Re: unsatisfied

Le Friday 21 March 2003 17:30, Paul Grosso a écrit:
> >
> >Paul, my last comments and your reactions to these comments for comment
> > #15 have not been updated.
>
> Comment 38 is where we processed your reply to our response
> to comment 15.

Ah, OK. I was thinking you would do an item #2 for that one

> Unfortunately, I am just the chair, not the editor of the
> Comments doc or the entire SG, and some other SG members
> tend to be more "terse" that I would be, so the official
> reply on comment 38 is less explanatory than I would like.
>
> In fact, the SG agrees that datatype handling in XSL is
> an area that could use more work, but such work is not
> within scope for an erratum.

I agree that would need to be addressed, really.

>  Furthermore, we are somewhat
> constrained by our requirement to support the CSS syntax
> which often causes problems for the XSL FO expression language.

I'm not sure how much this prevents from doing that in a much simpler manner.

> So for now, the point is that the <uri-specification> datatype
> is really a subtype of <string>.

This point should be stated in the spec then.

>  That is, the value of the
> src property follows all that for something of the datatype
> <string> but then in addition also has to follow the extra
> constraints defined by <uri-specification>.
>
> As such, the unusual quoting you noticed is, in fact, allowable.
>
> On the other hand, I'm not sure why that unusual quoting was
> chosen for the example in the first place, as I agree that isn't
> the way I'd usually quote such things.
>
> So if you still feel we have misinterpreted the spec as it now
> stands, please submit another comment.

I'm fine if <uri-specification> being a "subtype" of <string> is stated 
clearly in the spec. It's not a natural nor simple way of doing that, but I 
can live with it.

But so far the "subtype" aspect was not very clear in the spec.

-- 
The war against Iraq is both unjustified, illegal and amoral.