Re: Handling HTML and FO semantics

[Norman Walsh <ndw@nwalsh.com>]

<rob.cavicchio@emc.com> writes:
> I fear that this has the potential to make documents too
> output-specific.

I'm not sure there's anything output-specific about them, per se.
There's nothing to suggest that you couldn't take advantage of the
HTML rel semantic in some other system.

> I would say that in your two examples, using HTML and
> FO namespaces would inaccurately characterize the attributes as
> output-specific.

I could make the counter-argument that rather than being
output-specific, I'm explicitly asserting very particular semantics
with my elements.

> Link relationships could apply to many potential
> output formats,

That's true. Are there any significant examples other than HTML's @rel
out there?

> and keep/break formatting certainly does.

It does. But note that expressing it as fo:keep-together="always"
associates a very particular semantic with the element. That doesn't
prevent someone using DSSSL or some other publishing system from
observing and correctly interpreting that semantic. (Though perhaps
there are more exotic fo: attributes where that would be more
difficult to imagine.)

> Furthermore,
> a link relationship is not a presentational attribute, but a semantic
> one, and as such I think it belongs in the semantic markup.

I'm not sure I want to argue about whether @html:rel is presentational
or semantic. Though if I did, I think I'd be forced to argue that it
captures a very specific semantic not presentation. (In fact, AFAICT,
browsers make no presentational distinction based on @rel values.)

This description certainly strikes me as mostly semantic rather than
presentational:

  http://dev.w3.org/html5/spec/Overview.html#linkTypes

> Currently
> this could be done using one of the DocBook effectivity attributes.

Well. Maybe. I have to admit I didn't think of that. Of course, the
effectivity attributes all have the shortcoming that they're local not
global.

   <link condition="author">...</link>

might mean @rel="author" to me, but it certainly doesn't convey that
in any general sense.

> If
> it might be widely used, perhaps it should be under consideration as
> an additional attribute in a future version of DocBook.

Yes, that wouldn't be unreasonable for @rel.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com>      | As charms are nonsense, nonsense
http://www.oasis-open.org/docbook/ | is a charm.--Benjamin Franklin
Chair, DocBook Technical Committee |

PGP signature