Fixes Defects: PM48938,PM56094,PM59537,PM66907,PM67147,PM67399,PM67792,PM68052,PM68213,PM68658,PM68896,PM69279,PM69330,PM69503,PM70374,PM70385,PM70878,PM70970,PM70972,PM71714,PM72176,PM72260,PM72918,PM72979,PM73326,PM73432,PM73586,PM74114,PM74428,PM75101,PM75238,PM75306,PM75703,PM76372 This patch contains the following software changes: PM48938: Problem: When a time out occurs in a communication from ClearCase dynamic view client to the view server, then there might be EIO errors in the system log on the client and performance of the view server process could degrade. Fix: Fixes a problem in ClearCase which leads to degraded performance of the view server process and EIO errors in the system log of the ClearCase dynamic view client machine. PM56094: Problem: Changing the Extended Naming Suffix was not working on 64-bit Windows machines. So, the default suffix (@@) is still in effect. Fix: Corrects a problem where the Windows system registry key which stores a non-default value for the ClearCase extended naming suffix was redirected on 64-bit Windows systems, so it was invisible to MVFS. The registry key was changed to a nonredirected location.  P PM59537: Problem: ClearCase and ClearQuest removed the original MSI packages used to install 3rdParty components, upon cleanning up the Windows MSI cache, the packages had problems to perform Installation Manager Operations. Fix: Updated the ClearCase / ClearQuest to keep a copy of the original MSI packages inside the Install directory in case the machine's MSI cache is cleared, in this situation the MSI system will have the original MSI path and prevent problems on the Installation Manager Operations. PM66907: Problem: mklbtype -acquire -global allows you to acquire a lbtype that is locked The end result is a unlocked global lbtype with a locked local copy Fix: An inconsistency in the handling of locked label types was fixed; if the client type is locked, it is reported as not acquirable, and the operation continues with other types. PM67147: Problem: When attempting to create a Managed Code "wrapper" for the CCPolicy put_value() method, there seems to be no way to get Visual C++ 2010 to accept the CCPolicy::put_value method. Fix: The definition of the CCPolicy put_value variable was changed from BSTR* to BSTR, to correct a problem with use of that method with some compilers. PM67399: Problem: ClearCase Automation library crashes when called from Multi-threaded code Fix: A race condition caused by incorrect locking was fixed in the ClearCase Automation Library. PM67792: Problem: User could not determine how to create a key binding to navigate to the ClearCase Details view. Fix: Added a section on Accessibility features to the "Developing -> Developing software with Rational ClearTeam Explorer -> Team software development with ClearCase" Help topic and used the case of creating a key binding to navigate to the ClearCase Details window as a specific example. PM68052: Problem: The ClearCase registry may contain files with .new and .old extensions. Fix: This fix attempts to clean up these .new and .old registry files when the albd_server is starting, by restoring the registry file corresponding to the .new or .old file, and then removing the .old and .new files. PM68213: Problem: The IBM Rational ClearCase type manager for text files (text_file_delta) can handle individual lines up to 8000 bytes or characters in length. Normally if a user attempts to check in a text file containing a line over 8000 characters an error will be presented and the checkin will fail. In a case where a text file is checked in containing only 7998 bytes and no newline character the checkin will also fail. Fix: A problem in an edge case when calculating line length in the text_file_delta type manager has been corrected. PM68658: Problem: ClearCase online help might fail to launch properly after being used once, the system will throw back an error about library relocation. Fix: Updated the underlying runtime to a version that correctly handles the library relocation. PM68896: Problem: User will see label of type "label_name" already on element error when mklabel recursively if some branch is locked with -nuser while the user is not in the -nuser list. Fix: A change was made in the transaction handling of certain use cases of cleartool mklabel, to prevent incorrect error messages. PM69279: Problem: Lock manager tables can become corrupt and non-recoverable without manual intervention. Fix: Certain system-level error conditions can trigger this defect; sanity checking was added in the code to avoid the possible deadlock. PM69330: Problem: In older versions of the IBM Rational ClearCase cleartool man page for protectvob it states, "Execute this command only while a VOB is quiescent (no active users); it stops and restarts the associated vob_server process, which prevents access to storage pools." The man page in version 7.x and 8.x do not clearly state this.  Since this is not stated clearly some customer have attempted to run the command on a VOB being accessed and protectvob failures are seen with errors such as: vob_server.exe(2644): Error: Temporarily unavailable, admin process running as pid 65535 Albd(2284): Error: Server vob_server.exe (pid=2644) on "C:\vobs\VOB1.vbs" died on startup; marking it as "down". vobrpc_server.exe(3292): Error: Problem starting vob_server for vob vobserver1.company.com:C:\vobs\VOB1.vbs vobrpc_server.exe(3292): Error: See albd or vob error logs on host vobserver1.company.com WORKAROUND: Run the protectvob command only on a VOB which is not being accessed at all by any user or process. Fix: The following text has been restored: The protectvob command is used to change various properties of a VOB. Before executing this command, log in to the host where the VOB storage directory resides with the following credentials:   - UNIX system or Linux: root   - Windows: local administrator, VOB owner, or member of the ClearCase group Execute this command only when there are no active users of the VOB; it stops and restarts the vob_server process, which prevents access to the storage pools. PM69503: Problem: After manually selecting one or more file/folder elements in the "ClearTeam Details" tab in the ClearCase Remote Client (CCRC)/ClearTeam Explorer (CTE), and then changing the sort criteria, results in the selected elements being different ones from the initial ones. Fix: In CCRC/CTE, the current selection in the Details view is now maintained when the columns are resorted. PM70374: Problem: Clearmake failures occur when the view_db_crs file grows past 2 GB. Fix: The maximum size of the view_server's configuration record file was increased from 2Gb to 4Gb. PM70385: Problem: ClearCase 8.0 Information Center does not document IHS 8.0 SSL configuration. Fix: Documentation pertaining to IHS 8.0 SSL configuration has been updated in the ClearCase Information Center. PM70878: Problem: Cleartool crashes when using ct describe -fmt '%[versions]p' on an activity with a very large changeset Fix: A segmentation fault in certain use cases of cleartool commands describe and lsactivity has been fixed. PM70970: Problem: ClearCase primary and auxiliary groups not honored correctly on Windows platforms in the ClearCase Remote Client (CCRC) WAN Server. Fix: A change was made to to ensure that the environment variables related to the groups specified, are updated earlier in the processing of the command. PM70972: Problem: When specifying groups within ClearTeam Explorer, 770 permissions within a VOB cause permission denied when not owner Fix: A change was made to ensure that the environment variables related to the groups specified, are updated earlier in the processing of the command. PM71714: Problem: When running an IBM Rational ClearCase server on Microsoft Windows 2008 or 2008 R2, a sporadic port allocation failure may occur. Fix: Logic was added to correctly interpret winsock errors in certain conditions on Windows 2008 and 2008 R2. PM72176: Problem: Users experience "dead connection" error in ClearQuest when attempting to deliver ClearCase activities in a ClearQuest enabled project, using the ClearTeam Explorer, CTE.  This was caused by too many ClearQuest sessions and subsequent database connections being created during CCRC operations involving integration with ClearQuest. Fix: A change was made to ensure that the ClearQuest sessions do not increase with each operation by correctly including the JSESSIONID in request headers. PM72260: Problem: SQUID tracing does not work through the CCBE. Fix: A problem with the logging calls used in the server code was corrected.  Please refer to technote #1127287 for additional details on correctly enabling tracing. PM72918: Problem: In ClearTeam Explorer, when the selection state of a folders' contents were modified, the resource tree did not correctly update the selection state of the folders. Fix: Folder nodes in the tree correctly reflect when their contained resources are unselected, partially selected, or fully selected in ClearTeam Explorer. PM72979: Problem: Using the IBM Rational ClearCase Remote Client (CCRC) plugin to migrate to Rational Application Developer (RAD) 8.x from Rational Software Architect (RSA) 7.5.x succeeds, but a blocking dialog showing a property error and requiring additional user intervention appears. Fix: The property is now being fetched when needed resulting in no blocking error dialog. PM73326: Problem: In the ClearTeam Explorer, an RSA model comparison results in an incorrect diff. Fix: The ClearTeam Explorer has been corrected to order contributors according to Eclipse ordering, not the standard ClearCase ordering. This results in a correct RSA model diff, but the contributors are now displayed in a different order from ClearCase. PM73432: Problem: When using the ClearCase UCM/ClearQuest Web Integration, and a record is transitioned to the complete state by the integration, the order of operations is incorrect. Fix: The order of operations has been corrected when the ClearCase UCM/ClearQuest Web Integration does an automatic transition to complete on a record. PM73586: Problem: The ClearTeam Explorer slows down an Eclipse build of an uncontrolled project on a machine that does not have MVFS installed Fix: Fixes a problem in ClearTeam Explorer (CTE) configuration validation which was adding overhead to certain operations in CTE. PM74114: Problem: RCleartool (a.k.a. ClearCase Remote Client (CCRC) CLI) takes more time to complete a load of many directories and files than the ClearTeam Explorer GUI. Fix: RCleartool has been fixed to have equal or better performance than the ClearTeam Explorer GUI. PM74428: Problem: The Checkouts view in ClearTeam Explorer (CTE) may show files as checked in if they are checked out outside of CTE. Fix: Fixes a problem in ClearCase where the Checkouts view in CTE may show files as checked in if they are checked out outside of CTE. PM75101: Problem: In the ClearTeam Explorer, users cannot deliver in a stream context when the stream has web views attached Fix: Fixes a problem in ClearTeam Explorer where users cannot deliver in a stream context when the stream has web views attached PM75238: Problem: The IBM Installation Manager panel was not displaying the correct version for non-English operating systems. Fix: The code has been updated to have a generic extraction of the correct version on the IBM Installation Manager panel independently of the Operating System language. PM75306: Problem: The Rational ClearTeam Explorer (CTE) is unable to delete a IBM Rational Software Architect RealTime Edition (RSARTE) model fragment during an 'absorb' operation if the fragment has checkouts on any branch. Fix: Fixes a problem in ClearCase where The Rational ClearTeam Explorer (CTE) is unable to delete a IBM Rational Software Architect RealTime Edition (RSARTE) model fragment during an 'absorb' operation if the fragment has checkouts on any branch. PM75703: Problem: Creating a stream from CLI or ClearTeam Explorer may result in unexpected errors with VOBs at feature level 7. Fix: Fixes issues surrounding creation of streams in feature level 7 VOBs. PM76372: Problem: When a user intends to start ClearCase Mainframe Connectors (CCz) rccTSOServer in certain environment where more than one network adapters are configured, sometimes addrinfo list returned from  getaddrinfo() has IPV4 address as the first one and IPV6 address as the second one. But if IPV6 network is not configured well, socket creation will fail using IPV6 protocol and exception is thrown. And then rccTSOServer has no opportunity to create socket using IPV4 protocol. So at last rccTSOServer will fail to start. Fix: When socket creation  fails using IPV6 protocol, CCz rccTSOServer does not throw exception any more while we return with an error code and then  rccTSOServer can have an opportunity to create socket using IPV4 protocol.