GRBs retractions - Param or value?

[Roy Williams]

On Aug 7, 2006, at 2:44 PM, Paul Price wrote:
> What you differ over is how the tags that contain that information  
> is documented.  You seem to favour documenting the tags in a formal  
> XML schema, while the current standard leans toward individual  
> providers providing their own documentation, with the fields  
> subject to change without much in the way of notice (much in the  
> way the GCN currently operates).

Thank you Paul for this good summary.

Most interesting to astronomers, I think, is documentation of what a  
parameter actually means, and we have that right from the start with  
the Description tag. While some technical people are most interested  
in an Xschema description, I think this is of less interest to  
astronomers, and would in reality be a big roadblock to entry.

I feel that formality and prototypes should develop together, rather  
than crystallizing the schema before prototypes. The last thing  
anyone needs is an overly-rigid system of fixed words and concepts,  
so that authors of events cannot express themselves. We do not want a  
Description tag that reads: "This event has a value of alpha at  
roughly 4.66 ". No computer can extract anything from that string!

What we intend with the What section of VOEvent is a "path to  
formality" in several stages:
(1) Author makes Param elements to describe measurements, hopefully  
with a Description tag to explain the meaning.
(2) Author makes a standard Group of Params with standard meaning.  
For example every Group of type "BurstStatistics" might contain  
"PeakCount", "TimeDuration", "AverageRate". The Group entity is also  
documented as a whole.
(3) Once such a Group of Params is shown to be useful through  
prototyping, we can convert it to XML elements and a piece of schema,  
allowing automatic validation. A small set of VOEvent Authors and  
Subscribers works with this extended schema for a while.
(4) Gradually the new schema is standardized within larger groups  
such as Integral or Astrogrid.
(5) When we really know how to effectiviely describe a specific event  
type, we change the VOEvent schema.

Roy Williams

California Institute of Technology
626 395 3670