IVOA Roadmap for 2013B

This outlines the roadmap for development activities by the various IVOA working and interest groups in 2013/14 between the Hawaii and Madrid Interops.

---

---

Applications

• MOC 1.0: Following discussions between authors a PR has been submitted for publication in January 2014, and the RFC is running from 7 February to 7 March. If all goes smoothly, we hope to achieve REC status near the May 2014 interop.
• Code Repositories: Following a request from the Exec, the Apps WG will start a discussion on use within the IVOA of code repositories beyond those at the project level. The Exec made this request in the context of the discussion in the Time Domain focus session at the Heidelberg (May 2013) interop in which Mario Juric (LSST) expressed his disappointment at not finding a one-stop shop for IVOA-blessed implementations of VO standards, specifically in Python. We will consider the current status and future possibilities for the IVOA and its member projects in interacting with existing repositories (notably Astropy) or setting up new ones, taking into account the desirability, costs and benefits of the various options.

Data Access Layer

• pre-Hawaii Interop status:
  • DALI-1.0 TCG review almost complete, new PR doc with all changes to date has been submitted
  • WD-DataLink-1.0 is late (to be submitted asap, it is missing 2 sections), final issues to be discussed at Interop
  • WD-STC-S.1.0 was posted in Sept, session at Interop
  • no work on TSA-1.0
  • cube access prototypes to be discussed at Interop

• DataLink-1.0
  • WD in October 2013
  • WG review January 204
  • PR document: February 2014
  • RFC in March 2014
  • TCG review in April 204
  • Recommendation: May 2013
• STC-S-1.0
  • this has been deemed not needed for SIA-2.0 and AccessData-1.0; these will adopt simple geometry value(s) without metadata
• SIA-2.0
  • agreed that ObsCore will define the bulk of the metadata returned by SIA query, additional metadata when ImageDM is available
  • WD in November 2013, will include query capability, maybe metadata (get-gory-details) capability
  • implementations in next few months
  • WG review in March 2014
  • RFC?
• AccessData-1.0
  • WD after WD-SIA stabilises
  • implementations need to be done
  • more discussions in May 2014
• TSA-1.0 -- small standard that re-uses SSA-1.1 with the time series DM produced by DM-WG (presumably SimpleTimeSeriesDM)
  • WD: TBD
  • WG review:
  • RFC period:
  • TCG review:
  • Recommendation: April 2014

In order to deliver useful functionality for data cube access in a short timescale, with incremental improvements over the next year, DAL will develop a series of much smaller components that will be the capablities delivered by web services (RESTful web services followng the DALI recommendations). These capabilities are, in order of priority (with relevant standard listed where they are known already):

• discovery of data cubes:
  • SIA query interface (SIA query capability)
  • TAP + ObsCore
• download of complete data cubes:
  • direct link in query response for single-file case
  • DataLink
• get additional metadata (SIA metadata capability)
  • return instance of ImageDM, format TBD
  • needed to perform some of the more advanced data access operations
• data access (AccessData capability)
  • extract subsection of cube (doable without detailed metadata)
  • flattening along one or more axes (sum or avg) (req: metadata)
  • rebinning along one or more axes (req: metadata)
  • precision pixel-level access (req: metadata)

The DAL roadmap is simplified to include the capabilities: DataLink, SIA query, SIA metadata (get-gory-details), and accessData (from DAL) as well as dependencies on the DM-WG for ObsCore and imageDM. We expect several versions of SIA-2.x and AccessData-1.x to be produced before all use cases are satisfied.

Data Model

PhotDMv1-1 1.0

TCG review started on July 2012.

Work stopped due to coupling with utypes Tiger Team work as PhotDMv1-1 was intended to be in line with utypes recommendation. Process of adaptation to the ongoing discussions within Tiger Team.

It is decided (12/05/2013 TCG meeting) to decouple both activities, update PhotDMv1-1 with the comments obtained during TCG review and encourage TCG members to comment on it (to be done before the end of Spring 2013 interop)

PhotDMv1-1 will be adapted for 2013 Autumn interop following VO-DML approach

Spectral DM 2.0

"Trying to be final" version released to the working group in March 2013. No major comments received. To be presented to WG during Spring 2013 interop. Implementation pending to formally start RFC process

Characterization DM 2.0

Stable version released to the working group in September 2012. No major comments received. Implementation to be presented during Spring 2013 interop. Waiting green-light to start RFC process

Image DM

Draft to be presented during Spring 2013 interop. Wider list of authors to be identified

VO-DML

Draft to be presented during Spring 2013 interop. Prototypes and comments to be received between Spring and Autumn 2013 interops

Grid and Web Sevices

* PDL:

• May 2012: first WD: PDL-0.1-WD-20120516
• May 2013: revised WD in 1.0 and status discussion in Heidelberg GWS session 1
• August 2013 : RFC
• March 2014 : end of the TCG Review
• Spring 2014: REC?

* VOSpace:

• Comments and suggestions from the community to prepare the VOSpace WD 2.1 for Madrid
• REC ? : probably in Fall 2014 or Spring 2015

* UWS:

• Comments and suggestions from the community to prepare the WD UWS 1.1 for Madrid
• REC ? : probably in Fall 2014 or Spring 2015

Registry

SimpleDALRegExt 1.0

SimpleDALRegExt has become a REC.

VOEventRegExt

In the context of VOEvent 2.0, a working draft of the VOEventRegExt is currently being finalized for metadata inputs with the VOEvent mailing list group. Matthew Graham is coordinating this effort. The RWG will participate in the RFC process to evaluate the specification in context of the registry standards integration.

Registry Interface and RegTAP

The RegistryInterface 2.0 (RI-2) working draft has been distributed to the RWG for review and is pending revisions for progression to the next phase for RFC. Our RWG goal for this standard is to move forward into the REC process following review and coordination during the May 2014 interop meeting. RI-2 will retain the standard OAI protocol for Publishing Registry and harvesting from the Publishing registry into the Fully Searchable Registry. This RI-2 standard will deprecate the the earlier SOAP interface support and provide 2 forms of web based REST interfaces, APIs, for the application client community: a TAP protocol based on the RegTAP relational registry model standard, and a SimpleRegREST interface. The later is the key to completion of this updated standard registry interface definition. A consolidated approach based on the integration of the Simple REST prototype via the EuroVO registry and VAO registry REST interface will be proposed. This standard RFC process is a high priority topic of discussion for the May 2014 interoperability meetings with direct impact to IVOA science application tool development that has registry interface dependency.

An underlying Registry relational data model, RegTAP, has been developed which describes the core registry metadata to be exchanged using the RI 2.0 standard interface. This document is currently beginning the RFC phase.

Registry Metadata for Footprints

The RWG will continue to have discussions and evolve the current standards infrastructure in support of metadata for footprint descriptions, i.e. spatial regions of coverage. The advancement in the MOC and STC/s standards definition will be evaluated in the context of footprint representation for support of applications in development of data discovery. Future coordination between IVOA groups will involve potential development for spatial indexing in support of application integration with registry collections.

Registry Operations and Curation

The Registry WG continues to drive regular discussions during Interoperability meetings about IVOA Registries operations, including curation of VO Resources and validation of DAL services. The Registry WG is pending effort to develop a coordinated plan accepted by TCG and proposed to the Exec for developing the procedures and processes of curating nonresponsive or non-valid registered DAL services. This also includes consistency in Registry managed authorities and harvested resource metadata across the IVOA registry network.

-- GretchenGreene - 12 February 2014 last update

Semantics

The WG will continue to contribute to other WGs and IGs as appropriate.

* Extension of the UCD List for planetary data:
There are requirements from the Planetary community to use and extend the IVOA UCD list for using these semantic tags while mapping their metadata.
Some concepts , current in Planetary science are missing .
An effort to propose new possible UCD strings and validate them in the IVOA UCD maintenance group is taking place .
Proposed list by May Interop 2014

Data Curation & Preservation

The IG will continue discussions on general topics of interest to the community. Some of the topics we expect to focus on during the next 12 months include:

• The Unified Astronomy Thesaurus
  • Review of current effort described online at http://astrothesaurus.org
  • Participation of IVOA members in merging / mapping similar vocabularies (e.g. the Paris Observatory's Etymological Dictionary: http://dictionary.obspm.fr/)
  • Drafting editors to support the development and review of the thesaurus
• The Research Data Alliance
  • Follow developments of the RDA: http://rd-alliance.org/
  • Participate in working groups on persistent identifiers and data citation

-- AlbertoAccomazzi - 2013-03-19

Knowledge Discovery in Databases

Theory

The main goal of the Theory I.G. is to propose an access protocol to discover and retrieve simulated data.

From last InterOp, several implementations have been tested. We also had a few teleconferences with some members of the Theory I.G.

Search for simulations

Some technical difficulties have been identified. The most important of them, is that, contrary to an observationnal data, a simulated data may require several thousands quantities to be described in a way allowing us to answer our usecases. Because RDBMS are not efficient with tables with more than thousands columns, the access protocol for theoretical data should not rely on specific technologies (as TAP/RDBMS). During the second Theory session in Heidelberg InterOp we will discuss this important topic. A possible solution, would be to define a few APIs with a main one, the SEARCH/CUTOUT function, to discover interesting simulations. The query language would be a subset of ADQL. On the other hand, the exchange format will have to be defined. We will discuss the oportunity to use the serialization of SimDM in VO-Tables following the proposal of the utype tiger team.

Extraction of data

Concerning the extraction of data, the situation is more simple. Services will provide a Datalink document to describe the possible operations on data. This will be presented and discussed in the common DAL / Theory session about datalink.

Goals of this InterOp for the Theory I.G. are : 1) conclude on a solution to discover simulations that is independent on the technology used to store metadata, 2) define the large picture about the exchange format document in the communication client / server during the search phase of simulations, 3) advance on the syntax of datalink document.

Education

The newly formed Edu IG is planning to discuss its Roadmap during the Heidelberg 2013 Interop. The basis for the discussion are the Terms of Reference that can be read in the IG's web page together with the Rationale. At the Heidelberg Interop, the Edu IG will have both a morning and an afternoon session on Thursday May 16. At least one of the two sessions will be devoted to discussions of the Roadmap and its implementation. Contributions from anybody interested will be most welcome.

-- MassimoRamellaINAF - 2013-03-25

Time Domain

The TDIG is working on a note endorsing SimpleTimeSeries as an IVOA-approved format for time series data and also waiting for feedback from the Registry WG on VOEventRegExt (see above). We anticipate participation in the multi-dimensional data discussions. We will issue an update note covering the VOEvent Transport Protocol, and then push that towwards becoming an IVOA-approved standard.

-- MatthewGraham - 2013-05-09

-- JohnSwinbank - 2014-02-11

Standard and Processes

No evolution of the standards forseen in 2013.

Science Priorities

Science priority areas are: Multi-dimensional data, and Time Domain Astronomy. Focus sessions at the May 2013 interop meeting will engage projects and surveys that produce and use multi-dimensional data (including radio astronomy, IFU, high energy and simulation data) and Time Domain data (time series, light curves, transient event reports) in a discussion on the role of VO technologies in these areas. Multi-d and Time Series use cases are being collected as part of this effort.