Funnelback 13.2.0 release notes

Funnelback is now able to perform website accessibility checking against version 2 of the W3C’s web content accessibility guidelines (WCAG 2.0). See Accessibility_Auditor for further details.

Additional sources of information (e.g. "related clicks" and "related results") are now used in the Recommendations system, improving the quality of suggestions.

Improved crawl consistency when restarting from crawler checkpoints.

Presenting Images).

Ability to calculate upper and lower bounds for numeric metadata.

Improvements to result summaries (removing duplicate sentences and noscript content).

Separate configuration of Funnelback webserver ports and link-generation ports to better support load balancer configurations.

Inclusion of a high-performance XML splitting tool.

Improved performance of user account editing.

Fixed memory and file handle leaks in libi4u that could cause issues when using recommendations or search session and history.

Users will need to have 'wcag' role to access the WCAG Compliance Auditor (previously was report_viewer). Grant access in administration dashboard under Manage Users.

WCAG databases from earlier un-bundled accessibility checker versions must be upgraded by running the following

  # Linux
  $SEARCH_HOME/linbin/java/bin/java -cp "$SEARCH_HOME/lib/java/all/*" com.funnelback.wcag.database.AccessibilityDatabaseUpgrade collection_name
  # Windows
  %SEARCH_HOME%\wbin\java\bin\java -cp %SEARCH_HOME%\lib\java\all\^*" com.funnelback.wcag.database.AccessibilityDatabaseUpgrade collection_name

Custom collections no longer preserve data between crawls by default. Collections which need to preserve data will need to implement that behavior the custom_gather.groovy file.

The convertRelative parameter on the <@fb.IncludeUrl> tag now supports both syntax: convertrelative and convertRelative