Expressing 2- and 3-D coordinates

[Ed Shaya]

When coming up with standard interfaces one needs to have a set of goals 
and requirements which flow from the application or system that you wish 
to create.  So, we have been working several years on the issues related 
to archival ingest, registry, retrieval, and query.  This does not mean 
that we have solved the problems for all project in the field of 
astronomy.  The document structure for publishing or archiving data for 
posterity is probably not suitable for a project with the goal of having 
Institution A  alert Institution B that it saw something.  If we try to 
stretch our standards to accomodate requirements that are outside of the 
initial goals and requirements we compromise those goals.  We could 
provide separate document standards (or models) for this special case, 
but then the burden on the VO to support all special projects in the 
field of astronomy is just too too great.

It is clear from the trouble we are having making STC aceptable to all 
parties that we are dealing with cross-purposes and changing 
requirements.  Specifically, the goals and requirements that I have 
heard from Roy's descriptions of VOEvent sit well outside the VO goals 
of providing an electronic observatory where science researchers can 
easily gain access to full astronomical data.   For example a 
requirement that one should be able to send just an RA,DEC pair to 
describe a skycoodinate, is totally at odds with the goals of archival 
science.  And, allowing it in our standards would be highly detrimental 
to future generations of astronomers.

However, none of this need be considered a problem at all.  It is just 
an issue of understanding when and why to relax standards.  The how is 
simple.  The VO should provide a fairly exacting set of scientific 
standards.  Projects  are free to  create copies  of the schema and  
knock out  parts and change required  elements to  optional  ones.  The  
documents point to  the  relevant schemas, so they are valid within the 
more liberal schema.  In most cases, applications will be oblivious to 
the missing parts or simply won't do what can't be done without the 
missing parts.  The question is who is the creator and maintainer of the 
more liberal schema?   I don't think the VO has the manpower to take 
care of these nor should they.  Perhaps, the VO has some responsibility 
to design our standards in such a way that others can knock out parts 
and still have something usefull.

Ed
 

Alasdair Allan wrote:

>
> Ed Shaya wrote:
>
>> In other words, projects like VOEvent should not guide the VO.  It 
>> should be informed by the VO and VO should be informed of it.
>
>
> So you're saying VOEvent is not part of the VO? Err, that doesn't make 
> a lot of sense to me...
>
> If you're saying that it is part of the VO, but that the VO standards 
> process should not be guided by it, then again this doesn't make a lot 
> of sense to me. It's a core VO project, it needs certain things from 
> other parts of the VO, why shouldn't the evolution of those standards 
> by influenced by the needs of VOEvent?
>
> Al.
>