OASIS Darwin Information Typing Architecture (DITA) TC

Currently, the DTD module pattern requires declaration of parameter entities
for each element type defined in a module for later use in content models.
These parameter entities are in the .mod file.

However, with the advent of the ability in DITA 1.2 to do per-element
content model overrides, I think these entities should be in the .ent files,
so that they are available to content model overrides specified in shell
DTDs.

For example, if I have shell "myconcept.dtd" and I want to extend the
content model of title to allow a new phrase type (e.g., "my-title-phrase"),
without this change I have to declare the entities for each of the content
model tokens I'm retaining from the base content model, since they otherwise
occur *after* the shell-defined content model override entity is declared.

I don't remember any discussion of this in the past, but I also haven't been
necessarily paying close attention.

Is there any reason not to make this change?

Cheers,

Eliot

----
Eliot Kimber | Senior Solutions Architect | Really Strategies, Inc.
email:  ekimber@reallysi.com 

This sounds like a reasonable change to make in 1.3.  I'd really 
rather we didn't keep doing major reorganization to the 1.2 files
at this time.

paul

> -----Original Message-----
> From: Eliot Kimber [mailto:ekimber@reallysi.com]
> Sent: Wednesday, 2009 July 22 9:31
> To: dita
> Subject: [dita] Organization of Tagname Parameter Entities in DTD
Modules
> 
> Currently, the DTD module pattern requires declaration of parameter
entities
> for each element type defined in a module for later use in content
models.
> These parameter entities are in the .mod file.
> 
> However, with the advent of the ability in DITA 1.2 to do per-element
> content model overrides, I think these entities should be in the .ent
files,
> so that they are available to content model overrides specified in
shell
> DTDs.
> 
> For example, if I have shell "myconcept.dtd" and I want to extend the
> content model of title to allow a new phrase type (e.g.,
"my-title-phrase"),
> without this change I have to declare the entities for each of the
content
> model tokens I'm retaining from the base content model, since they
otherwise
> occur *after* the shell-defined content model override entity is
declared.
> 
> I don't remember any discussion of this in the past, but I also
haven't been
> necessarily paying close attention.
> 
> Is there any reason not to make this change?

I think I'd also prefer to hold it to 1.3. I know I tried to do something
of the kind earlier in the process, to make constraint definitions easier,
but ran in to trouble with domains. Moving the definition of some entities
to the ENT files meant that they were defined ahead of the domain section,
meaning that you couldn't extend the elements by re-defining them (first
definition in the ENT file won). At least, that's what I think happened.

I'm not sure that this is the same thing, but it sounds similar enough that
I'd really rather hold off until we have more time to test it. Also - I'm
not sure I entirely understand the issue - which makes me vaguely wonder if
it's not already possible to do it today (that is, the only place I've seen
any trouble from the location of entity definitions is with constraints).

Robert D Anderson
IBM Authoring Tools Development
Chief Architect, DITA Open Toolkit

"Grosso, Paul" 

On 7/22/09 9:58 AM, "Robert D Anderson"