are going to have to give the user a multipage document in order to use this we have failed. Could a description have a type? 0..n
  • use Definition ONLY where it is appropriate. If there are places with Definition that needs a "description" has an Intent - describing purpose and usage
  • InstanceVariable Name -  InstanceQuestionName - be able to specify a name on these
  • Be able to specify a base name on ConceptualVariableName
  • Label remains
  • write up short version and implement - let MT review

Next week:

  • A couple new DDILIFE issues (3522, 3523, 3524)
  • DMFQA-71, 66
 20160121

 ATTENDEES: Wendy, Achim, Guillaume, Jeremy, Jon, Larry

AGENDA:

  • DDILIFE-3511
  • Summary of TC Q1 review

DDILIFE-3511

  • Multiple ways to work with OutParameters so we need guidence and
  • Unclear about reuse - what parts of a question grid can be reused?
  • Addressing an array must be more explicit
  • Versioning of In and Out Parameters is a special case and needs to be addressed.
  • How is the versioning different? The OutParameter is a means of providing input to another process/variable.
  • General issue of reusability of OutParameters. The OutParameter can be different if it parent versions, or if the source context changes.
  • It is a temporary variable and is intended as a pointer.
  • A category is a definition of some text. If it changes it's version changes.

ACTION:

  • Write the documentation completely with a variety of examples address different situation for OutParameters, Grid Types, and versioning situation.
  • Provide guidance for use in all of the exmples
  • Verify the reusable aspect of an OutParameter and and QuestionGirds.
  • Does a 3rd dimension make sense in addressing the content of an OutParamter?

ADDITONAL comments:

  • The more general issue of reusability for Questions: is the QuestionConstruct actually the instantiated Question. In DDI 4 we have the Represented Question.
  • Is language an issue?
  • Is the time span of this too long to put in 3.3? We need to wrap this up in a month (by end of Feb) or it goes to 3.4

Summary of TC Q1 review

  • Finish up in next 2 weeks

Add minutes from 1/14
 

 20160114

ATTENDEES:  Wendy, Larry, Jon, Johan 

DDILIFE 3620, 3504

 20160107

ATTENDEES: Wendy, Johan, Larry, Jeremy, Dan S., Dan G., Flavio

Reviewed DDI Lifecyle issues 3504-3519 all but 3511 are now in process. 3520 will be addressed next week. 3521 was added based on discussions of 3511.

Members asked to review list of recommended changes from xs:integer to xs:noneNegativeInteger by next week DDILIFE-3504

Try to wrap up outstanding issues next week.

Agenda Topic Index Key

Pre-2016 Minutes Page

Agenda TopicMeeting Date
Alliance Roles/Activities

20160908

20160901

20160818

Atlassian Tool User Access

20150611

20150212

20150129

20150122

Bitbucket Repository20151105 
Controlled Vocabularies

20160915

20160303 

CSPA LIM20150423
Design Principles20160428
DDI 3.2 and GSIM20140522 
DDI 3.2 Documentation

20140821

20170202

20140619

DDI 3.3

20170216

20170209

20170112

20161117

20161110

20160922

20160908

20160901

20160811

20160804

20160721

20160630

20160519

20160505

20160421

20160331

20160310

20160204

20160121

20160114

20160107

20151217

20150618

20150507

20150326

20150305

20140619

20141120

20141113

DDI examples for users20140417
DDI Licensing & copyright

20150514

20150423

20150416

DDI4 Comment Page

20150219

20150129

20150122

20150115

20141113

DDI4 Documentation

20150611

20150514

20150416

20150326

20150305

20150212

20150205

20150129

20150122

20150115

20141218

DDI4 Issue Review Q2

20270202

20170119

20170112

20161222

20161110

20161103

DDI4 Issue Review

20160225

20160218

20160211

20160128

20151217

20151119

20151112

20151001

20150924

20150917

20150910

20150903

20150827

20150820

20150813

20150806

20150723

20150716

20150709

20150702

20150325

20150618

DDI4 Modeling Guidelines

20160623

20160616

DDI4 Presentation - Town Hall20150423 
DDI4 Production Process

20150514

20150430

20150212

20150115

20141002

DDI4 Q1 2015 release issues

20150521

20150430

20150423

20150416

20150402

20150319

DDI4 Q2 2016 release issues

20160929

20160811

20160630

20160512

DDI4 Versioning and Binding

20150521

20150430

20150305

20150226

20150219

20150212

20150205

DDI4: TC issues list20141211
DDI4: TC role in review

20141120

20141113

20141002

20140821

20140417

Expression language20130801
IASSIST Sprint20140522
JIRA implementation

20160915

20150430

20141120

20140522

20140515

20140417

20130801

Modeler Issues: Collections / Classification

20140918

20140717

Modeler Issues: Core content20141522
Modeler Issues: Identifiers

20140703

20140626

20140612

Modeler Issues: Name, Label, Description

20140626

20140612

20140515

20140417

Modeler Issues: Q1 Content20141211
Modeler Issues: Reference

20140703

20140612

20140612

20140515

20140417

Product URLs

20160331

20150219

20150205

20150129

20160303

20150122

RDF Vocabularies

20161222

20160630

20160428

20160331

20150212

20150115

20140515

20140417

SDI (Weighting, Sampling, Questionnaire Development)

20140821

20140515

TC confluence site

20150319

20150205

TC Priorities/Activities Scheduling

20160609

20151210

20140619

20140410

 Town Hall - informational 20150423
  • No labels