VOEvent and Heliophysics Knowledge Base: new schema

[Rob Seaman]

Hi Elizabeth,

This is great progress.  Thanks!

> A few months ago I emailed the VOEvent list about Lockheed Martin's  
> Heliophysics Knowledge Base, which uses a modified form of VOEvent  
> packets to record solar events. At the time, the VOEvent packets  
> products by the IDL SolarSoft routine vobs/ontology/idl/ 
> export_event.pro did not validate against the VOEvent 1.1 schema.  
> Following discussions at September's ADASS VOEvent BoF, I have put  
> together two new schemas and an example solar VOEvent packet:
>
> 1. original HPKB solar flare voevent packet
> http://www.lmsal.com/helio-informatics/hpkb/Flare_example.xml
>
> 2. solar flare voevent packet rewritten to validate against new  
> VOEvent-v1.1_lmsal and lmsal_v0.1 schemas
> http://www.mssl.ucl.ac.uk/~eca/Astrogrid/xml/ 
> Flare_example_ECA291007.xml

This looks much better to me.  I will also be interested in the  
comments from the WG.

> 3. LMSAL schema incorporating solar-specfic mandatory and optional  
> parameters from the HPKB website:
> http://www.mssl.ucl.ac.uk/~eca/Astrogrid/xml/lmsal_v0.1.xsd
>
> 4. VOEvent 1.1 schema with new element "ExternalEventMetadata" as  
> an optional child of "What". ExternalEventMetadata has child  
> "SolarEvent" from the lmsal_v0.1 schema.
> http://www.mssl.ucl.ac.uk/~eca/Astrogrid/xml/VOEvent-v1.1_lmsal.xsd
>
> I'd be interested to hear your comments, especially on the proposed  
> "ExternalEventMetadata" element under "What".

We either need a new subelement like this (the name may be a bit too  
generic) or we need a new top level element such as <Whatever> as  
discussed at the ADASS and IVOA.  I like your suggested usage, but  
I'd like to hear from both the XML and HTN wonks as to the  
suitability versus <Whatever>.

> One question - the IVOA Documents page lists VOEvent 1.11 as the  
> latest tech spec, but the IVOA xml page still lists VOEvent 1.1 as  
> the latest schema. Is there a v1.11 schema lurking somewhere that I  
> could use?

Hmmm - I don't believe any of us deemed a 1.11 schema to be  
necessary.  The official location for such would be http:// 
www.ivoa.net/xml/VOEvent.  I'd have to reacquaint myself with the  
detailed changes between v1.11 and v1.1 of the document - I suspect  
these were all textual things requested during the RFC process, so no  
change would have been needed to the schema.  Since the process has  
been revised since then I'm not sure how much effort it is worth to  
revisit the RFC issues.

In general, one might think that incrementing whole number versions  
for a standard would definitely require a new schema, that  
incrementing tenths might or might not - but that incrementing  
hundredths never would.  Which is to say that any change to a  
standard requiring a new version of the schema would have to be at  
the level of a tenth in the version or greater.  On the other hand,  
the text is normative and simply updating a schema to better reflect  
the standard may not require a new version, per se (although we  
clearly want to keep track of change management for the schemata).

A heads-up to the peanut gallery.  The basic idea is to work on v2.0  
of the standard.  V2.0 is intended to be backwards compatible, i.e.,  
a v1.1 packet should validate against a v2.0 schema (other than  
technical XML issues, I suppose).  In addition to permitting the  
enclosure of external schema (in <What> or <Whatever>), the other  
four items suggested so far for v2.0 are:

	XML signatures
	Time series (via SDM elements in <What>)
	Orbital elements (via STC elements in <WhereWhen>)
	SKOS vocabulary support in <Why>

The consensus was that this was definitely a major enough set of  
changes to require a whole number version increment.

Rob