Re: [xml-dev] So maybe ID isn't a problem after all.

[David Brownell <david-b@pacbell.net>]

> To summarize, I think the reasonable choices are:
> 
> (a) Remove the RawName construct in XPointer
> (b) Change the semantics of the RawName construct in XPointer to say that 
> it refers to, say, a element with a xptr:name attribute with a particular 
> value
> (c) Add an xml:idatt(s) to XML

Seems like (b) is isomorphic to the xml:id construct, so if you
believe xml:id doesn't work, then for the very same reasons
it'll be true that xptr:name doesn't work.  Both have the same
problem transitioning old apps to new attributes.  Not unlike
moving to "id=..." in current HTML; solvable.

On the other hand, I'm not sure I bought that claim, or the
problem definition requiring DTD stripping not to discard any
"important" information.

I clearly don't want to see (a) happen, and (c) does seem to be
avoidable complexity (so many in-line declarations).

- Dave