Difference between revisions of "Discovery Telecon 2011-09-13"

From Earth Science Information Partners (ESIP)
Line 110: Line 110:
 
** recursion
 
** recursion
 
** OGC has a simple view on the kinds of services that wil be deployed
 
** OGC has a simple view on the kinds of services that wil be deployed
** Chris mentions that OGC uses "should" liberally (and "optional" is more heavily weighted to "should")
 
 
* OGC has [http://www.genesi-dec.eu/presentations/OGC_201009.pdf  CSW 3.0 Part 4: OpenSearch Query Interface] (presentation by Pedro Gonçalve)
 
* OGC has [http://www.genesi-dec.eu/presentations/OGC_201009.pdf  CSW 3.0 Part 4: OpenSearch Query Interface] (presentation by Pedro Gonçalve)
 
** CSW 3.0 Part 4 is documented in the spec OGC 10-032
 
** CSW 3.0 Part 4 is documented in the spec OGC 10-032

Revision as of 15:33, September 13, 2011

<< Back to the Discovery Telecons page

Action Items

(As perceived by Eric...)

  • Chris: Will add details to latest whitepaper section about the importance of using general tools (e.g., OpenSearch, Atom)
  • Chris: Will add a section to the whitepaper where Discovery Cluster participants can "sign" the submission
  • Chris: Will add a page to the ESIP wiki extracting some details on the OGC OpenSearch Draft
  • Chris: Will look at the OGC CSW 3 part 4 specification (on OpenSearch) over the next month
  • Erin: Please ask Carol to review the section of the whitepaper on ESIP federation background
  • Ken: Will try to get the ServiceCasting section of whitepaper written (himself or someone else)
  • Hook: Will add section to whitepaper about Discovery Cluster governance
  • Someone: Will need to write the whitepaper conclusion (likely after all other sections are completed)
  • Eric: Try to submit proposal for Discovery convention implementation by Thursday
  • Ruth or Chris: Decide who will attend the charrette

EarthCube white paper status and plan (Chris)

  • Background
    • NSF Earth Cube program
      • "NSF seeks transformative concepts and approaches to create integrated data management infrastructures across the Geosciences. In a new partnership, the Geosciences Directorate (GEO) and the Office of Cyberinfrastructure (OCI) recognize the multifaceted challenges of modern, data-intensive science and education and envision an environment where low adoption thresholds and new capabilities act together to greatly increase the productivity and capability of researchers and educators working at the frontiers of Earth system science."
      • Earth Cube Webinar on July 11, 2011 (Powerpoint presentation)
        • "A need to transform the conduct of science and education over the next decade"
        • "NSF seeks transformative concepts and approaches to create a sustained, integrated data management infrastructure spanning the Geosciences."
  • Current White Paper Draft now on wiki
    • Title: A Grassroots-Federated Approach to Earth Science Data Discovery
      • are the words grassroots and federated redundant?
    • Section on Use of General Tools and Technologies (e.g., OpenSearch, Atom)
      • James: warns that discussing simple, general tools gives some people a bad impression
      • James: recommends that, in discussing general tools, also describe why they are useful
        • Some Reasons: general tools are important to adoption, specific systems based on general tools can leverage the many utilities designed for general tools
    • Authorship
      • should it come from the Cluster or from individuals?
        • Using the cluster as an author downplays individual IP (which is good for this charrette), but all affiliated must consent
        • Having an individual as a first author may give the appearance of agency origins (which is bad)
      • Erin: could also have a section for people who support it can add their name to the list.
    • Outline of grand challenge
      • Explains the commonalities between ESIP and NSF challenges and why ESIP solutions may serve as a template for EarthCube
      • Need Carol to review the ESIP background section (Erin will make this request)
    • Technology framework
      • Chris has contributed section on Federated Search Framework
      • Ruth has contributed section on DataCasting Framework
      • still need author for section on Servicecasting Framework. Ken volunteered.
      • Chris is also writing a section on integrating all the framework together (looking for input from Brian)
    • Example Scenario (written by Ruth)
    • Governance
      • Hook volunteered.
    • Looking to the future
      • expanded approach of Discovery to the overall Earth Science Collaboratory (ESC).
    • Conclusion
      • Someone will need to write the conclusion (likely after all other sections are completed).
    • Other Suggestions
      • Chris: Should there be a plug for the NSF to coordinate with ESIP no matter the outcome?
        • Erin: yes, b/c ESIP wants to have a larger role with NSF
  • Charrette
    • NSF will hold a charrette (community-driven meeting)
      • Now scheduled to be during first week of November.
      • Coincides with ESDSWG on Nov 1-3. (conflict for Hook and Chris)
      • Who will attend?
        • Is Ruth going to Charrette?
          • NOTE: Ruth is not going and will be at the ESDSWG meeting instead.
        • Rahul (UAH) may also be going.
        • James knows someone from OPeNDAP who will attend

RFP for the Discovery Testbed proposal (Hook)

  • the ESIP Testbed serves as a forum for innovative collaboration across all sectors of the Federation by enabling Federation committees and clusters to develop, test and improve availability and accessibility of member products and services.
  • we submitted a proposal to the ESIP Products and Services Committee for a Discovery services testbed.
    • an idea is to have a reference implementation for the ESIP Discovery Cluster related conventions
  • Discovery Testbed Proposal
    • Title: Discovery Services and Clients: Interoperability Testing, Advertisement, and Assessment of Data to Service Broker
    • Description: The Discovery cluster provides a medium for Federation members to coordinate on development, deployment, and creation of interoperable specifications for Discovery services such as OpenSearch, DataCasting, and ServiceCasting. As such, a common testbed to deploy services and/or clients would enable the Federation members to test the interoperability of Discovery services and clients. More over, having a common testbed where Federation members from various data centers can access would provide a common platform that can be used to advertise their services. Currently it is difficult to know where services exist, what they do, and how to access them. Several machine-processable mechanisms for advertising services (such as service casting) have been evaluated by the NASA Tech Infusion Working Group. The testbed will provide an environment to demonstrate how Federation-wide member services can be advertised, described, and chained together using existing technologies. From this, new capabilities can also be explored such as enabling an association of data to services where for each data collection or granule, compatible services provided by Federation members can be shown.
    • ESIP Community Benefits: Improved interoperability of ESIP Discovery services and clients leading to broader adoption. Increased awareness and usage of Discovery services. Improved brokering of data collection and granules to compatible services that can use them.
    • Cost: $8K to (1) setup the testbed for allowing service and client implementations to be deployed onto the testbed, (2) setup a service cast of Federation member’s services offered, and (3) explore concept of data to service mapping.
  • testbed proposal was approved
    • P&S pitched it to the SIP Finance Committee.
    • setup Configuration Board to review the implementation proposals.
  • the committee is now is soliciting ESIP members to propose to implement the functionality.
  • note here that the testbed idea proposal and implementation are not necessarily done by the same people
  • last month, Ken Keiser sent out an email of this.
  • A proposal should be one page or less in length and should specify:
    • Which task(s) the proposer is responding to;
    • The capability of the proposer to carry out the task(s);
    • Any special skills or intended outcomes that the proposer can bring to the task.
  • Proposals
    • may be submitted by any ESIP organization, an individual within such an organization, or a team of such individuals.
      • Civil servants are restricted from receiving ESIP funds.
    • should include a CV for any participants (not part of the 1-page proposal).
    • should be sent to keiserk@uah.edu by Sept 15, 2011.
  • ESIP Testbeds will be more prominent on the ESIP Portal.
  • Architecture of testbed
    • some ideas include virtualization of testbeds
  • Eric Rozell is interested in submitting an implementation proposal.

DCP 3 (superseding DCP 2): Proposal, schedule (Brian)

  • DCP-2: Canonicalizing Granule-level OPeNDAP Links
    • was submitted for review in 2011-06-15.
    • essentially killed as new discussion has brought up more ideas.
  • upcoming DCP-3 will be superseding DCP-2.
  • DCP-3 will be more generalized (addresses more than just OPeNDAP).
  • justification of needing a DCP that adds to the OGC spec. ESIP's needs for OpenSearch goes beyond OGC's spec.
    • adding esip-specific qualifiers for rel types (e.g., type of enclosure)
    • recursion
    • OGC has a simple view on the kinds of services that wil be deployed
  • OGC has CSW 3.0 Part 4: OpenSearch Query Interface (presentation by Pedro Gonçalve)
    • CSW 3.0 Part 4 is documented in the spec OGC 10-032
  • within the next month, Chris will write a ESIP position paper on comparison of ESIP and OGC position on OpenSearch.
  • we should get a better understanding of CSW 3.0 and OGC OpenSearch first, before pursuing DCP-3.
  • Brian will finish table of IANA rel values for DCP-3.
  • Chris willing to help on writing DCP-3.

Comparison of ESIP Federated Search with OGC OpenSearch

  • The OpenSearch Query Interface was approved by the OGC Technical Committee to use the Fast Track process to become an OGC Candidate Standard.
  • In July 2011, there was a public Call for Comments as part of the fast track process.
  • OGC documents are difficult to access
    • it took Chris a few weeks to get access to it.
  • OGC authors have many recognizable people
    • Pedro Gonçalves
    • Andrew Turner
    • Doug (FGDC guy)
  • "optional" in OGC reads more like "should"

Attendees

  • Chris Lynnes (NASA GSFC)
  • Hook Hua (NASA JPL)
  • Erin Robinson (ESIP)
  • James Gallagher (OPeNDAP)
  • Jeff McWhirter (UCAR)
  • Ken Keiser (UAH)
  • Eric Rozell (RPI, Discovery Cluster Fellow)
  • Brian Wilson (NASA JPL)
  • Mike Daniels (Manager of the NCAR Earth Observing Laboratory Computing, Data and Software Facility) ?