Comments on mapping the IRA-CINECA schema to the SimDB model.
-------------------------------------------------------------

The model (I'll refere to it as IRA below) can be almost completely embedded in the SimDB model.
Here a proposed identification/mapping (left the Element from the IRA XSD, right the proposed SimDB UTYPE).
I have restricted myself

Experiment -> SimDB:/simdb/experiment/Experiment

Tool(s)->SimDB:/simdb/protocol/Protocol
Unclear how to interpret multiple tools for 1 experiment, see below.

Data->SimDB:/simdb/experiment/Result

DataObject->SimDB:/simdb/experiment/Product

Author->combination of SimDB:/simdb/Contact its referenced SimDB:/simdb/Party

Authors->SimDB:/simdb/Resource.contact (i.e. the collection, not a class)

Science->SimDB:/simdb/experiment/Experiment.parameter (i.e. the collection, not a class)

Parameter (as child of Science)-> combination of SimDB:/simdb/experiment/ParameterSetting with 
its SimDB:/simdb/protocol/InputParameter reference.
Parameter (as child of Data)-> combination of SimDB:/simdb/experiment/StatisticalSummary with 
the SimDB:/simdb/protocol/Property referenced by its axis reference.

Note that in the IRA schema there seems to be a similar struggle how to deal with parameter values of different types.
IRA can solve it using an xsd:group with a choice element. This was not an option for us as we also need to suport
relational databases. In UML we could have used a base-type Value or so, with subtypes String, Integer etc. 
We have a cardinality on the InputParameter definition that allows one to model the StringList etc.
But again, having multiple values for a parameter is harder to map to a relational database.
We would have to do some kind of serialisation like arrays in VOTable.
It would be nice to get some examples of parameters that have a list value.
 

Some questions/comments.

1. DataCollection has 1 Experiment. 
Is it possible for there to be multiple DataCollection-s with the same Experiment?
If not DataCollection and Experiment together can correspond to SimDB:/simdb/experiment/Experiment.
DataCollection, with its collections of Data and Tools could maybe be a SimDB:/simdb/Project, apart from the restriction
on having only 1 Experiment 
Are the Data of a DataCollection those produced by the experiment?


2. How should we interpret the collection of Tools.
Here an Experiment can be joined to multiple Tools through the common DataCollectionTools. 
In SimDB and Experiment corresponds to 1 Protocol, the actual executable.
A Project can have multiple protocols and corresponding experiments .
We used to have a CompositeProtocol, but no firm use cases were ever proposed that could not be (better) interpreted 
using a Project. Do you have one here?
 
3.
Some explicit SimDB features missing from IRA, though they may end up in the Info string.
- a full description of the protocol. In IRA model Tools can not be reused, has to be redefined. 
- indication of the properties that are being calculated in the experiment and are available in the Product.
- target: the goal of the experiment
- what physics is used in the experiment