2012Jun20

From FilteredPush
Jump to: navigation, search


Etherpad for meeting notes: http://firuta.huh.harvard.edu:9000/FP-2012Jun20

Agenda

  • Meeting Date and Time OK?
  • Report from SPNHC
  • Collaborations
    • Kepler
    • Specify/Symbiota
    • NE Vascular Plants TCN
    • SCAN TCN
  • Develop Botany 2012 Goals
    • User Tests and Debugging for Annotation Processor.
    • Develop Demonstration Plan
  • Review Roadmap

Critical Path Things to Do

  • Find Duplicates Use Case (Supporting NE Vascular Plants TCN)
  • Remote Data Capture Use Case (Supporting NE Vascular Plants TCN)
  • Core Infrastructure
    • Delivery of annotations to interests based on related data not found in annotation Requirement20
    • SPARQL Push messaging system
    • Multiple node network with no single point of failure
    • Lightweight annotation system (Supporting iDigBio, Morphbank, Symbiota)
    • Network configured to encrypt all data in motion Requirement14
    • A second case of a Driver
  • Quality Control Data Use Case
  • Deploy ApplePie network (Supporting NE Vascular Plants TCN)
  • Deploy SCAN network (Supporting SCAN TCN)

Refactoring Targets

  • Ontological Representation of DarwinCore
  • AnnotationProcessor/ClientLibrary/Driver
  • Facades for Jobs,JobRunner,NetworkCapabilities allowing development/testing outside Glassfish.

Botany 2012

  • User Tests and Debugging for Annotation Processor.
  • Develop Demonstration Plan

Reports

  • Paul
    • At SPNHC
    • Updated Roadmap to indicate things not accomplished for SPNHC, and expand list of potential next step targets.
    • Worked on Kurator proposal with rest of team (Saturday).
  • Bob
    • Tinkered with refactorization of SPQARQLPush
    • Developing SPARQL queries to represent interests about ApplePie use cases, e.g. InsertDetermination
    • Consulting with David about Fuseki
    • Kibitzed on Kurator proposal with rest of team (Saturday)

Notes

Filtered Push Team Meeting 2012 June 20 Present: James; Heather; Paul; Bob; Jim; Lei; David Agenda

  • Meeting Date and Time OK?

July 4th is a Wednesday, Canadians have the 1st off, Otherwise OK.

  • Report from SPNHC

James: Lots of "Can I use this tomorrow?" questions. TODO: Add audio track to video, upload to youtube channel. James to try recording audio track, Paul to add to video, Lei to make channel link available. TODO: Jim to send Paul contact at Smithsonian Center For Astrophysics group, Paul to contact about more elaborate video.

  • Collaborations
    • Kepler

Bob: Bob and Jim working out contacts for Designers. Jim: Need clarification on some of the notes.

    • Specify/Symbiota

Nothing further yet.

    • NE Vascular Plants TCN

Paul to Invite Patrick and Chuck for next week's meeting. Need Software engineering discussion with Ed Gilbert about colaboration in Symbiota.

    • SCAN TCN

Confirmed meeting dates in August. Time to book travel.

  • Develop Botany 2012 Goals July 7-11, Talk on 11th. Workshop on 12th
    • User Tests and Debugging for Annotation Processor.
    • Develop Demonstration Plan
      • 15 minute talk using screen shots
      • 20-30 minute live demo at workshop (like at SPNHC)

James: Asked by iDigBio to give (1) 15 minute talk, and (2) be part of a one day workshop (about 30 participants) to give a FP demonstration. Demonstration much like SPNHC. TODO: Get Specify installed on James' laptop, and test MySQL over SSH tunnel from his machine. Paul: A few minor bugs and some minor feature requests from demonstration, will file bugs. Add fillin for highertaxon/rank on mapping. Date issue. Leave oauth error handling to next week when Maureen gets back. TODO: Paul to stop in with Jim this Afternoon to check in authentication. James: Could use further help on UI development.

  • Review Roadmap

Jim: Timelines for TCN deliverables? Need to develop these and work out how to use funding. Need to look at burndown rate. Integrate all of these into the planning process. James: Ontological representation of DarwinCore. Bob: Following the community (e.g. Joel's TDWG task group), not too much formal except DarwinSW. Roger Hyam's TDWG ontology hasn't had traction. James: Should we add in other knowledgable people? Bob: We are pretty close to being correct in our claims of agnostisism. David doesn't have much DarwinFP left in his. Mostly need to pay attention to activity on that list. Perhaps we need to discuss some of the technical requirements and restrictions they are placing in more detail with them. Paul: Current state? Bob: Not Roger Hyam's TDWG ontology, rather DarwinFP started with Darwin-SW, removed some restrictions, and added some vocabulary from TDWG ontology. Paul: Specific issue, Georeference class needs DwC bug, if John W hasn't filed one yet. More General, evaluate DarwinFP against AppleCore. Roadmap: Critical Path Things to Do

  • Find Duplicates Use Case (Supporting NE Vascular Plants TCN)
  • Remote Data Capture Use Case (Supporting NE Vascular Plants TCN)
  • Core Infrastructure
    • Delivery of annotations to interests based on related data not found in annotation Requirement20 (Bob working on, evaluating SPARQL Push).
    • SPARQL Push messaging system
    • Multiple node network with no single point of failure
    • Lightweight annotation system (Supporting iDigBio, Morphbank, Symbiota)
    • Network configured to encrypt all data in motion Requirement14
    • A second case of a Driver
  • Quality Control Data Use Case

Roadmap: Deployments

  • Deploy into iDigBio appliance infrastructure
  • Deploy ApplePie network (Supporting NE Vascular Plants TCN)
  • Deploy SCAN network (Supporting SCAN TCN)

Roadmap: Refactoring Targets

  • Ontological Representation of DarwinCore
  • AnnotationProcessor/ClientLibrary/Driver
  • Facades for Jobs,JobRunner,NetworkCapabilities allowing development/testing outside Glassfish.

James: Priorities, ApplePie and TCN engagement gives us nice traction, how do we leverage this. We need something that works. Discussion: Lightweight annotation system would be good something to deploy that works. TODO: David and Bob to produce a brief description of VM capability and software requirements (SPARQL Push, Fuseki, etc) to deploy an instance. Paul: How fragile to changes in domain ontologies? Bob: Not fragile. David: Domain ontology removed from Java code, all defined in configuration. Paul: What are the implications of deploying, using, and then changing the vocabulary? Bob: For development, about 2 hours of configuration. For use, makes questions more complex, are questions (interests) to reflect only old interests, or only new, or both? A specific risk is that existing interests would not pick up new annotations. Phrase other way around, when someone expresses an interest (a SPARQL query), they want a notification whenever a new annotation is made that meets that query, plus the first time you subscribe, all existing matches to the query. What do you expect to happen when you register a new interest with respect to things in the old vocabulary (will depend on reasoning linking vocabularies). James: A few higher level suggestions, coming from Maureen. Meetings taking too much time. Good to have specific weekly targets. Paul: Good to tighten up Wednesday meetings, down to 1 hour. Bob: Refactor meetings to put deliverables discussion that programmers need to be involved in first, then research goals later. James: Would be good to have more control of meeting software. Jim: Kansas/Minesota/Arizona team discussions used effective system. Bob: evo.caltech.edu http://evo.caltech.edu/evoGate/ Paul: Will give another try at using roadmap to develop weekly targets.