TDIG Chair

[John Swinbank]

Dear TDIG,

(As you might have heard rumoured previously…) following last month's InterOp in Heidelberg, the Exec has asked me to take on the role of Chair of the Time Domain Interest Group, a position which I'm now happy to have accepted. I'm delighted to report that Mike Fitzpatrick has agreed to take on the job of Vice Chair.

I'd like to thank Matthew for all the work he's put in as the previous chair of the TDIG, and wish him every success in his (continuing) role as Vice Chair of the TCG.

Perhaps this is a good moment to take stock of what the immediate priorities for the group are. Off the top of my head, I'm aware of the following work in progress:

- Endorsing SimpleTimeSeries as an IVOA-approved format.

I understand that this will involve the production of an appropriate IVOA note. As far as I'm aware, Matthew is taking the lead on this.

- VOEventRegExt.

A working draft of VOEventRegExt has been submitted to the Registry WG for comments. Matthew submitted this document to the TDIG list for comments in October last year, and is available from <http://www.ivoa.net/pipermail/voevent/attachments/20121001/f8bef60c/attachment-0001.pdf>. I believe that still represents the current latest draft, and we are effectively in a holding pattern waiting for the Registry WG before we make further progress here.

- VOEvent Transport Protocol.

I have redrafted the VOEvent transport note by Bob & Alasdair, and it is our hope to push this through the IVOA standardization process. The latest version is at <http://tinyurl.com/20130513vtp>, and your comments and corrections are welcome. I hope to work on polishing this up further over the next few months, and then start trying to figure out what bureaucratic hurdles it needs to jump through to become official!

In addition to those immediate items on the todo list, we can consider longer-term discussions. In no particular order:

- VOEvent security.

How can we be sure that a VOEvent is from who it claims to be from? How can we provide events only to authenticated subscribers? This discussion has rumbled on for a long time, and I personally think security would be a major bar to widespread adoption – I would be reluctant to trigger my telescope unless I was really sure I was dealing with a genuine event. Bob and I rehashed a lot of the issues last year, but unfortunately rather ground to a halt, due chiefly to a lack of time on my part (for which I apologize). This is definitely something I'd like to return to.

- Future evolution of the VOEvent standard.

We discussed this a little last month, and it's something that I'm sure we'll return to again in the future (perhaps most immediately in person at http://www.hotwireduniverse.org/). I don't see a pressing need to work on a major redraft – to be honest, I'd rather focus on adoption of what we have now – but we will need to be ever alert for and responsive to the requirements of our community, and particularly the special requirements of major new facilities like LSST (while being careful not to disenfranchise existing users, of course).

- Time series data modelling.

Is SimpleTimeSeries enough? Will fully incorporating time domain information into the VO require a more elaborate model? If so, what are the requirements? …that should keep us busy for a while.

So that's what's on my mind at the moment. What have I missed? Please pitch in and tell me what else we should be worrying about.

I'm also planning to reach out to the representatives of the time domain community who joined us at the InterOp in Heidelberg and encourage them to engage with us and let us know how we can best help them reach their goals: whether that will result on any new members of this list I don't know, but I'll certainly pass along any feedback.

Cheers,

John