Re: [Bug 7198] can inheritable attributes appear, on elements with simple types

On 3 Aug 2009, at 22:30 , bugzilla@wiggum.w3.org wrote:

> http://www.w3.org/Bugs/Public/show_bug.cgi?id=7198
>
>
> --- Comment #3 from Mukul Gandhi <gandhi.mukul@gmail.com>   
> 2009-08-04 04:30:25 ---
> (In reply to comment #1)
>
> Thanks for the explanation. I think, I am now quite clear about the  
> inheritable
> attributes concept.
>
> ... Inherited attribute can be referred
> in conditional type alternatives, and I think, also in assertions.
>
> So is it right to believe, that inheritable attributes are designed  
> in the
> spec, to be usable only in conditional type alternatives (CTA), and  
> assertions?
> Or they can have usage, other than these two facilities (CTA and  
> assertions)?
>

In connection with bug 7242, reported today, I have just reviewed
the text of the relevant portions of the spec; I apologize for not
doing this earlier.  I should have corrected the misimpression
reported in the second paragraph quoted above:  in reality, as
currently specified, inherited attributes are visible in the XDM
instances used to test conditional type assignment, but not in
the XDM instances used to test assertions.  It's a natural thing
to think the two cases will be, or should be, treated the same way,
but for technical reasons the attributes used in tests for
conditional type assignment are untyped, while the WG elected to
make the attributes and descendants used in testing assertions
carry their full PSVI information, including their type assignments
and validity.  In that context, including inherited attributes
would introduce some serious complications:  either we would need
a story about why the type assignments are not consistent from
one instance to another, or we would need some way to ensure that
type assignments are consistent.  We were able to resolve bug 5003
successfully largely because the technique of using untyped
attributes allowed us to avoid those complications.  (The treatment
of assertions would have been much simpler, too, if we had used an
untyped XDM, but the WG chose otherwise.)

So my earlier response to your second question was misleading:
instead of saying that inherited attributes were designed to
be useful only for assertions or conditional type assignment, I
should have said:  really, only for CTA.  (The definition of a
term for them in the PSVI does not actually make much difference
in practice, since the information in question has always been
present in any case:  any attribute among the [inherited attributes]
of an element is necessarily present among the [attributes] of
an ancestor element.)

Sorry for not noticing this point earlier.

-- 
****************************************************************
* C. M. Sperberg-McQueen, Black Mesa Technologies LLC
* http://www.blackmesatech.com
* http://cmsmcq.com/mib
* http://balisage.net
****************************************************************

Received on Friday, 7 August 2009 20:29:00 UTC