Params cannot carry descriptions in VOEvent 1.1

[Matthew Graham]

Hi Bob (et al.), 

On Mar 26, 2010, at 7:54 PM, Bob Denny wrote:

> Matthew et al. --
> 
> A few thoughts:  
> We should ask ourselves if a <Description> and/or <Reference> inside a <Param> makes sense. What happens if you have 100 <Param>s with the same name but different <Description>? What would the <Description> elements be used for? And what's the use case for a <Reference> on something as elementary as a <Param>? I don't know the answers, but the questions should be asked.
Roy gives an example of using a Param and a Description in our recent IVOA note:

<Param name="Sun_Distance" value="132.21" unit="deg" ucd="pos.angDistance">
	<Description>Angular distance from the sun to the event position</Description>
</Param>

Similarly a Reference could be used to link to a full description of a particular parameter but I agree that we should decide whether we actually want this before "fixing" the schema.
> "Bob, is this correct in your schema?" and "Well, it should mean that any Reference should not validate at all since no type is specified but it shows how forgiving most validators are!" In my mini-schema, <Reference> is typed. It had to be in order to make a validator that works (as well as a binding). Please see the VOEvent Mini-Schema V1.1 (a graphical/hypertext view of the schema) for answers to questions about same. The validators that I have used would all barf on the definition in the original schema. They are not forgiving at all! 
So we definitely need to update the "definitive" schema that is registered with IVOA, Rob.
> "The default interpretation for a missing [reference] type is url:" A reference is not only a uri but also name and a type. Thus "reference" must be defined as a complex type. Again refer to the mini-schema (and the original).
Agreed.
> Note that a <Group> can contain one or more <Param>s plus  zero or more <Description> and/or <Reference> elements per both the original and the  mini.
Yes.
> This issue is a perfect example of why a schema is a more precise/rigorous way  to state the spec, combined with explanatory amplification only where needed.
<Alert>Big can of worms</Alert>

> Matthew, at some point in the past you mentioned that you felt that bindings are not the right way to handle XML. When you have a moment, I'd like to take your input on this. I'm interested in alternatives. A phone call would be far more efficient. I'm in the office most days during business hours (will be out all day Tuesday the 30th). You can reach me at +1 480 396 9700, or drop me a line at rdenny at dc3.com with a time and number to call.

Bob, I'll give you call some time next week.

	Cheers,

	Matthew
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.ivoa.net/pipermail/voevent/attachments/20100326/40e285f3/attachment-0001.html>