Re: [dita] Some bookmap updates

[Robert D Anderson <robander@us.ibm.com>]

Hi Nancy,

The idea with this the change to lists is that users may want a special
type of index. If they do, then they would logically try to specialize off
of the <indexlist> element. However, doing so would also prevent them from
including both types of index - you can use the original, or the new type,
but not both. I know that some IBM books contain a part number index (our
SGML DTD allows for one per book). The author index was another suggestion
I heard. Changing the content model allows users to create a specialized
bookmap that with a model of (indexlist?, authorindex?, partindex?) in
place of indexlist*.

The user request that came to Don specifically called out indexes, but the
same argument could be made for the other lists. Users could use
<tablelist> for all tables, or they could specialize <ordinaryTableList>
and <myOptionsTableList>. For figures, users could create
<syntaxdiagramList>. For trademarklist, your organization might distinguish
between types of trademarks and need to list them separately, or on some
occasions, only list one type. I've just now come up with the ones in this
paragraph, so they do not represent actual user requests, but I'd guess
that realistic cases exist.

Does that make sense?

Thanks for the review-

Robert D Anderson
IBM Authoring Tools Development
Chief Architect, DITA Open Toolkit
(507) 253-8787, T/L 553-8787

> * Change the cardinality of list elements from (0 | 1) to (0 to many). An
> external comment indicated that some users need to create special types
of
> indexes (samples given were keyword indexes, author indexes, and a part
> number index). The same logic applies to other list types.
>
> ***
> not sure I understand the reasoning here
> ***
>