Last week, we released a new, tested, binary release of the integration platform (the source has been on CodePlex for a while).  This release primarily contains a bunch of bug fixes.  The bugs came from our internal use of server to server synchronization tools, customer reports and internal testing. http://visualstudiogallery.msdn.microsoft.com/f854dd59-8eeb-4673-8d9a-ae012989bfa2 The list of bugs fixes is:

• VC adapter: code review should use cloaked path for pending change comparison
• VC adapter: Branch, Merge, Edit code review failure
• 2 way sync conflict not handled correctly
• Created Date greater than Resolved Date for some WIs ported by TFS Migration tool service
• VC adapter: Branch, Merge, Edit code review failure
• Icons branding shell evoke tools and toolkit
• Dev10 Work Item Tests are hard-coded to use ReflectedWorkItemId field
• Integration service (SqlChangeGroup) threw OOM exceptions
• A directory , which contains an sub directory which is not visible over the Config Spec causes tool to exit with error.
• VC adapter: content conflict resolution (take right) is not working when conflict was raised while pipeline flows from right to left
• Shell: help link in edit view is not working
• TFS 2010 WIT adapter: custom network-topology link gets deleted in bi-directional sync
• VC migration does not correctly handle the case of multiple filter paths when one path is a child of another
• TFS merges with identical content should appear as changes in CC
• Labels added to ClearCase when “LabelAllVersions” is true should be logged at Info level
• Bi-directional WIT tests: right side filter string generated does not included the work items migrated from the left to right
• WIT: Migration / sync of work item to work item links does not occur until a field change is made to one of the work items
• nightly WIT tests are failing with nullref
• WIT migration/sync: Conflicts of type WITUnmappedWITConflictType and WorkItemTypeNotExistConflictType should not cause the current trip to stop
• Some changesets are migrated from TFS to CC more than once
• ClearCase (forum post): User unable to configure ClearCase to TFS migration in Shell
• ClearCase (forum post): User unable to run ClearCase to TFS migration due to exception parsing date in CC history row
• Null Ref exception in Subversion Adapter
• Null Ref exception in SubversionShellAdapter
• Undelete tests are timing out due to unresolved code review conflicts
• FilterPairs with AreaPath values block on ValidationException: TF51011: The node specified is not found in the classification hierarchy
• Docs should say that TestCases migration is not supported
• Null reference exception occurs in ClearCase adapter when CC ls command output is empty
• ClearQuest – TFS sync: Duplicate bugs are created in TFS
• ClearQuest – TFS sync: Editing a bug in TFS that originated as a CQ Defect causes Runtime Error with COMException on call to ClearQuest API
• ClearQuest – TFS sync: editing State field of a TFS Bug causes ArgumentNullException and stops sync if change would result in invalid CQ state transition
• ClearQuest – TFS sync: adding an attachment to either a CQ Defect or a TFS Bug does not work
• ClearQuest adapter fails to detect changes in ClearQuest when timezone is greater than UTC because it uses local time for the HighWaterMark
• ClearQuest – TFS sync: changing the only the State of a ClearQuest Defect does not cause it to be sync’d until another field is changed
• Dogfood: HWM data stored in SYNC_POINT table is wrong causing ServerDiff WIT to ignore many differences (and other problems)
• Pioneer Upgrade: Hierarchy links don’t sync properly
• WIT sync fails for fields of type double with “decimal point is comma” setting
• ClearQuest – Migration is blocked by error on query for ClearQuest items to migrate if the CQ database server uses a Date format that is not the ISO standard format
• ClearQuest – ServerDiff Wit command does not work at all when the CQ filter string specifies a CQ stored query (very common)
• Key constraint violation occurs on the TFS_IntegrationPlatform DB when resolving a Namespace Conflict

Together, these represent a significant improvement in stability, correctness and manageability.

Brian