Vocabularies and VOEvent

[Rob Seaman]

> Shall we identify a few pre-existing time domain vocabularies and  
> start rendering them down to a common voe list?  ATEL, GCN, CBAT, ...
>
> I suggest we be prepared to support this common list as a default,  
> but also allow prior transient publishers to continue using their  
> current vocabularies when their telegrams/messages are instantiated  
> as VOEvent packets.

Let me expand on this (and note I've narrowed the distribution to  
voevent).

Time domain astronomy - in particular, transient response astronomy,  
but also synoptic monitoring of time varying sources - will benefit  
from a comprehensive alert format and transport network, namely  
VOEvent.  We therefore wish to invite prior transient alert publishers  
and their subscribers to benefit in turn from our comprehensive, VO- 
based, solution.

To put it more bluntly - we can't succeed if we don't offer the  
benefits Roy is seeking to our potential collaborators at an  
attractive cost point.  Because why?  Because a potential collaborator  
is a possible competitor.  If they can offer a better format and  
better transport with lower overhead of various sorts, astronomical  
customers will move to that alternative rather than to VOEvent.

Worse yet would be a situation that splits the market.  A stable, but  
suboptimal, solution would assign one type of transient to one system  
and another to a second (or third, etc, and we're soon back to  
Babel).  But one person's transient is another's distraction, and a  
truly useful system will permit both transient signal and transient  
noise to be handled by the same components.  After all, single surveys  
often discover multiple types of phenomena.

The great strength VOEvent has is that, unlike earlier transient alert  
systems, we are a peer-to-peer network of heterogeneous publishers.   
We don't want to force others to fit a mold - rather, we want a  
suitably accommodating paradigm that will permit publishers to use  
common formats and technology as best satisfies their own requirements  
and schedules.

On the other hand, subscribers not only want the one stop shopping of  
such a unified system - they (meaning the client applications that  
will later flow like Soviet harvests) want enough interoperation from  
these disparate publishers that it will matter little whether a newly  
discovered Royesque Banana (BNe in the plural) was published/announced  
by PQ, CBAT, ATEL, NOAO, etc and so forth.  Similarly with each follow- 
up observation in turn, and similarly for issuing SEAP queries or  
assembling parameters for passing to VO web services or community  
resources such as the IAU naming authorities at CBAT and MPC.

A facility similar to SKOS that permits a CBAT vocabulary to be mapped  
to an ATEL vocabulary and both to be mapped to VOE (and later to the  
IAU and various journals) is part of the glue that will make VOEvent  
(or any possible candidate for a comprehensive transient alert system)  
work.

One has to also think that such a thing will be key to any VO that  
ever hopes to reach outward to an often skeptical astronomical  
community.

Rob