metadata and objects

Am Dienstag, 07.10.03 um 17:35 Uhr schrieb Simon Pamies:

> Hi!
>> I knew some-one c/would do this for me :-))
> aha ;-)

what else can you do when you're lazy and stupid :-)

>> this sounds - all in all - like a proposal for a set of default object
>> classes in Ticle,
>
> not really into the ticle core but at the Zope / Plone Layer. Handling
> of special objects should be done by an App Layer not by the engine 
> itself !!?

it seems you're right - at least, this is a point.
Maybe, I see objects and metadata to closely bound to each other: I 
don't want to loose the metadata when I make - for instance - an 
XML-export of my object following a defined XML-Schema so that others 
(with complete different databases, like SQL-based dbs) at least could 
import all the important information. This would only be possible if 
the export is done from the Zope/Plone instance and the special 
representation of my objects, right? Otherwise, the metadata 
information would be lost.
Another problem of separating object data (in the Ticle-Repository) and 
the metadata in the Zope Application could come up, when you want to 
define your object-class and the belonging metadata in one XML-Schema. 
Wouldn't it then (in case of the separation) be necessarily to have to 
schemas?

Hm, I have a certain feeling that my problems come from a basic 
misunderstanding of the whole thing ... but at the moment it looks to 
me this way.

>> "Do you want to create objects compliant to
>> [ ] Dublin Core
>> [ ] OAI
>> [ ] IMS
>> [ ] something else ..."
>>
>> What do the others think?
>
> Think that's ok that way!
>
> Simon
>
> P.S: Please try to add some urls (spec, proposal, example) every time 
> you suggest a feature.
> It makes life easier to have urls chosen by people working with that 
> stuff, as to search it myself ;-) thx

Ok, - and, maybe we should start a collection of links to metadata 
standards / proposals ... ?
People (in my environment) are getting more an more interested in these 
things. Especially, it seems as if no software will have any chance to 
be accepted in the future if it does not offer some sort of 
"configurable metadata" (ok, Zope does in a way - and we should not 
hesitate to stay closer to Zope if things become to difficult 
otherwise) or a default set of metadata standards to chose from. 
Everybody seems to believe, that Semantic web and Onthologies could be 
a solution to the problem of billions of google results - but no-one 
things about the two main problems
- "putting the real world into semantic boxes" to me seems impossible 
even for clearly defined scientific areas (because you'll always learn 
very fast, they are _not_ clearly_ defined, or there is - at least - 
misunderstanding about the semantics ..)
- "who you gonna get to do the dirty work" (of filling in all those 
masses of data) "when all the slaves are free" - Jony Mitchell)

Ciao

Bernd



-------------------------------------------------------
This SF.net email is sponsored by: SF.net Giveback Program.
SourceForge.net hosts over 70,000 Open Source Projects.
See the people who have HELPED US provide better services:
Click here: http://sourceforge.net/supporters.php