[OAI-implementers] confusion about the repository structure - Again

[Fadi Qutaishat]

Hi all,

I am really very new to the OAI/PMH and I am doing a
research where I have to come up with general
solutions which are out the scope of my research!!!
Therefore, I beg your pardon if I ask a very basic or
silly questions.

I've done some initial readings but the thing which is
I am (at this stage) confused about is how the
structure of the data providers' metadata repositories
should be. 

should the service providers produce their metadata
using the normal Dublin core (and then it would be
automatically changed by OAI/PMH into the structure
below when the metadata is harvested) or they still
should use Dublin core but they have to follow this
structure when creating the metadata files:

**************************************
header (mandatory)
 ­  identifier (mandatory: 1 only)
 ­  datestamp (mandatory: 1 only)
 ­  setSpec elements (optional: 0, 1 or more)
 ­  status attribute for deleted item

metadata (mandatory): where DC should  be used or
other metadata standareds 
 ­  
about (optional)
 ­    rights statements
 ­    provenance statements
...
...
...
***************************************
Again sorry for asking a very basic question.

Any help is really appreciated.

Fadi 



__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com