Re: ISSUE-160: Allowing collections in semantic relationships

Hi Tom,

> I wanted to point out that "Dublin Core" is more than a set
> of fifteen elements used with string values (a usage which
> is now referred to as "Simple Dublin Core").

It is not about what DC is at present - it is what it was in 1995 when it was 
first proposed and subsequent years.
Look how long it took to make DCMES unambiguous and functional?
People who did not have enough background in descriptive metadata interpreted 
the DC semantic in many different ways simply because the standard itself did 
not offer enough constraints.

Hence the comparison betwen SKOS and DC approaches should be considered in this 
light and my point was exactly that SKOS need not to spend 10 years of learning 
on mistakes in order to become more useful to wider community (as opposed to be 
useful on to those who have resources to extend it).
There is no doubts that formats can be extended and that is quite easily to do 
so - but application formats make data integration much harder if fundamental 
things (such as the one of difference between node label and concept) are 
misunderstood.

One of the good approaches in standards creation is to, create a scalable model 
from the outset and whenever possible offer, full set as well as sub-set of core 
elements.
If this is not possible then at least to make sure that the main format is 
robust enough and that it can be downscaled easily.
I think to keep SKOS intentionally simple in order to create an illusion that 
pinning down semantics is simple may not be the best approach.

aida

Received on Tuesday, 16 December 2008 13:00:37 UTC