Registry usecases

[Niall Gaffney]

Roy,

I think were going the right way. I have added some comments on 2 and 3 
to sort of look at what sorts of advanced registry queries we can handle 
to help the user choose the right services for their problem.

Roy Williams wrote:

>2> I have discovered a subclass of dwarf galaxies that fall into a
>particular color, color, absolute V mangeude space.  I need to get
>
....

>I'm not sure what are the dependancies you are talking about. Perhaps
>you could define more closely this "Sloan - 2Mass cross correlation
>color service"?
>
This is just someones results of combining Sloan and 2Mass which now 
resides as a catalog availble from the VO.  It has the coverage region 
of Sloan. What it is is not important, just that it has the information 
with UCDs that can be made into the UCDs that are what I need.

So not to go into implemetaton too much (I'll implement a couple of 
users but not the registry)...here is a couple of flows of how this 
might be done.  

1> Query registry for photometric catalogs for these objects that have 
the UCDs I need
2> Discover that there is derived photometry in a catalog that does the 
trick for region X
3> Query registry for data sources that cover NOT X that have data from 
which I can derive the UCDs that are of interest to me
4> Discover that for the region NOT X there are data providers to do the 
trick from two data sources

Its this handling of the regions and logic (we might need NOT X AND NOT 
Y too) in both the query and the results that is the power here.  

>>3> I want to make the highest S/N image of a region of the sky in the V  
>>  
>>
>>band from all images available to study a very distant galaxy cluster's structure.  
>>
...

>This wants metadata about the survey itself (sky coverage), plus
>metadata about an image survey (resolution, wavebands). Probably the
>"independent observation" part you would have to check carefully
>yourself, I wouldn't want to trust a computer with that.
>
I agree that checking it yourself is probably the way to go, but if the 
registry gives me the metadata and methods to backtrack the parents 
lineage, I could do the research quickly.

However my second use case shows a different reason for having the 
registry know about some form of lineage.  By knowing which dataparent 
was not the limiting factor in the all sky coverage of the catalog, you 
could make sure you use the 2-mass data to fill out the other half of 
the sky, combining it with some other souther sky optical survey of our 
choosing.

Looking at Rays use cases I think most of what is needed is listed in 
there.  I think the power of logic on regions and some simple lineage 
infomation will take us the rest of the way.

Niall