INC-135523 · Issue 590437
Tab panel read meaningfully by screen readers
Resolved in Pega Version 8.3.5
When the tabPanel received focus with accessibility, the entire tab content was read instead of the W3 standard of only reading "Edit ". This has been resolved by ensuring the Tabpanel has the proper aria-labelledby.
INC-139023 · Issue 591165
Tab will exit portal side menu
Resolved in Pega Version 8.3.5
It is expected that the Tab key will take users out of the portal menu and to the next focusable element, but on the side navigation menu both tab and arrow keys moved to the next menu item. This caused users to have to navigate the entire menu. The correct behavior was seen for the Navigation rule when called from an Action (for example on the button). This was traced to the system using tabindex as "0" for every element in the menu, and the tabs code has been updated to exit out of the menu bar when pressed.
INC-138354 · Issue 594356
Third party cookies reminder added for Mashup in incognito browsers
Resolved in Pega Version 8.3.5
When working in incognito/private browser windows, attempting to use Mashup displayed a message indicating login credentials were not recognized ("Unknown password and/or username"). This was due to recent changes originating with the browsers regarding blocking third-party cookies which are necessary for Mashup to run. For informational purposes, a new API has been added which will check for the PegaRules cookie; if the cookie is not present, the message "Third party cookies must be enabled for Mashup to function." will be displayed. This message may be customized by configuring the setting "pega.web.config.xCookiesDisabled" with the required message.
INC-133951 · Issue 584148
Try catch added for templated refresh when
Resolved in Pega Version 8.3.5
After configuring a refresh when condition on a dynamic layout triggered by changes on D_worklist.pxResults, a component stack mismatch error was generated if client side UI template is enabled. Without templatization it worked as expected. This has been resolved by adding a try catch block to DynamicLayoutGenerator.java.
SR-D16427 · Issue 497220
Multi-nodes rebuild LibraryMetadata to ensure all Rule-Utility-Functions are present on change
Resolved in Pega Version 8.3.1
When performing a complete Application import into a clean installation, references to certain Rule-Utility-Functions went unresolved during the initial assembly. Investigation showed that after introducing a new Rule-Utility-Library or Rule-Utility-Function on one node in a cluster and then generating that, the other nodes in the cluster did not have the correct LibraryMetaDataCache for that Rule-Utility-Library. Therefore assemblies on those other nodes could be bad and throw a runtime UnresolvedAssemblyError. This has been resolved by modifying the way the Library subsystem processes the node changes events for Library Generation to ensure that each node completely rebuilds the LibraryMetadata for that Rule-Utility-Library so it contains all the Rule-Utility-Functions.
SR-D32441 · Issue 502572
Multi-nodes rebuild LibraryMetadata to ensure all Rule-Utility-Functions are present on change
Resolved in Pega Version 8.3.1
When performing a complete Application import into a clean installation, references to certain Rule-Utility-Functions went unresolved during the initial assembly. Investigation showed that after introducing a new Rule-Utility-Library or Rule-Utility-Function on one node in a cluster and then generating that, the other nodes in the cluster did not have the correct LibraryMetaDataCache for that Rule-Utility-Library. Therefore assemblies on those other nodes could be bad and throw a runtime UnresolvedAssemblyError. This has been resolved by modifying the way the Library subsystem processes the node changes events for Library Generation to ensure that each node completely rebuilds the LibraryMetadata for that Rule-Utility-Library so it contains all the Rule-Utility-Functions.
SR-D22113 · Issue 498312
Multi-nodes rebuild LibraryMetadata to ensure all Rule-Utility-Functions are present on change
Resolved in Pega Version 8.3.1
When performing a complete Application import into a clean installation, references to certain Rule-Utility-Functions went unresolved during the initial assembly. Investigation showed that after introducing a new Rule-Utility-Library or Rule-Utility-Function on one node in a cluster and then generating that, the other nodes in the cluster did not have the correct LibraryMetaDataCache for that Rule-Utility-Library. Therefore assemblies on those other nodes could be bad and throw a runtime UnresolvedAssemblyError. This has been resolved by modifying the way the Library subsystem processes the node changes events for Library Generation to ensure that each node completely rebuilds the LibraryMetadata for that Rule-Utility-Library so it contains all the Rule-Utility-Functions.
SR-C49664 · Issue 435310
Updates made for usage validation utility Run report checkbox and buttons
Resolved in Pega Version 8.1.5
The 'Review mode only' checkbox wasn't working for any browser, and the 'Run Report' button wasn't working in IE. This has been resolved by updating the javascript used to invoke the Run Report button to be better integrated with IE, and by adding additional handling to set the ID for the checkbox.
SR-C74677 · Issue 418339
ADM pulse timing changed to inprove performance on very large clusters
Resolved in Pega Version 8.1.5
ADM uses a pulse that checks if any models need to be updated. This uses a Cassandra query which has no "where" clause, causing poor performance for very large clusters. For immediate relief of this issue, the pulse configuration has been updated to increase the timing from 30 to 300 seconds so it fires less often. Model updates typically happen every x hours and often the pulse will not detect any models that would need to be updated. Further improvements in this process will be available in a future release.
SR-C77730 · Issue 437101
Resolved case lock release improved
Resolved in Pega Version 8.1.5
Cases created through the create work object were unnecessarily holding onto locks for up to two hours. This was traced to a defer-write scenario for mobile devices which did not return the correct status needed for "no further assignments". In order to resolve this issue, a new parameter "returnNextActionInfo" of type boolean has been added to svcPerformFlowAction. This parameter will be always set as false when calling this activity from sync up so all locks will be closed and released.