RofR

[Tony Linde]

> >Let's distinguish between full and publishing registries. 
> >
> I was never really comfortable with this distinction. My view 
> is that each registry *may* have any of three interfaces: 
> publish, query, and harvest. A registry may harvest only a 
> few other registries, or it may harvest all. If people can 
> publish or query there, I don't see what that has to do with 
> their harvesting policies. Please can somebody explain the 
> advantage of this  full/publish scheme?

Registries do not publish anything - publishing registries started out being
called 'local' registries to indicate that they were invisible on the global
scale. Registries only support harvesting and query interfaces. 

If a registry supports a query interface, it is a full registry - it must
contain all resource records (therefore must harvest from other registries
in order to do so).

A registry which does not support the query interface is a publishing (or
local) registry. It is simply somewhere someone (like a data centre) can
stick resource metadata which is harvested by one other full registry. That
full registry is the one which manages the publishing registry's authIDs.

In effect, the same software is likely used for both types of registry but
the different names indicate the different ways they function.

Cheers,
Tony. 

> -----Original Message-----
> From: Roy Williams [mailto:roy at cacr.caltech.edu] 
> Sent: 11 April 2005 14:17
> To: Tony Linde
> Cc: registry at ivoa.net
> Subject: Re: RofR
> 
> 
> 
> Tony Linde wrote:
> 
> >I'm unclear about what the registry of registries (RofR) is going to 
> >do. I understand that *every* registry is listed there, 
> whether full or 
> >publishing.
> >  
> >
> I am also unclear. I think it is good to know where the 
> registries are running, but I do not see why the RofR needs 
> to be a registry itself. 
> The problem then is that the owner of the RofR is perceived 
> as being the center of the IVO, which is politically 
> unattractive, unless you are part of the project that runs 
> the central registry, in which case it is very attractive.
> 
> Perhaps the RofR could start out as a web page or a Twiki 
> page, just so we know where everyone is. Forget the formal 
> structure for now? A URL for humans and an endpoint to get 
> the service?
> 
> >Let's distinguish between full and publishing registries. 
> >
> I was never really comfortable with this distinction. My view 
> is that each registry *may* have any of three interfaces: 
> publish, query, and harvest. A registry may harvest only a 
> few other registries, or it may harvest all. If people can 
> publish or query there, I don't see what that has to do with 
> their harvesting policies. Please can somebody explain the 
> advantage of this  full/publish scheme?
> 
> Roy
>