cocoon-dev mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Stefano Mazzocchi <stef...@apache.org>
Subject Explicit vs. Implicit dependencies [Re: Ant/Maven/Centipede discussion]
Date Thu, 01 Jan 1970 12:52:05 GMT

On Tuesday, Sep 30, 2003, at 16:36 Europe/Rome, Stephen McConnell wrote:

> What I have not resolved in my head yet is how the dependency url for 
> cocoon blocks translate to computational descriptions (or to put this 
> into Merlin terms, how can I translate a cocoon block uri into 
> computation service request that could be used in Merlin to return a 
> serialized Merlin block descriptor)?

You need the full power of an ontology to be able to do this.... 
something that we decided we didn't want to do so that human knowledge 
is distilled into explicit dependency, implicit version ranging and 
explicit discovery services.

Implicit dependency was proposed and it might be considered in a 
potential future of the system, when (if?) explicit dependencies become 
too hard to maintain or keep in synch and we need machines to help us 
finding the associations.

But for now, if you consider behavioral URIs as "functional 
abstractions", they could well be used as a taxonomy of behaviors, 
then, in the future, the explicit dependency between a functional 
requirement and its implementation might be inferred from ontological 
analysis capabilities.

But the complexity of such a system is *huge*, it has the full 
complexity the semantic web and writing an ontology of cocoon 
functionality would be an incredibly complex and expensive job (in 
terms of development *and* community maintenance).

This is why I proposed an lazy explicit dependency approach where the 
URI is just an identifier and the ontological evaluation is done via 
simple versioning ranging rules and by the block librarian which does 
discovery but it's driven by human-edited information.

--
Stefano.


Mime
View raw message