28 May, 2009

2 commits


22 May, 2009

1 commit

  • Found a race condition affecting WRQ. Apparently, notice of update is sent before screen copy is updated. But if I wait until keyboard is unlocked, screen update has been performed.
    
    git-svn-id: https://svn.forgerock.org/openicf/trunk@4741 05b3e5af-d696-470f-a577-fd7599f74d3c
    george_hetrick
     

21 May, 2009

3 commits


19 May, 2009

1 commit


15 May, 2009

1 commit


12 May, 2009

1 commit


11 May, 2009

1 commit


07 May, 2009

3 commits


06 May, 2009

3 commits


05 May, 2009

3 commits


04 May, 2009

2 commits


01 May, 2009

2 commits


30 Apr, 2009

2 commits


22 Apr, 2009

1 commit


21 Apr, 2009

1 commit


17 Apr, 2009

1 commit

  • This allows the reporting of OWNER (for acount, group, and connection) to not require determining the object class.
    
    LdapUtil needs to be updated to add in the Ldap portion to Uids when makiing Ldap calls, and may need to determine object class of OWNER before doing a set, but I can't determine this until I have a running Ldap system.
    
    git-svn-id: https://svn.forgerock.org/openicf/trunk@4562 05b3e5af-d696-470f-a577-fd7599f74d3c
    george_hetrick
     

16 Apr, 2009

1 commit

  • . improve handling of NULL values
      . be more consistent in handling of Uids
    Speed up delete (use the fact that Expect4J can recognize error conditions
    
    I've found that to construct the Uid for the OWNER attribute, I need to issue a SEARCH command, since OWNER can be either an ACCOUNT or a RacfGroup. Because of this, I need to revisit the use of Ldap-style Uids.
    
    
    
    git-svn-id: https://svn.forgerock.org/openicf/trunk@4561 05b3e5af-d696-470f-a577-fd7599f74d3c
    george_hetrick
     

13 Apr, 2009

7 commits


10 Apr, 2009

4 commits