TCG F2F Meeting - 2017-05-14 in Shanghai, China

present:

Patrick Dowler - PD
Mireille Louys - ML
Brian Major - BM
Giuliano Taffoni - GT
Marco Molinaro - MM
Carlos Rodrigo - CR
François Bonnarel - FB
John Swingbank - SW
Kai Polsterer - KP
Cui Chenzhou - CC
Mark Allen - MA
Pepi Fabbiano - PF
Mark Taylor - MT
Francoise Genova - FG
Pierre Fernique - PF
Laurent Michel - LM
Mark Cresitello-Dittmar - MCD
Markus Demleitner - MD
Matthew Graham (Over WebEX) - MG

action items from last telecon

• MM to do something with registration/profile page and email TCG with options [defer to f2f]
• MM to automate doc page on submission [ip]
• ALL to review "how to publish" page [defer to f2f]
• WG chairs - recommend move old/dormant specs out of main repo listing, also xsd files [attempted by BM, ongoing]
• MM: inform ivoadoc exactly what to do [ip]
• MCD: review by application to specific models seems to be working [ip]

new action items

• MG: send SODA and DALI to PF for exec review
• FG to introduce discussed changes (notes below) to the errata process
• PD to update TCG page with Endorsed Notes and Errata sections.
• FG to update the standard and RFC template to remove chair names and change vote section to a table
• MA to discuss updating the main IVOA page before future interops for better advertising
• MD to ask document coordinator to add the TAPRegEx errata to TAPRegEx landing page
• MD to add the TAPRegEx errata to the TCG errata section.
• MD to announce the TAPRegEx errata to the IVOA community
• PF to ask document coordinator to add the errata to VOTable landing page
• PF to add the VOTable errata TCG errata section.
• PF to announce the VOTable errata to the IVOA community
• MM to ask document coordinator to add TAP the errata to TAP landing page
• MM to add the TAP errata TCG errata section.
• MM to announce the TAP errata to the IVOA community
• MT to update the Publishing in the VO Page to link to the new IVOA Validators Summary page.
• MM to try to rename the Publishing in the VO Page (remove New)
• All WG/IG chairs to fill in the 2017A roadmap by the end of May.

Introduction

CC: introduction and welcome

MA: Standards that meeting priority of multi-D are complete. Effort to have more scientists involved now. MG: Science engagement has become more general. We have made progress with current science priorities. Try to identify new science goals now.

Agenda review:

• FG: Suggest we go through the consequences of the new document process.
• PD: Yes, added to the late morning agenda
• PD: New item for agenda: David Schade has asked to bring up idea of process of standardizing an external model, in this case the CAOM data model.
• MG: SSO document has same requirement. Added as additional agenda item for this afternoon.

Review of standards:

• 5 standards are for exec committee to review
• FB: SODA and DALI * have received all approvals from TCG.
• PF: please send those two to exec as well
• ACTION: MG: send these to PF for review
• MG: Status of vo dml?
• MCD: Expects it ready any day now, maybe during the interop
• MG: Final version?
• MD: Maybe go through the document one more time for details?
• MCD: Sounds reasonable
• MG: Brian, what is the state of the vospace rfc?
• BM: A long document, but have revived feedback from 2 chairs already
• MG: Keep in mind this is 2.1 not 3.0
• MM: TAP 1.1 and adql 1.1 coming in the near future
• MG: Can ops comment on the impact of having many new revisions right now
• MT: Doesn't feel that the community is too aware of these new versions. doubt there will be any issues. from ops perspective, will not be a concern. many are minor versions so don't impact compatibility.

Status of standards impacting various WG/IG:

• MCD: Much implementation and examples work happening on various data models and vo dml annotations. working out particulars. documents are mostly up to date. on the mapping side: new syntax is being implemented. tom d and omar are working on sample files and serializations. laurent working on healpix. getting good feedback from a science point of view.
• FB: Long discussion about time series proposal. do you (mark cd) see this impacting the models?
• MCD: Mild effects from those comments, but have been addressed.
• PD: ND Cube model: originally separated minimal requirements. has any of the work with models looked at supporting future soda versions?
• MDM: SIA 2.0 was kept in mind. will have to work together for sia/soda next versions
• FB: Time domain now the DAL main priority.
• MA: Minimal things allowed us to reach a milestone. new versions will need to come from CSP requirements which aren't there yet. must come from real projects and use cases
• MCD: Have tried to not go too far ahead
• MA: Need to look at how time domain fits into the data model
• MG: Agree with mark a. since the data model is stable. next address multi-d in data model, then time series. a reasonable schedule?
• MCD: Dependency on stc coordinate system. it's relatively stable, but changes will have impact.
• MD: No rush on new data model aspects. Many things happening at the same time. Priority should be getting STC right.
• Conclusions: we need feedback from community before going to sia 1.1 or soda 1.1. Make the most of feedback from science group feedback from South Africa.
• MG: To MA a: can we close this chapter of supporting the basic use cases?
• MA: Yes.
• PF: Agree with Mark A, but keep the pressure on to get cube model work done.
• FG: Must be sure we do this data model work well. Need to get STC right.
• PD: Goal is to property annotate VOTables.
• MCD: Yes, that is the reason we are focus on example implementations.
• MCD: Need to make this a milestone before moving on.
• MD: Yes, for time series, but not for ND cubes. Driver of ND cube requirements should be time series.
• MA: Do we have examples in the past on how to iterate over this model?
• KP: ND cube can be used for time series. But perhaps we consider other techniques for time series.
• FG: Thinks it essential that Applications is involved with STC.
• MG: We should advertise that we have reached a milestone, but clearly more work to be done.

UTypes and VO-DML:

• MCD: VO-DML is mostly accepted so considered nearly done. Mapping is being tested with examples with time series, etc.. Need more people trying to use it.
• FB: I question the word 'annotation' for the mapping work, because this is for data exchange.
• MCD: Believe it is a broad enough, but might agree that it is not the best wording.
• MG: Should we address the UTypes now that the mapping is progressing?
• FG: We should discuss first.
• LM: there is a session in DAL on mapping with room for discussion on this topic.
• MCD: Would like to hear about specific use cases/implementations.
• FB: Not possible to approach this yet--DAL has many references to UTypes.
• PD: Need a transition period, so certainly not possible to change existing standards.
• ML: VO DML Mapping: believe the mapping can be the core as to how we develop the mapping but also, what kind of simple representation can we bind? For example, we can ask: can we bind simple UTypes in the provedence model? We can use the XML to see how we can handle these mappings in other standards.
• MD/MCD/FB: This is all true already. For IVOA, VOTables are a priority, but it should be able to be used with other types of data too. The VO-DML standard and VO Mapping are not coupled.
• MG: Ensusre we make clear the purpose of VO-DML when we announce the RFC.

Time Domain:

• FB: Asterix meeting in Strasbourg with long discussion on time domain. Use cases for time series discussed with various groups. This will be reported on during the DAL sessions.
• MCD: Jiri created a proposal on how to use the cube model to represent time series data in February. This was valuable in determining where changes needed to be made. Worked well, only annotation level changes possible remaining.
• JS: Really good to see that time series is getting traction.
• JS: Notice that there isn't representation from the large projects at this interop.
• FB: There is a talk on time series with regards to Gaia
• MG: Would like to see some big project examples and how they deal with the complexity of the standards.

Documents Standards 2.0:

• FG: A new template has been built for TCG comments to make things more clear.
• FG: We need an errata link from the main page of the standards documents. Giulia has provided an example.
• MM: perhaps have a running number that represents the number of errata present for the applicable document.
• PF: Or perhaps individual links for each errata?
• MM: There have been comments that we should not make it too complicated.
• MT: make it more visible
• MCD: Show that there are no errata when that is true
• PF: Should the authors be listed in the errata?
• All agree that:
  • A blank errata page is created with all standards
  • Date to be displayed of last errata (or say 'no errata')
  • Make the link more visible with non-clickable text
  • Author(s) to be added to errata
• ACTION: FG to introduce these changes to the errata process
• FG: Where to keep the list of errata? TCG page seems like the proper place, can be managed by the TCG.
• ACTION: PD to update TCG page with Endorsed Notes and Errata sections.
• PD: One errata per version or or one page with all errata?
• MM: An errata page will have links to the details of each errata
• FG: The TCG vote and comments section have been separated
• MD: Are the names of the TCG chairs necessary?
• Group: No
• FG: On the RFC template, replace the chair names with links to the working group home pages.
• FG: will change the 'vote' section to be a table
• ACTION: FG to update the standard and RFC template with this feedback

Interop Programme Review:

• Apps & DAL: * not enough room for all the people who wanted to contribute
• Data Model:
  • Hoping to have key participants over webex.
• MT: Not clear that what the interop meetings are about and how to contribute
• ACTION: Mark Allen to discuss updating the main IVOA page before future interops
• Time Domain:
  • Anticipate an eventful interop because of LSST participation
• MA: CSP has introduced concept of Lightning Talks. Will be looking for more participants.
  • General Discussion of how to involve the science community in the Interop Meetings
  • Discussion of how to share information/attendance between conferences such as Astroinformatics, ADASS

Errata Review:

TAPRegEx 1.0:

• MD: TAPRegEx Errata Summary: Allow anyURI instead of stdID in TAP Registry data model identifiers.
• TCG endorses this errata
• ACTION: MD to ask document coordinator to add the errata to TAPRegEx landing page
• ACTION: MD to add the errata to the TCG errata section.
• ACTION: MD to announce the errata to the IVOA community

VOTable 1.3: COOSYS un-deprecation

• PF: need to reintroduce COOSYS. Example (in 1.3 format) on the errata page. Some collections have already added this field back.
• MD: has already reapplied COOSYS.
• PF: Multiple COOSYS entries are allowed in one VOTable
• TCG endorses this errata
• ACTION: PF to ask document coordinator to add the errata to VOTable landing page
• ACTION: PF to add the errata TCG errata section.
• ACTION: PF to announce the errata to the IVOA community

TAP 1.0:

• MM: Summary: Reshaping the spec to adjust to the common usage of TAP. There are five errata in total.
• MM: Reviewed each of the errata
• ACTION: MM: The final errata (VOTable table usage) is to be deferred.
• TCG endorses this set of errata (with the final one being deferred).
• ACTION: MM to ask document coordinator to add the errata to TAP landing page
• ACTION: MM to add the errata TCG errata section.
• ACTION: MM to announce the errata to the IVOA community

ADQL 2.0:

• MM: Summary:
  • Clarification in BNF grammer
  • Mathematical function text correction
• MT: Item 2 is an error which should be addressed soon and has been encountered (and addressed) by Gaia.
• Errata to be deferred for now

• FG: Suggest that the TCG be able to endorse this ADQL errata by email
• All agree with this suggestion and agree that errata in general can be endorsed by email.

Endorsing external Data Models.

• PD: What is the process? Specifically wrt to CAOM which is currently used by CADC and MAST.
• FG: A process has been discussed where:
  • It is assigned to a certain working group
  • There is a brief RFC period
• PF: HiPS has gone through a similar process in about one year. Thinks it was a valuable process with good discussion.
• FG: If it is only to be given visibility, it can be a note. If it is to be fully adopted by the IVOA, then it should be a standard.
• MG: What is the scope of this particular proposal?
• PD: Not entirely certain, but think that it cannot be done in the same way as HiPS because it goes through revisions more quickly.
• MG: Perhaps it should be a note describing a "best practice" in how to use CAOM.
• MA: Agree with MG with the proposal of a note.

STC:"> Status of STC:

• MG: There is a lack of expertise around the STC standard.
• MCD: Arnold is retiring so need a replacement in the domain expertise.
• MCD: The models need to be exercised in more domains, though the need for this is diminished.
• MG: When will STC 2.0 be ready?
• MCD: In the next semester. It will be more of a components model, less specific to space and time.
• MG: We're looking for experts in space/time coordinates
• MG: What are the dependencies on data cube?
• MCD: Small dependency on STC. Would be a gamble not to wait for STC 2.
• PD: There is actually an implementation dependency on STC, so it must come first and be stable.
• MCD: Yes, that's correct.

Review of "Publishing in the VO" page.

• MT: Can "Validation Tools" links be updated to use "IVO Validators Summary" page?
• ACTION: MT to update the Publishing in the VO Page to link to the new IVOA Validators Summary page.
• MT: Remove the 'New' from the page link?
• ACTION: MM to try to rename the page (remove New)
• ML: Add the year to the publishing page to make it look current?

Roadmap 2017A:

• ACTION: All WG/IG chairs to fill in the 2017A roadmap by the end of May.