VOEvent used for solar knowledge base

[Rob Seaman]

On Jul 18, 2007, at 10:05 AM, Frederic V. Hessman wrote:

> I wasn't expecting LMSAL to throw out their schema (although I  
> wouldn't mind) - the point was to show that the conversion could be  
> easily made, so LMSAL could easily produce a filter which converts  
> their internal document to a public one so that they could publish  
> the details in a VOEvent-compatible format outside the trivially  
> generated but generally unparseable <Group>'s and <Param>'s.

We're conflating <What> and <How> here.  Grouped params are precisely  
intended to provide a straightforward, general purpose mechanism in  
the absence of a vast data model covering all possible astronomical  
measurements.  Presumably purpose-specific schemata will emerge that  
can later be plugged into the <What> element.  Params will remain as  
an option with a low buy-in cost.  On the other hand, RTML addresses  
the still complex, but more manageable, chore of describing how to  
make those measurements.  In short - engineers are better behaved  
than scientists.

Independent of VOEvent usage, it certainly seems productive for the  
RTML community to wrestle with the requirements of novel  
instrumentation on an ongoing basis.  At any rate, we should  
encourage the VOEvent community to use RTML, but I don't think we can  
require this.

> Hmmmm......  this would give us explicit structure, but I guess  
> that this information is generally not needed for ground-based  
> telescopes.   With SOHO, the telescopes all have to look in the  
> same direction, so the organization via <Device> makes sense where as
>
> 	<Observatory name="ESO">
> 		<Site name="Cerro Paranal">
> 			<Telescope name="Antu"/>
> 			<Telescope name="Kueyen"/>
> 			<Telescope name="Melipal"/>
> 			<Telescope name="Yepun"/>
> 		</Site>
> 	</Observatory>
>
> is formally correct but who cares, at least for RTML, VOEvent or  
> HTN purposes?    Well, maybe but just maybe for VLTI purposes, but  
> then one could use
>
> 	<Device name="VLTI" type="interferometer">
> 		<Telescope name="Antu"/>
> 		<Telescope name="Kueyen"/>
> 		<Telescope name="Melipal"/>
> 		<Telescope name="Yepun"/>
> 	</Device>

The future of astronomy is as a systems of systems.  A configuration  
like VLTI is tightly bound and the telescopes were predesigned to  
work together as in the second example.  VOEvent and HTN are  
heterogeneous by design, however.  Very diverse assets on one or more  
mountaintops or space platforms might well be combined toward a joint  
observing project with a common experimental design.  These assets  
will not therefore constitute a single device, but simply function as  
separate components in a loosely bound observing system.  (Or the  
system might include compute and VO assets and spill over beyond  
simply "observing".)  I don't think either example above captures the  
essence of autonomous astronomy that we've been discussing for the  
last couple of years.

Rob