Those of us who continue to use SGML are kind of forced into opt 2, are we not?
I'd prefer to avoid further divergence between the two versions (SGML/XML).

Mark

At 03:34 PM 11/30/01 -0500, Norman Walsh wrote:
>/ Eduardo Gutentag <[EMAIL PROTECTED]> was heard to say:
>| Option 2 makes much more sense to me.
>
>Why?
>
>My thoughts are:
>
>1. We're moving towards more modular, reusable documentation. Sooner or later,
>    probably sooner, someone's going to want to include a fragment that 
> uses one
>    table model along with a fragment that uses the other. And that won't be
>    possible.
>
>2. This is exactly the problem namespaces are supposed to solve, isn't it? :-)
>
>3. Won't tool vendors have to support mixed namespaces "real soon now"
>    anyway, for things like XLink, SVG, MathML, etc. So tools will
>    actually be able to handle this?
>
>| Norman Walsh wrote:
>| >
>| > See 
>http://sourceforge.net/tracker/index.php?func=detail&aid=472229&group_id=21935&atid=384107
>[...]
>| > There appear to be two solutions:
>| >
>| > 1. Use namespaces.
>| >
>| > 2. Force the user to make a top-level choice by having, effectively,
>| >    two DTDs. This would mean a document could contain *either* HTML tables
>| >    *or* CALS tables, but not both.
>
>                                         Be seeing you,
>                                           norm
>
>--
>Norman Walsh <[EMAIL PROTECTED]>      | He who fails to become a giant
>http://www.oasis-open.org/docbook/ | need not remain content with being
>Chair, DocBook Technical Committee | a dwarf.--Ernest Bramah

Mark B. Wroth
<[EMAIL PROTECTED]>

Reply via email to