RE: [xml-dev] xml:href, xml:rel and xml:type

["Rushforth, Peter" <Peter.Rushforth@NRCan-RNCan.gc.ca>]

Hi Len,

> 
> It's a side issue but what relationship is explicit by enclosure?  
> 
> IS-A or HAS-A?
> 
> Seems simple enough but this is where this has fallen apart 
> in the past.
> We think it is a simple, easy to recognize relationship and 
> it turns out what we think we know just ain't so.  
> 
> The evidence only supports HAS-A.

OK, I don't know.  I always thought of it as "I use it any way I
see fit", but the mechanism is static.  It is the mechanism that
yields some interoperability though.  It's a hard-wired convention.

> 
> A post-2000 discussion of architectural forms reinforces the 
> basic dilemma of web technology:  it is a system attempting 
> to be THE systems of systems and I've no knowledge of any 
> system succeeding at that for
> more than one turn of the wheel.

Sorry. Are you talking about architectural forms, or the web?

>  A lot of nonsense can be avoided by
> saying "We are only spec'ing this for the web.  Full stop." 
> 
> And then not nudging or winking that the only system of 
> concern for others is the web.  That is the nonsense that 
> defeats separation of concerns.

Is the proposal for xml:href, xml:rel and xml:type is
such a nudge or a wink? _I_ think it's a proposal for a simple hard-wired 
convention, something analogous to the nesting of elements implying a relationship
between whatever those elements are representing.

Cheers,
Peter