May 2016 InterOp

[John Swinbank]

Hi Rob,

> On 28 Apr 2016, at 06:01, Rob Seaman <seaman at lpl.arizona.edu> wrote:
> 
> Le 06/04/2016 00:31, John Swinbank a écrit :
>>>> 
>>>> - Just today I (finally…) submitted the VOEvent standard to François Bonnarel with the intention that it moves to PR state: watch this space for details.
> 
> VOEvent (v2 even) is already a standard. I presume you are talking about VTP here.

Yes, of course. Sorry if my typo caused confusion!

>>>> - Jean-Paul, I would love to hear more about the latest news from Svom and the success you’re having with VOEvent transport over XMPP. Will you be at the meeting? If not, can you suggest somebody who will be there who can give an update?
> 
> Just pointing out that an XMPP proof of concept was demonstrated many years ago now. By all means discuss how this is being used by Svom, but the basic functionality isn’t in question.

I don’t think anybody’s questioning what’s possible; I’m simply soliciting descriptions of what’s been successfully deployed in practice and what lessons have been learned.

[…]

>>>> - I spent much of this afternoon sitting in an LSST discussion about their plans for announcing transients, touching on the format and content of alert packets and the transport mechanism. They’ve already announced ambitious plans for what they want to include in events (see http://ls.st/lse-163 p29), and they aren’t backing down. What can we do to be ready?
> 
> It would help if such discussions were announced in advance to VOEvent / TDIG. If held in Tucson (or there’s a video link), I would be happy to attend, but you should also reach out to the other ANTARES folks, and if solar system use cases are included, also to the new IAU/MPC ADES format folks (http://minorplanetcenter.net/iau/info/IAU2015_ADES.pdf).

Sorry if I wasn’t clear — this was an LSST internal meeting, to which the general public wasn’t invited (not even TDIG members!). I mention it merely for your background information. Several LSST folks will be at the InterOp (and may be at Hotwired), so there will be plenty of opportunities for future discussion.

> On 20 Apr 2016, at 09:24, Thomas Boch <thomas.boch at astro.unistra.fr> wrote:
> 
>>> VizieR has hundreds of catalogues with attached time series (full list available at http://vizier.u-strasbg.fr/vizier/welcome/vizierbrowse.gml?media ). It would of course make sense to publish them in a homogeneous format as this would allow generic tools to offer at least some preview and visualization features. I was however reluctant to invest time in this as I was unsure of the potential adoption of this standard. Does anyone serve (or plan to serve) time series in SimpleTimeSeries format?
> 
> SimpleTimeSeries were found to be acceptable by several VOEvent projects. As Thomas implies, it is a chicken and egg issue. We reached closure on SimpleTimeSeries long ago and IVOA carried it no further. It has been raised a second time two or three years ago and IVOA carried it no further. It was raised last fall and somebody suggested it was inadequate for some narrow purpose – but also that the alternative was equally inadequate – and IVOA is likely to again go no further. If community uptake is desired, standards have to actually reach the standards stage.
> 
> The entire notion was to adopt SimpleTimeSeries as a “simple” option and to *also* move forward on the “non-simple” solution. That SimpleTimeSeries doesn’t attempt to address every possible edge case is a feature, not a bug. If you standardize it, they will come.

Neither IVOA in general, nor TDIG in particular, has FTEs we can deploy to push particular technologies through a standardization process: this work is carried out on a volunteer basis, as and when the technology is required to meet a particular need.

As you describe, despite several attempted starts, we haven’t seen wide adoption of SimpleTimeSeries. Neither have we seen volunteers stepping forward to lead an effort to develop a “non-simple” solution. Perhaps that means the need hasn’t been sufficiently pressing to date? Perhaps the available technologies have been insufficiently capable?

In any case, I actively encourage those members of the community who are facing these challenges to consider the available technologies, to build new ones when necessary, and, where appropriate, to step up and help drive the process of standardizing them and making them available for others.

Cheers,

John