Re: [office-metadata] data-typing

[Michael Brauer - Sun Germany - ham02 - Hamburg <Michael.Brauer@Sun.COM>]

Hi Patrick,

Patrick Durusau wrote:
> Bruce,
> 
> Apologies for the long silence but doing email on 30 minutes a day is 
> hard! Low end hotel so I have to buy email access in 30 minute increments.
> 
> Bruce D'Arcus wrote:
> 
>>
>> On Mar 2, 2007, at 11:43 AM, Svante Schubert wrote:
>>
>>>   3. Introducing new attributes (e.g. meta:value and meta:value-type),
>>>      but reusing as well the attribute value set of the existing
>>>      attributes (office:value and office:value-type)
>>>
>>> Today I tend to solution three, but I am curious about the opinion of 
>>> Patrick and Michael as editors of the ODF spec.
>>
>>
>> The more I think about it, the more I also agree.
>>
>>
> I understand the desire to re-use existing ODF values but as Bruce 
> points out there are problems with the existing "date" type in ODF.
> 
> I am not sure what little we may gain from re-use is really worth the 
> effort, since we would have to in any case allow any data type in the 
> RDF/XML. Seems to me that to re-use an existing value implies a 
> restriction that we won't be making.

I have no objections to not re-using the datatypes we have for other 
fields if they don't fit. But if we don't do so, we need to know

- whether we want to allow datastyles to display for instance date 
values in human readable ways.
- we want to be able to validate the values (validation is the reason we 
have multiple attributes for other fields).

My proposal would be that
- we support datastyles for exactly those W3C datatypes that are covered 
by the office:value-type attribute.
- we have a meta:value-type attribute that lists those datatypes 
explicitly in the schema, but allows arbitrary other values, too.
- we have a single meta:value attribute, those values cannot be validated.

Michael