Funnelback 10.0.0 release notes

Released: 13th October 2010

New Features

  • Query completion suggestions available on search forms.

  • Quicklinks to sub pages available for top search results.

  • Explore option for discovering related results from a single starting result.

  • New .warc storage option for very large web collections.

  • Simplified collection configuration files.

  • Event search custom display mode.

Upgrade Issues

  • ActivePerl is no longer required to be installed separately to Funnelback, as it is now bundled with the Funnelback installation. You may wish to remove any prior version of ActivePerl as part of your upgrade process.

  • Analytics, since version 9.0.0, have included requests where start_rank was greater than one (i.e. where something other than the first page of results was viewed). This change has been reversed in version 10, which may decrease the reported volume of queries.

  • Please be aware that stemming is now enabled for new collections by default. Stemming can be disabled by removing the -stem2 query_processor_option.

  • To simplify administration, a collection’s collection.cfg file will now contain only settings which differ from the default values defined in $SEARCH_HOME/conf/collection.cfg.default or $SEARCH_HOME/conf/collection.cfg. Any setting may be copied into the collection’s collection.cfg file to override the value set in one of the shared configuration files, and the administration user interface now provides a simple drop-down menu listing all of the inherited options and their default values.

Selected improvements and bugfixes

  • Improved logging for analytics updates.

  • Improved handling of sites with base hrefs.

  • Better handling of passwords containing special characters during installation.

  • Fixed halting of web collection instant updates.

  • Handling of binary document URLs without file name extensions using in-line filtering and mime types.

  • Display of 'Top Query' for contextual and faceted navigation reports.

  • Improved handling of existing in-place data during web crawling.

  • Improved quality of title fixing.

  • New interface and syntax highlighting for collection.cfg editing.

  • Improved TRIM metadata mapping.

  • Click tracking for RSS feed results.

  • Improved exclude pattern handling of special characters in URLs during web crawling.

  • Better handling of phrase queries with s:Select tags.

  • Logging for padre error messages and query logging permission errors.

  • Inclusion of header row in CSV analytics exports.

  • Sorting of administration interface user list.

  • Fixed handling of query changes within a faceted navigation category.

  • Improved interface for editing web collection include and exclude patterns.

  • Activated basic search query stemming for new collections by default.

  • Added support for a server-wide reporting blacklist in addition to collection specific ones.

  • Improved date sorting for collections including dates in the future.

  • Improved scrolling behaviour within form and long config file editing.

  • Improved charting boundary conditions within analytics reports.

  • Changed analytics to exclude queries with a start_rank greater than one (i.e. subsequent result pages).