[sc34wg3] A plea for CTM and a request for more input on requirements and evaluation

[Kal Ahmed]

Hi Gabriel,

See my comments below.

Gabriel Hopmans wrote:

>Hello Kal,
>Thanks for your comments and additional requirements. I will add them
>to the presentation we are preparing.
>
>On 7/26/05, Kal Ahmed <kal@techquila.com> wrote:
>  
>
<snip/>

>>(v) -- XTM -> TMDM -> CTM and back is not necessarily a lossless
>>transformation. For example XTM -> TMDM loses information about the
>>grouping of topic references under association roles and the nesting of
>>variant names.
>>    
>>
>
>Ok, and I think there will be more examples. 
>But dont know if reformulating the requirement to "Roundtripping
>between XTM and CTM should be supported as best as possible" is a good
>idea?
>
>  
>
Actually I think you can just drop this round-tripping requirement as 
you already have requirements to deserialize into TMDM and to support 
all constructs of TMDM. So XTM <-> TMDM <-> CTM is a logical conclusion 
of that.

>>Two more requirements that I think you should consider:
>>
>>1) CTM must support the creation of modular topic maps with minimal
>>"import" mechanism. Something like LTM's INCLUDE directive.
>>
>>    
>>
>
>yes, I was yesterday asking LarsH and Robert if CTM should support the
>use of prefixes.
>Robert answered with integrating external knowledge should be
>possible. I think that captures what we need (including directives,
>prefixes) ?
>
>  
>
Yes I think so - though there is an important difference between a MERGE 
and an INCLUDE and I think it would be useful to support both directives 
in CTM.

Cheers,

Kal