Re: DOCBOOK: RFE 136: co in userinput

[Norman Walsh <ndw@nwalsh.com>]

/ Nik Clayton <nik@nothing-going-on.demon.co.uk> was heard to say:
| > What we'd really like is <co> as an inclusion on <screen> but we can't
| > have that.
| 
| Why not?

Because XML doesn't have inclusions.

| > In order to fix this in XML, we're going to have to add co to the
| > content model of some elements. As a consequence, the DTD will not be
| > able to distinguish co in userinput in screen from co in userinput in
| > para.
| 
| Why is that an issue?

Some people on the TC are concerned that callouts inside paragraphs
are not motivated by a use case and could result in authors producing
documents for which there was no useful processing expectation.

| > I suggest:
| >
| > 2. We add co to the content model of prompt, userinput, computeroutput,
| >    replaceable, and systemitem. (And nothing else.)
| >
| > Comments?
| 
| This would add extra markup requirements.
| 
| Currently, my DocBook DTD extension has this in it:
| 
|     <!ENTITY % local.cptr.char.mix "|CO">
| 
| which lets me put co inside screen.

That's unnecessary. CO is already allowed inside screen and no one is
suggesting that we remove that. The problem is CO inside userinput
inside screen.

The question is, what elements besides prompt, userinput,
computeroutput, replaceable, and systemitem do you use inside screen
inside of which you might want a CO.

                                        Be seeing you,
                                          norm

-- 
Norman Walsh <ndw@nwalsh.com>      | Waste no more time arguing what a
http://www.oasis-open.org/docbook/ | good man should be. Be
Chair, DocBook Technical Committee | one.--Marcus Aurelius