OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

 


 

   Re: [xml-dev] Rethinking namespaces, attribute remapping

[ Lists Home | Date Index | Thread Index ]

It seems from all the various postings on XLink/HLink, et cetera, that one of the real problems of XML 1.X is a missing data type, the pointer.  If the various attributes could have been declared as:
 
<!ATTLIST    img
    src    PTR    #REQUIRED
        :
        .
>
 
<!ATTLIST    a
    href    PTR    #IMPLIED
        :
        .
>
 
Then we would have had a way to identify links [anything with the type PTR], regardless of the name of the attribute.  So if want to correct the HLink/XLink fiasco, why not add a type of attributes that indicates that it is a pointer.  We might event use the notation mechanism so that there can be a way to identify validation constraints on pointers, without requiring predeclaration of every pointer in an external entity:
 
Thus:
<!NOTATION  URI PUBLIC "some agreed upon string for href-style URIs" >
<!NOTATION XPointer PUBLIC "some agreed upon string for XPointer URIs" >
 
Would declare "Attribute types"
 
which could then be used like:
<!ATTLIST    a
    href    NOTATION    XPointer    #IMPLIED
>
 
<!ATTLIST    img
    src    NOTATION    URI    #IMPLIED
>
 
So anyway, if we [as Tonto once said -- We who, Kimosabe?] end up refactoring XML, I think that something like this should be considered.  It seems to eliminate a lot of the trouble posed by links today, without a lot of new syntax, and no colons necessary [for those Averse Religiously to such devices].
 
    Keith
 
 
Engineering is what happens when science and 
mathematics meet politics.  Products are what
happens when all three meet reality.         
 




 

News | XML in Industry | Calendar | XML Registry
Marketplace | Resources | MyXML.org | Sponsors | Privacy Statement

Copyright 2001 XML.org. This site is hosted by OASIS