Changes from DITA 1.2 to DITA 1.3

DITA 1.3 adds a number of new features and refinements to DITA.

Enhancements to keys

Added scoped keys
Added facility for key-based, cross-deliverable addressing

New XML grammars

The DITA standard now supports RELAX NG.

Enhancements to conditional processing (filtering)

Enabled information architects to associate DITAVAL conditions with specific topic collections in a map (branch filtering). Prior to DITA 1.3, a set of DITAVAL conditions could only be associated with a root map.
Expanded syntax for filtering attributes. The original property attributes (@audience, @platform, @product, and @otherprops) now accept grouped values, using the same syntax for generalized attribute values within @props and @base.

Enhancements to addressing

Added new (abbreviated) syntax for addressing an element within the same DITA topic

Enhancements to specialization

Enabled individual element types of structural specializations to be reused by other structural specializations without requiring one to be specialized from the other. For example, allowed the <steps> element from <task> to be reused in place of an <ol> within a <remedy> section of the troubleshooting document type.

Increased support for producing troubleshooting information

Added a new value, "trouble", for the @type attribute of the <note> element

Increased support for producing user assistance

Added the following new attributes to <resourceid>:
- @appid
- @ux-context-string
- @ux-source-priority
- @ux-windowref
Added a new element, <ux-window>, designed to specify a window or viewport in which an online help topic or Web page is displayed

New base elements and attributes

Added the <div> element. This element enables arbitrary grouping of content within a topic. It is primarily intended as a way to group content for content referencing or as a basis for specialization.
Added new conditional processing attribute, @deliveryTarget, to replace the now-deprecated @print attribute. Users can set controlled values for the @deliveryTarget attribute by using a subject scheme map.
Added @orient to <table> and @rotate to <entry>. These new attributes enable authors to control formatting more precisely for print-based output formats.
Added @scope and @headers attributes to the <entry> element in tables. These new attributes are intended for extremely complex tables where accessibility cannot be handled based on existing header markup.
Added a new attribute, @cascade, to better control how metadata attributes cascade within a DITA map

Modifications to the base content model

Enabled <draft-comment> in more places
Enabled <text> in more places
Extended content model of <data> to allow additional phrase level markup.
Extended the list of values allowed for the @style attribute on <prop> and <revprop> elements in DITAVAL files.
Enabled <ph> within <indexterm>. This makes it possible to use specializations of the <ph> element, such as <sup> and <sub>, within index entries.
Enabled @keyref for <object> and <param>
Added <data> to the content model for lists
Added new default values for the @format and @scope attributes
Added @rev attribute to <title>
Added <cite> to the content model of <title> and similar elements
Added <xref> to the content model for <shortdesc>

Refinements to base domains

Added element for sorting, <sort-as>, to the utilities domain
Added new elements <line-through> and <overline> to the highlighting domain