OASIS XML Localisation Interchange File Format (XLIFF) TC

 View Only

Re: [xliff] Final Consolidated Restype Attributes List

  • 1.  Re: [xliff] Final Consolidated Restype Attributes List

    Posted 03-27-2003 15:32
     MHonArc v2.4.5 -->
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    
    

    xliff message

    [Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


    Subject: Re: [xliff] Final Consolidated Restype Attributes List


    Title:
    Hi all:

    We've returned to this discussion more than a few times in the past couple of months.  After our most recent discussion of this topic,  I believe it was left that we would explicitly list all possible restypes.  I personally take the view that the list should be normalized down to less granular categories or what Mark calls "root restypes".  Anything requiring a more specific or granular definition should be added via attribute namespace extension.

    What's the consensus of the group on this issue?  If we're split,  then I can set up a ballot using our new Kavi system,  and we can determine consensus that way.  This needs to be resolved before we move on.

    Regards,
    Tony


    Lieske, Christian wrote:
    Hi all,
     
    It might make sense to say a word again about my original motivation for including both
    XUL- and XSD-related items:
     
    I was assuming that we where looking for a kind of 'superset' of items to which items from specific
    sets (e.g. those belonging to Windows RC files) could be mapped. My specific line of thought was
    to create mapping tables which reveal how the individual sets would be mapped to the superset. This
    approach appealed to me since on the one hand the superset would give us a comprehensive layer of
    abstraction, and the mapping tables would leave no doubt about what representations for example
    of Windows RC items have to look like in XLIFF.
     
    The only alternative to this 'superset and mandatory mapping' approach to me would be the consequent
    use of the namespace mechanism (which for example might render Windows RC specific information
    in XLIFF as 'item="winRC:X"'). Of course this alternative approach could boil down to a completely different approach
    for tool suppliers. Possibly, they would no longer work on rendering XLIFF, but would look into XLIFF files,
    see that they are really 'wrappers' for say Windows RC files, and then would use Windows RC
    related functionality for further processing.
     
    Best,
    Christian