Listing HAPI endpoints in IVOA registry

[BaptisteCecconi]

Registering HAPI endpoints in the IVOA registry would increase the discoverability of HAPI data streams, and expand the use to other communities in astronomy.

We have discussed this point several times in the past years.

How can we present the case? I think there are three elements:

• First, the IHDEA is willing to work closer with the IVOA (recommendation of using VOUnits, collaboration with the Semantics WG...). So, strategically, that would be a new pathway for closer collaborations.
• Second, there is an IVOA Note describing a "simple time series" access protocol, but it never went any further. The HAPI protocol could be a proposed solution. This probably implies to review the context and requirement in that note, and see how we could accommodate those within HAPI.
• Third, the Solar System IG (heliophysics-related) teams are already using HAPI endpoints to publish data streams and are referencing them in EPNcore table (in their Datalink metadata).

Let me summarise what I think could be done to achieve this goal:

• To be listed in the IVOA registry, we need to get an IVOA Standard Id. This IVOA Standard Id needs to be present in a document endorsed by the IVOA.
• Hence, we need to contact the TimeDomain and SolarSystem Internet Group (IG) leads and present our case.
• Then there are two paths to explore:
  1. We manage to convince the Technical Coordination Group (TCG) that the IHDEA and COSPAR have adopted HAPI and we are willing to expose endpoints in the Registry.
     This would be the simplest path, but it is not clear how we connect the HAPI specification to the IVOA process.
     Note: we will need to mention the ivo-id of the HAPI standard in the specification, or in a document related to the IVOA.
  2. The IG leads will discuss with the Data Access Layer Working Group (DAL WG) for proposing a new recommendation to the Technical Coordination Group (TCG), which will eventually endorse the recommendation.
     In this case, we would need to go through the whole formal IVOA recommendation review process and we would also need to adopt the IVOA document template. The endorsement process takes time, more time than our versioning time scale. I don't know how we can keep the current HAPI specification and the IVOA recommendation document in parallel.

I have talked with the head of the TCG, Marco Molinaro, and he says that path 1. would be easier, but we've never tried such a process. So we're not sure how it will be seen by the other WG leads sitting at the TCG.

We have a meeting on Feb. 26th. If you are ok, I'll prepare a few slides to discuss this. I'll also check with Brian Thomas and Arnaud Masson, so that they are aware of this.