Branch filtering

[Bob Thomas <bob.thomas@tagsmiths.com>]

I would like to write a conditional-processing white paper before writing a branch-filtering white paper. Approaching branch-filtering without a solid understanding of conditional-processing is an exercise in futility.

1.3 has added new features to conditional-processing such as groups and the @deliveryTarget attribute. This paper is also necessary because I don't believe that one was written for 1.2 conditional processing.

Once the conditional-processing white-paper has been written, we will have a solid foundation for discussing branch filtering.

My initial thoughts are that they paper would be structured somewhat like this:

Introduction

Basic capabilities

• product attr with 2 values, abstract filtering settings
• product attr with 2 values, ditaval filtering settings
• product attr and audience attr each with 2 values each, ditaval settings
• Greedy nature of exclude
• One attr evaluating to exclude kills the element irrespective of the other att eval
• Element evaluating to exclude poisons that element and all of its descendents
• ditaval include/exclude defaults
• flagging

Groups

• groups, simple example
• multiple groups
• implicit groups
• ditaval for groups

Delivery target domain

Best practices

• managing complexity
• localization
  

One advantage to this structure is that the basic capabilities section would also be applicable to DITA 1.2.

Best Regards,

--

Bob Thomas

+1 720 201 8260

Skype: bob.thomas.colorado

Instant messaging: Gmail chat (bob.thomas@tagsmiths.com) or Skype

Time zone: Mountain (GMT-7)