Re: Do we need a local markup for Domain & resourceRef

Am 20.03.13 21:45, schrieb Yves Savourel:
>>> The question is: Is it worth to take to pain
>>> to add local markup for Domain?
>> Maybe another question is: who would implement this? In the
>> usage scenarios we have at http://www.w3.org/TR/2013/WD-mlw-metadata-us-impl-20130307/
>> I don't see an example of processing domain information locally - I may be wrong.
> We would support it if the WG decided to define it.

Sure - just trying to understand what usage scenario would back such a 
definition, and eventually what implementation.

> But I'm not asking for more work...
>
>
>> If I'm not wrong and domain information also in XLIFF is only
>> in a global manner, that is not on individual segments: can't you
>> define in the mapping namespace "your" global domain attribute and
>> have a its2xliff rules file to point to it?
> That's not quite true. XLIFF documents can be made of data coming from various source, so potentially one could have different domain in the same document, the lowest level is probably the trans-unit.

ok ... so could you then define the domain attribute in the mapping 
namespace and point to all its occurances in an its2xliff rules file?

- Felix

>   Although technically, if the ITS domain selects an inline element, a domain may be set at a span level. One could do this for example for citations, etc.
>
> I'm not sure that could be supported very easily, but it's technically possible to specify. MT system often allow that too. For example the engine to use is a parameter of the segment-level API in MS Hub, so technically again, one could switch engine every other segment if you use that API.
>
> -ys
>
>
>

Received on Wednesday, 20 March 2013 20:58:37 UTC