Release Notes v6.32

What's new?

• Created URSA-USAGE module, to help customers track on their own Ursa Studio usage metrics.
• KEEP INTERVAL IF OVERLAPPING and IS OVERLAPPING INTERVAL patterns.
• Added a category filter to the User Manager dashboard.
• Added utility to clean up duplicate value set values or lookup table keys, for RBDMSes that do not support unique contraints. The cleanup tool is right alongside the duplicate warning.
• Added option to include database optimizations in module submission. By default these optimizations are not carried into the target environment.
• Add a utility into the hamburger menu of Object Workshop to generate performance tuning recommendations. This utility will search across three areas:
  • List unused fields. Fields on an object are "used" if they are (a) used in a restriction/reshape pattern on an immediate downstream object, or if they are (b) published on an immediate downstream datamart or measure, or (c) they're published downstream some number of generations before triggering (a) or (b).
  • Fields for which all the rows of the current table for that object are null.
  • (Redshift only) Explain plans containing the following red flags: DS_DIST_INNER, DS_DIST_OUTER, DS_BCAST_INNER, DS_DIST_ALL_INNER, DS_DIST_BOTH.

What improvements have we made to existing features?

• Allow users to hide revision comments for objects, measures, reports, models, and semantic mapping templates. Hidden revision comments will, by default, not appear in the revision history table, but there will be an option to reveal the hidden comments, whereupon the comments can be permanently unhidden.
• When cloning assets or performing a "recover-as-new" to an asset, the revision history of the asset will be carried over to the newly created asset. Any comments that are marked as hidden at the time of the creation of the new asset will not be carried over.
• Allow the creation of cohorts for integer fields, including the application of integer cohorts as filters against integer fields. Integer cohorts can be used as a filter against string fields, but string cohorts cannot be used as a filter against integer fields.
• If users enter strings in the "Consider as Null" input in semantic mapping objects in the form YYYY-MM-DD, then that guard will be used against date fields of mapped registered table objects.
• Allow filter dimension to be set when pinning maps to boards.
• Merge "top 5 list" board tile option into the more fully-featured "data table" board tile option.
• Make the Saved ELT table sortable.