Newbie question : how to specify a source instrument in a VOEvent ?

[Jean-Paul Le Fèvre]

Don't rejoyce too fast because I've thousand of questions to ask about VOEvent stuff ;)

In fact I think that the sets of parameters will be very different : one instrument is a gamma ray camera, another one a Xray camera, another a one is working in the visible band, etc. So the second option, i.e.  2 or more streams, would be preferable.

However I'm confused about what sounds inconsistent to me : 
the recommendation 1.11 states that the resource key of the IVORN must be the publisher identifier (key point written in bold face in the document)
while the voeventstream working draft proposes that the resource key is the stream name.
I would appreciate comments about that.

Moreover I've the feeling that inserting too many contents inside the IVORN string does not respect the spirit of the VOEvent XML philosophy. Isn't it better to parse a xml document using a xml schema than to chop a string to figure out the type of the packet ?  And I don't understand why the <How> element should be despised.

PS :
We are not yet ready to play with samples to test different approaches, I hope that by the first semester of 2010 we will have something available.


--
-----------------------------------------------------------------------------
Jean-Paul Le Fèvre 
Svom FSC project manager - CEA Saclay Irfu
Mél :  jean-paul.lefevre at cea.fr
Tél : +33 1 69 08 34 58


>-----Message d'origine-----
>De : Roy Williams [mailto:roy at cacr.caltech.edu]
>Envoyé : jeudi 5 novembre 2009 16:03
>À : Le Fèvre Jean-Paul
>Cc : voevent at ivoa.net
>Objet : Re: Newbie question : how to specify a source instrument in a VOEvent ?
>
>Jean-Paul
>
>I am so happy that Svom is becoming part of the VOEvent world. As you
>know, we are running an event publisher at Caltech, and would be very
>happy if you could publish there, it is called skyalert.org, and the
>specific instructions for authoring are here:
>http://www.skyalert.info/static/author.html
>
>As for the different instruments, you can think this way. Suppose
>instrument A returns events with data parameters A1, A2, and A3, and
>instrument B returns parameters B1, B2, B3. You can choose
>-- to have a single event stream which allows A1, A2, A3, B1, B2, B3,
>and every event  provides values for either the A or B parameters. This
>is the SWIFT approach. Or
>-- to have two separate event streams SvomA and SvomB, where the first
>kind of event has A parameters, and the other kind of event has B
>parameters.
>
>In any case, I would like it if you could try to make one or five
>"sample events" as defined on the above link. I hope we can work
>together on this!
>
>Roy Williams
>
>
>Le Fèvre Jean-Paul wrote:
>> Hi,
>>
>> We have just started a discussion about how to format VOEvents for the Svom
>mission. (A future French-Chinese space mission dedicated to GRB).
>> There will be at least 5 instruments involved in the mission capable of
>generating data. My question is the following :
>>
>> What is  the best way to specifiy in a VOEvent packet which  instrument is the
>source of data  ?
>>
>> - a token in the localID of the IVORN (Fermi and Swift seem to have chosen this
>solution),
>> - using a <Group> element with a different name attribute for the different
>instruments,
>> - the <How> element referencing a rtml document describing the different
>instruments,
>> -  other.
>>
>> It is not clear for me which solution is the best, so I need advice !
>> Thank you in advance for your help,
>>
>> --
>> -----------------------------------------------------------------------------
>> Jean-Paul Le Fèvre
>> Svom FSC project manager - CEA Saclay Irfu
>> Mél :  jean-paul.lefevre at cea.fr
>> Tél : +33 1 69 08 34 58
>>
>>
>>
>
>--
>
>California Institute of Technology
>626 395 3670
>