tor 2008-12-04 klockan 18:40 +0000 skrev Pete Johnston:
> Hi Mikael,
> 
> > Could you consider importing the term definitions into the 
> > IEEE template? Just use the attached file. 
> 
> I'm just working through doing that now.... 
> 
> One quick question: in
> 
> http://dublincore.org/educationwiki/DCMIIEEELTSCTaskforce/LomDCAMAnalysi
> s
> 
> the mappping for 2.1 Version indicated a requirement for a class
> lom:Version (which would be the range of the lom:version property), and
> that's how that property is described currently in
> 
> http://dublincore.org/educationwiki/DCMIIEEELTSCTaskforce/LomTerms
> 
> In the new PDF version, the mapping (correctly, I think) uses a
> LangString. 
> 
> So I think that means the range of the lom:version property should be
> lom:LangString and we don't need the lom:Version class?
> 
> Is that right please?
> 
> Alternatively, I guess we _could_ still have a resource of type
> lom:Version as the value, with the same value strings, but that would
> seem slightly inconsistent with other cases like
> lom:installationRemarks, lom:otherPlatformRequirements.

It *is* inconsistent. Necessarily so. Compare Coverage, where the value
is, for example, a Jurisdiction, but described by value strings picked
from a LangString. Same with Rights.Description - the value is a
RightsStatement with value strings picked from a LangString.

I've tried to think about whether the value is naturally a language
object or something else. Version is somewhere in between, but I think
about it as something else than just a natural language thing (I'm
thinking it can be given codenames, release dates, etc etc, which an
"otherplatformrequirements" cannot be).

Maybe lom:Version is even a frbr:Expression? :-P

/Mikael


> 
> Thanks
> 
> Pete
> 
-- 
<[log in to unmask]>

Plus ça change, plus c'est la même chose