VOEvent II topics

[Rob Seaman]

Hi,

Looking forward to seeing folks again.  Like most projects we have a  
selection of short term and long term topics to debate.  Here's my list:

1) Should we establish a voevent namespace for "VOConcept" derived  
UCDs?  If not, what's the alternative?
2) Should VOEvent continue to rely on STC - and only STC - for  
coordinates?  What's up with XLink?
3) I plan to modify the schema to allow XML signing.  Suggest that we  
use same.
4) XML IDs?
5) We have a lot of support for referencing external content.  Should  
we provide the same options for internal content, i.e., data rich  
packets?  An application that actually intends to benefit from all  
that content is going to have to fetch it anyway.
6) We've demonstrated the value of a rich network of web services  
responding to simple packets.  How about the value of relatively  
simple applications responding to rich packets?
7) <Group> and <Param> provide a way to build moderately complex data  
structures in an efficient, ad hoc fashion.  They are not a  
substitute for a detailed, application dependent, schema.  <What> and  
<How> (in particular) could benefit from some mechanism for  
permitting such sub-schemata.  Am convinced we'll want such support  
eventually.  What are the implications for the specification and for  
the authors, publishers and subscribers?
8) Am less convinced that functional ontologies will become a reality  
any time soon.  What pragmatic compromises should we continue to make  
within <Why>?  ...and so back to #1

The true value of VOEvent (and maybe of the VO in general) will be  
realized when the specification, network and services are mature  
enough to represent and transport packets with the semantic richness  
of an arbitrarily complex IAU Circular but the near zero latency of GCN.

Rob
seaman at noao.edu