VOEvent transport protocol Working draft online

[John Swinbank]

Dear all,

Thanks, François, for posting this, and Rob for getting the ball rolling with some comments.

A little back-story may be in order for those coming on this for the first time. An IVOA note describing VOEvent Transport Protocol has been available for several years. It’s become a de-facto standard for transferring VOEvents over the internet and is currently in use by a number of projects. Unfortunately, the original note has a few rough edges and ambiguities. This is an attempt to resolve those issues and polish the document up to be usable as a standard. Given the need to maintain interoperability with existing systems, it is deliberately not a root-and-branch effort to design a modern protocol suitable for shipping VOEvents.

There are a number of existing implementations of part or all of VTP. I’m aware of:

* Comet, http://comet.transientskp.org/
* The Dakota VOEvent Tools, http://voevent.dc3.com/
* PyGCN, https://github.com/lpsinger/pygcn
* The GCN VOEvent Client Demo, http://gcn.gsfc.nasa.gov/gcn/voevent_client_demo.c

(GCN presumably also has a server implementation which acts as a counterpart to that client, but as far as I know it’s not publicly available.)

The VTP document itself is available on GitHub [https://github.com/jdswinbank/vtp]. While I welcome your feedback in whatever form you care to provide it, for minor issues in the document (typos, formatting, etc), GitHub pull requests are preferred.

Cheers,

John



> On 22 Sep 2015, at 17:20, Rob Seaman <seaman at noao.edu> wrote:
> 
> Howdy,
> 
> Good to see motion on this. “Long period of maturation” is about the most polite way this has ever been stated ;-)
> 
> My only comments on the draft are:
> 
> 1) It is a lot more mature than the current version v0.1 claims. Please increment to v1.0, v0.99 or whatever.
> 
> 2) Any path forward will likely include the transport of diverse message types over the same channel, whether these are portfolios, containers, or even non-IVOA messages. For instance, IAU Comm-20 adopted a new XML-based MPC astrometry format at the GA:
> 
> 	http://minorplanetcenter.net/iau/info/IAU2015_ADES.pdf
> 
> and it is easy to imagine interleaving VOEvent messages with MPC messages. Perhaps a paragraph explicitly supporting such usage can be drafted?
> 
> To provide context for the DAL group, VTP "formalizes a TCP-based protocol”, not THE transport protocol. Diverse communications channels have always been anticipated for VOEvent and other alert messages.
> 
> Rob
> —
> 
>> On Sep 18, 2015, at 7:00 AM, François Bonnarel <francois.bonnarel at astro.unistra.fr> wrote:
>> 
>> Dear interop members,
>> 
>>    After a long period of maturation in the Time Domain Interest Group, it's now time to move the discussion on "VOEvent Transport protocol" to the DAL mailing list. The Working draft can be found here http://www.ivoa.net/documents/VOEventTransport/
>> 
>> Best regards
>> François Bonnarel and Marco Molinaro for the DAL WG
>> John Swinbank and Mike Fitzpatrick for the Time Domain IG
>