Schema Compare for Oracle v3.0.2.348 released

richardjm

My second big release of the Deployment Suite for Oracle is now available via check for updates. The full release notes for Schema Compare for Oracle are below.

Features

• Attempt to detect column renames. When a potential column rename is detected, a "map" icon is shown next to the object name in the comparison results. Column rename detection can be turned on or off in the Comparison Options. By default, rename detection is turned for new projects and off for projects created in earlier versions of Schema Compare
• Populating from scripts progress messages now match those for live population

Command line

• Command line options now match options in the GUI
• TNS name lookup is now case-insensitive
• SCO returns exit code 62 if the deployment fails
• Abort-on-warnings option returns exit code 63 and aborts deployment if medium or high-level warnings occur during a /deploy run
• Warnings are displayed in the command line

Warnings

• The Warnings page in the Deployment Wizard is shown if there are warnings to be reviewed
• Warnings are no longer duplicated
• Skip/Ignore warning for non-existent users is no longer ignored
• Dropped table and dropped column data loss warnings are now high- level warnings

Fixes

• Saving an individual snapshot no longer fails to load schemas if no Oracle client is installed
• A reports directory is now created on startup, preventing errors when you tried to save a report without specifying a directory

Schema Compare Engine

• Rename column detection
• Scripts folder population is approximately 25% faster
• Improved nested table support
• Materialized view definitions are no longer always in brackets
• Materialized view comment support - default comment excluded
• Skip GRDMLTR_* (auto-generated) triggers for SDO_GeoRaster columns
• Copy NOVALIDATE for constraints if specified
• Removal of some unused schema population
• Improved logging of population speed
• Joined indexes depend on all joined tables
• Script parsing errors:
  • Interval literal support for materialized views e.g. "NEXT trunc(sysdate,'HH') + interval'1'"
  • Fixed view parsing to cope with "cast(trunc(systimestamp, 'DD') as timestamp)"
  • Parsing of materialized view starting with newline would fail
  • Allow U'' style string literals
  • Tables with external data failed if column names ended in "location"

Release notes history