[tmql-wg] Proposed changes to existing requirements

Kal Ahmed kal@techquila.com
Thu, 10 Apr 2003 21:24:47 +0100

Lars Marius Garshol wrote:

>* Lars Marius Garshol
>|  "TMQL should be able to query topic maps stored in a distributed
>|  fashion on different TMQL processors."
>* Kal Ahmed
>| Are you saying that TMQL should define a mechanism for aggregating
>| results sets ? Or that TMQL will define a mechanism for accessing
>| distributed TMQL processors (e.g. by specifying a web service
>| interface that a TMQL implementation must support) ?
>This one came from Holger, so please don't take what I say as gospel,
>but as I understood it it was about being able to query topic maps
>stored in a distributed fashion as if they were a single topic map.
>To do that you'd need to be able to aggregate result sets and access
>distributed processors, but TMQL wouldn't necessarily specify how to
>do that. (Holger may disagree on the "necessarily", I don't know.)
>| If the former, then +1, but make it clearer in the requirement.
>| "TMQL should define a mechanism for aggregating results sets".  If
>| the latter then -1. It is important, no doubt, but should be tackled
>| by a separate standard/recommendataion.
>I would tend to agree with your -1. Your +1 I'm less sure of. Could
>you expand on how you'd see that fitting into the standard and how it
>would be useful?
Let me propose: "TMQL results sets will be topic maps. Aggregation of 
TMQL results sets will be performed according to the topic map merging 
rules of the SAM". How's that for fitting into the standard ? ;-) As for 
useful? <shameless-plug>Wait till you see peer-to-peer topic maps in 
action at XML Europe 2003</shameless-plug>

>| +/- 0 until I read what the modifications made by RFC 2732
>| are...unless you can elaborate on this for me.
>RFC 2732 adds support for IPv6 IP-addresses in URIs. So 2396+2732
>means you need a more up-to-date URI parser than with only 2396.
>That's really all.
Oh, in that case I have no problem with that at all.