RofR

[Tony Linde]

> I guess the picture i have a hard time with is really the 
> VObs and the proposed "hands-off" nature of the networked 
> registries...  I see things as more dynamical less 
> static and the need for flexibility and change.

But in order to allow 'flexibility and change', you need systems that can
adapt and support that level of dynamism - or you have systems that cannot
cope with it and simply leave everything to the user to decide.

I'm aiming at systems that can cope with significant dynamism without
constantly going back to the user looking for guidance. 

> CDS has done an excellent job at maintaining a large set of 
> managed metadata resources and perhaps they would have more 
> insight to this than i,  but the complexity of metadata that 
> is proposed i would think will require constant updates and 
> the simpler the better.

I don't think constant updates will be necessary. Once a resource is
catalogued, even with extremely complex metadata, updates will only be
required if the dataset itself changes. 

The goal has to be to create software which responds to changes in datasets
and automatically updates the resource metadata.

Cheers,
Tony. 

> -----Original Message-----
> From: Gretchen Greene [mailto:greene at stsci.edu] 
> Sent: 12 April 2005 18:45
> To: Tony Linde; registry at ivoa.net
> Subject: RE: RofR
> 
> it is a direct url link to the OAI resource publication.  I 
> have to admit i did advocate adding this to the schema to 
> have a reference instead of building another required service.
> 
> This predated the authority IDs mechanics and the simplicity 
> works to overcome the QA and maintenance barriers.  Note any 
> XML reader can verify the resource metadata source.  This 
> should be a plus.  
> 
> Okay,  a little rambling here.
> 
> I guess the picture i have a hard time with is really the 
> VObs and the proposed "hands-off" nature of the networked 
> registries.  This would parallel the world view of 
> ATM/Banking.  How to get scientific data to fit into that 
> regime is hard to see.  I see things as more dynamical less 
> static and the need for flexibility and change.
> 
> CDS has done an excellent job at maintaining a large set of 
> managed metadata resources and perhaps they would have more 
> insight to this than i,  but the complexity of metadata that 
> is proposed i would think will require constant updates and 
> the simpler the better.
> 
> -Gretchen