VOEvent v2.0, some thoughts

[Joshua Bloom]

>
>>    4) time series via SDM (<What>)
> I think this can be done, if we all agree to use the same time scale  
> -- otherwise it will be difficult to control which time standards  
> are allowed, and parser will become difficult. Suggest Barycentric  
> Modified Julian Day for this.

Agreed that the simplicity of a single time model is appealing, but  
there will be many applications for which the marshaling into such a  
framework either is conceptually awkward (e.g. pulsar timing) or is  
limiting. Indeed any short episodic event that needs to be tracked  
over a significant time in the Earth's motion around the Sun would be  
more naturally described in heliocentric terms.

That said, selling the timeseries addition in 2.0 as a beta  
representation is reasonable way to gain acceptance. (Afterall, MJD  
should satisfy 90%? of the current VOEvent users).  So I'm for the  
proposal.

Since the ultimate question here is how clever a parser can be, we  
should hear from those more familiar with XML schema validation as to  
how complex the timeseries model can be.

>>
-Josh