Metadata Distribution Subcommittee of TAC

["Cantor, Scott" <>]

On 8/16/13 8:56 AM, "Tom Scavo" 
<>
 wrote:

>>I don't understand what that means, unless what you're talking about as a
>> "dynamic query" is the idea of querying on some arbitrary criteria.
>
>Yes, precisely. I want to assume that everything of interest can be
>known in advance and can therefore be pre-computed, so that all we
>need to do is map the query to a file in the file system.

Ok. Let's call that "ad hoc" rather than dynamic. Particularly because the
SP plugin for MDX use is called "Dynamic". ;-)

> This is what we do now for JSON metadata, in fact. I'm not keen on
>building a high
>availability, high latency system that supports the full range of MDX,
>that is, queries for arbitrary subsets of metadata. (Related to this,
>I suspect eduGAIN is a disaster waiting to happen.)

Yes, I would note that much of that flexibility wasn't really the original
scope we had for the protocol.

>Yes, that *is* what I mean but I think we can do more. The MDX syntax
>is easily extended to entity attributes, so it shouldn't be too hard
>to let users create entity attributes and thereby pre-register their
>queries (so that the target aggregate can be pre-computed).

Right.

>Of course all of this depends on an online signing key (online in the
>sense that these processes can be automated), which is why your notion
>of a trusted HSM for the signing key is crucial.

We definitely need to capture why that is. As we discussed, certainly we
can precompute a lot signed documents, but with a manual process, we just
can't do it often.

-- Scott