6.0.13

6.0 service packs

6.0.13

6.0.13 has been built on February 13, 2023. Overview of this service pack:

Supported databases	Microsoft SQL Server 2012 SP4 is no longer supported. Minimum supported version of SQL Server is now 2014. PostgreSQL 10 is no longer supported. Minimum supported version of PostgreSQL is now 11. See Supported databases for more information.
End of support of the browser "Internet Explorer"	Internet Explorer will no longer be supported. EBX® will still be accessible with IE but no issue specific to IE will be fixed in the future.
Backward compatibility – Rolling log files	Since 6.0.6, the rolling files names have changed from xxxx.log.yyyy-MM-dd (or xxxx.log.yyyy-MM-dd-i) to xxxx.MM-dd-yyyy.log (or xxxx.MM-dd-yyyy-i.log). By default, the former format (xxxx.log.yyyy-MM-dd or xxxx.log.yyyy-MM-dd-i) is used. The rolling format can be specified in ebx.properties file by setting the property ebx.log4j.appender.ebxFile.rollingFormat. If this property is specified, the filenames will have the following format: xxxxx.rollingFormat.log and will not be append to the end of the file name with the extension. To configure same file name format than since 6.0.6, the value must be %d{MM-dd-yyyy}-%i. The property will accept both a date/time pattern compatible with Java SimpleDateFormat and/or a %i which represents an integer counter. The integer counter is used in conjunction of the backup threshold, and will increment to create new files when the current file size will reach the threshold.
Closed issues	See 6.0.13 Closed issues section.

6.0.12

6.0.12 has been built on January 3, 2023. Overview of this service pack:

Records deduplication	As a side effects of some issues on indices, existing repositories might be impacted by duplicated records in a same table. A new service is available to fix records duplicates. The property ebx.persistence.boot.checkRecordDuplicates = true must be set in the EBX® main configuration file (ebx.properties) to launch this service. When set to true the records deduplication service is launched when starting the EBX® repository. During the execution of this service the repository cannot be accessed through the UI, REST and SOAP requests. Once the service has been ran the server is automatically shut down. The server must be restarted and the property ebx.persistence.boot.checkRecordDuplicates must be removed or set to false to make the repository available. The records deduplication service is activated with Container Edition using EBX_DB_BOOT_CHECK_RECORD_DUPLICATES environment variable. Value is false by default.
Health REST built-in	New health category operation helps to monitor the EBX® server. See Health operations for more information.
Java API – Repository	The new method Repository.getDefaultWhenFullyInitialized() blocks the calling thread until the repository is fully initialized and become able to handle: SOAP, REST Built-in, REST Toolkit requests and UI. See Repository.getDefaultWhenFullyInitialized for more information.
Closed issues	See 6.0.12 Closed issues section.

6.0.11

6.0.11 has been built on October 28, 2022. Overview of this service pack:

Disable logging the content of validation reports	To specify whether the content of validation reports is logged when validating a dataset or a table, use the new property ebx.validation.report.logContent, from the EBX® main configuration file (ebx.properties). By default this property is set to true.
Closed issues	See 6.0.11 Closed issues section.

6.0.10

6.0.10 has been built on September 29, 2022. Overview of this service pack:

Confidentiality	It is now possible to define the confidentiality of a data model element. Confidentiality changes the level of permission check performed when running queries. For more information, see Defining confidentiality.
Search strategy on foreign keys	It is now possible to change the search strategy of a foreign key field. This allows defining a Code search strategy that performs a 'starts-with' search, instead of a 'contains' by default. This is recommended for large volumes of data, especially to ensure good performance when searching or matching. This behavior impacts foreign key combo-box search (in a form and in advanced search), quick search, and matching. See Search strategy for foreign key fields for more information.
Permission to launch a workflow	It is now possible to delegate permission evaluation for launching workflows by API. For this a new property and a new method are available. For more information, see Permission to launch.
RequestResult Java API	The Javadoc of RequestResult.isEmpty() has been clarified; RequestResult.refresh() and RequestResult.refreshIfNeeded() have been deprecated. For more information, see RequestResult.
EBX® Container Edition	A new batch mode is available for the installer. See also Running the batch installer
Closed issues	See 6.0.10 Closed issues section.

6.0.9

6.0.9 has been built on August 22, 2022.

6.0.8

6.0.8 has been built on July 27, 2022. Overview of this service pack:

Workflow publication	To improve performance, workflow publication is now divided in two services that correspond to two different cases: When publishing an edited workflow model (legacy 'Publish' button), only this workflow model is published. The new service 'Mass publish' offers a dedicated step to select the workflow models to be published. See Publishing workflow models for more information.
User service without display and called from an association	A user service without display (in other words, a service that closes itself during its initialization, is now supported if called from an association. For more information, see UserServiceExtended.initialize and UserServiceNext.nextClose.
Backward compatibility – XPath predicates with parameters	XPath predicates using parameters and String literals containing a dollar sign and a parameter's name have been treated incorrectly in previous versions. Example for such a predicate: ./field1=$id and ./field2='$id'. Starting from 6.0.8, the first $id will be treated as an XPath parameter, while the second '$id' will be treated as a String literal, in accordance with the supported XPath syntax. Additionally, a predicate containing a parameter whose name starts with an underscore (e.g. $_param) was incorrectly parsed as a literal (i.e., '$_param'). Starting from 6.0.8, it will be treated as an XPath parameter; As a consequence, requests that incorrectly relied on the previous behavior will throw a "java.lang.IllegalStateException: The parameter $_param has not been declared".
REST access to history table	To improve the performance of the REST select operation on a history table, the new parameter includeMergeInfo can be used. See REST access to history table for more information.
Backward compatibility – Code generated from OpenAPI	For the built-in RESTful services, OpenAPI schema names have been changed to be more compatible with code generators: Snake case has been replaced by Camel case and some schema prefixes switch to suffixes. Therefore, entity names for newly generated code will change accordingly. See OpenAPI operations for more information.
REST Toolkit	In development run mode: Anonymous access to OpenAPI endpoints is now granted to any user. See OpenAPI permissions for more information. The Swagger UI is now available from the Administration area, to visualize the OpenAPI document. For more information, see Visual documentation.
URL for built-in schemas	The URL for common built-in EBX® XML Schema documents now uses a secured access (https://schema.orchestranetworks.com/ instead of http://schema.orchestranetworks.com/).
Closed issues	See 6.0.8 Closed issues section.

Notifications for upcoming releases:

• Oracle Database 12c will come out of support in an upcoming release. We advise to upgrade the database to a recent supported version.

• Microsoft SQL Server 2012 SP4 will come out of support in an upcoming release. We advise to upgrade the database to a recent supported version.

• JRE 8 will come out of support in an upcoming release. A recent LTS version will also offer global performance improvements.

6.0.7

6.0.7 has been built on June 24, 2022. Overview of this service pack:

Improved search on fields with 'Code' strategy	When searching on fields having the 'Code' search strategy, a 'contains' is now performed by default, instead of a 'starts-with'. Foreign key combo-box search (in a form and in advanced search) and the quick search are impacted. This change generally makes searching more user-friendly but it could negatively affect performance on large volumes. This can be configured in the data model by using a new parameter for 'Code' strategy. For large volume tables, restoring the previous behavior should be considered.
New audit trail	A new 'audit' logging category is available, to log user activity. In this release, it is limited to views access and some user services launch. It includes information on the search and filters applied. See Audit trail for more information.
Mass update optimization	Mass update performance has been optimized by limiting the number of intermediate processing such as bloc accesses, record serializations, and bloc compressions.
Query engine isolation improvement	The isolation mechanisms inside the query engine have been improved, now when a query is executed and its result is iterated, it is guaranteed that all internal persistence and indexing entities used in this process will utilize the same snapshot. For more information, see Query.getResult.
EBX® Container Edition	New Support Policy for Kubernetes. Default Java memory configuration has been decreased to 40% of the container to comply with Memory management. See also Memory configuration of the container.
Language packs	Language packs are available on eDelivery, since 6.0.4. With each release, there is a certain delay between the GA and the availability of the corresponding pack; in this case, the previous pack can be deployed (but may not cover all messages). See Language packs software for more information.
Closed issues	See 6.0.7 Closed issues section.

6.0.6

6.0.6 has been built on April 29, 2022. Overview of this service pack:

Filters preserved on view change	When the end user changes the view applied on a table, the local filter is now preserved (in previous versions the local filter was reset).
Accessibility	An accessibility mode has been added, see the dedicated section for more information.
User Service to create a record	In order to improve User Service integration, a new interface has been introduced: UserServiceForCreate. It is now recommended that implementations of UserServiceRecordFormFactory.newUserServiceForCreate return an instance of UserServiceForCreate rather than UserService.
Last modification date of a table	The new method AdaptationTable.getLastModificationDate returns the date when any record creation, modification and deletion last happened for this table.
Data services pagination	In case no page size or 0 are specified in a Data Service select operation, the pagination policy has been adapted regarding the new behavior that was introduced in 6.0.0. This simplifies the previously mentioned backward compatibility issues for SOAP and REST clients, by specifying default and maximum page sizes. See Data services pagination for more information.
Asynchronous indexing	Depending on the system load and the computing power available, indexing can now be performed concurrently. The log lets you know when this feature is enabled. For more information, see CPU.
Backward compatibility – Log configuration	Properties ebx.log4j.appender.xxxxx only support com.onwbp.org.apache.log4j.ConsoleAppender and com.onwbp.org.apache.log4j.net.SMTPAppender. Other types of appenders are not supported (they were not documented in previous versions). Renderers are not supported. Properties ebx.log4j.appender.xxxxx.layout support only com.onwbp.org.apache.log4j.PatternLayout Property ebx.log4j.disableOverride is no longer supported.
Toolbar configuration and workflow launcher	In the Data Model Assistant, if you edit a toolbar configured in a previous release, you may have to remove the service 'Workflow Launcher' from the 'Excluded services' list, before submit ('Workflow Launcher' is always excluded from the toolbar, so it is no longer offered as an option).
H2 Database Engine upgraded	The H2 Database Engine included in this release has been upgraded to version 2.0. This new version does not support an automatic upgrade of existing repositories from an older H2 version. These must be manually updated through the dedicated procedure.
Closed issues	See 6.0.6 Closed issues section.

6.0.5

6.0.5 has been built on February 18, 2022. Overview of this service pack:

Performance and stability	Performance has been improved to handle the case when many small transactions occur (the problem was a noticeable latency correlated with the latency of the disk storing Lucene indexes, and independent of the size of the tables updated by the transactions; for instance, this issue could be observed when starting a workflow). Memory management and stability related to the indexes have been improved.
Memory allocation	The memory allocation requirements, for both the application server and the operating system, have been clarified. For more information, see Memory management.
Coding recommendations for transactions	For a better performance of transactions (Java implementations of Procedure and also of TableTrigger), any unnecessary index refresh should be avoided. For more information, see Unnecessary index refresh.
Java 17	The Java Runtime Environment 17 LTS is now supported. For more information, see Supported application servers.
Java 8	JRE 8 will come out of support in an upcoming release. We advise to upgrade the runtime to a recent LTS version, to take advantage of the performance improvements it offers.
Backward compatibility – Permission checks in hierarchy	As a consequence of fixing CP-22261, hierarchy views relying on hidden fields will now fail with a QueryPermissionException, as expected. See Improved permission checks when executing queries for more information.
Backward compatibility – Workflow properties	The values of the workflow properties in the EBX® configuration file are now constrained as follows: ebx.manager.workflow.legacy.userTaskMode must be true or false (case-insensitive). ebx.manager.workflow.hierarchyPlanView.hidden must be true or false (case-insensitive). ebx.workflow.activation must be true or false (case-insensitive). Any other value throws a fatal exception at server launch.
Closed issues	See 6.0.5 Closed issues section.

6.0.4

6.0.4 has been built on December 22, 2021. Overview of this service pack:

Closed issues

See 6.0.4 Closed issues section.

6.0.3

6.0.3 has been built on November 9, 2021. Overview of this service pack:

EBX® Container Edition	This service pack includes a utility to build an image of TIBCO EBX® Container Edition. This image is Linux-based and can optionally include addons. For more information, see Building the image, Running the image and Customizing the image.
Supported databases	Oracle 18c is no longer supported. PostgreSQL 9.6 is no longer supported; the minimum required version for PostgreSQL has been upgraded to 10. See Supported databases for more information.
Improved the workflow diagram	The PDF export wizard has been enhanced. It is now possible to display the step identifiers of the PDF on demand, and to add a step index at the end of the generated file. See Actions for more information.
URLs configuration	Two new properties have been added to the ebx.properties file to force the use of the HTTPS protocol and its related properties for building URIs. In the user interface, the 'URL Policy' configuration in the Administration area becomes deprecated. See URLs computing and URL policy (deprecated) for more information.
Backward compatibility – Workflow activation property	As a consequence of fixing CP-22130, the value of the property ebx.workflow.activation (in EBX® configuration file) must be true or false (case-insensitive). Any other value throws a fatal exception at server launch.
Closed issues	See 6.0.3 Closed issues section.

6.0.2

6.0.2 has been built on September 15, 2021. Overview of this service pack:

Supported databases	A new database is supported: Google Cloud SQL for PostgreSQL, version 9.6 or higher.
Improvement of the workflow diagram	The boxes representing the steps have been optimized, allowing more text to be displayed. This avoids truncation in the step names. Note: for this evolution, the step boxes have been enlarged. This may somewhat degrade the display of existing or imported diagrams: the links may overlap on the step boxes. This can be easily corrected in the workflow modeling: just by selecting the step boxes and moving them a bit, so that the display fits well, then saving the layout. For larger diagrams, it is recommended to use the 'Default Layout' feature in the 'View' section (but in this case, the old disposition will be lost). Note that it will be necessary to publish the models, for the display to be updated in the Data Workflows section. A new action is available in the View menu: 'Show/Hide steps identifiers'. This new action allows displaying the step identifiers, on demand, directly in the workflow diagram (hidden by default). The generation of the pdf will take into account the choice made prior to displaying/hiding these identifiers. See View for more information.
Backward compatibility – Check on default constraint on primary keys of type xs:string	As a consequence of fixing CP-21856, the check of the default constraint defined on primary keys of type xs:string is now stricter. The default constraint that forbids non-collapsed whitespace is now blocking for all operations. It is still possible to override this default constraint, by declaring a specific pattern that allows non-collapsed whitespace on primary keys of type string. See Whitespace handling for primary keys of type string for more information.
Closed issues	See 6.0.2 Closed issues section.

6.0.1

6.0.1 has been built on June 25, 2021. Overview of this service pack:

Performance	Many features offer better performance. In particular: View access (CP-21576) Selection in a view Filtering lists and fields in groups Search on a table validation state whose validation report contains many items Queries involving computed fields (the computation during queries is deferred until the value is needed; if a computed field is actually not needed for the execution of the query, its value is not computed) Validating a table reference that has no defined table reference filter.
Disk latency	The documentation has been updated to indicate a low disk latency as a requirement for a better overall performance.
Breadcrumb in history view	In the breadcrumb trail, steps related to the history are now clearly identified.
Workflow editor	The button for step creation also appears on 'mouse over', making this feature more intuitive. The DEL shortcut is now functional, and allows deleting a step.
Workflow API	The work item key is now directly accessible from the Session interaction API. See SessionInteraction.getWorkItemKey for more information. A method is added to CreationWorkItemSpec, to create a work item allocated to a specified user, with the possibility to reallocate it to a collection of profiles. See CreationWorkItemSpec.forAllocationWithPossibleReallocation for more information.
Toolbar API	A new method in the 'Toolbar' API allows defining whether the current button should be highlighted by default. See ToolbarButton.setHighlighted for more information.
Pattern	The validation of patterns on string is now greedy, and no longer requires extra parentheses with the | operator. For example, the pattern [0-9A-Z]{8}|[0-9A-Z]{11} will validate strings of 8 upper characters, without requiring extra parentheses around the pattern.
D3 broadcast	The D3 broadcast Destination section has been added to the documentation.
Browser support	Microsoft Internet Explorer 10 is no longer supported.
Closed issues	See 6.0.1 Closed issues section.

Note: 6.0.0 has been built on March 17, 2021.

Overview of 6.0

New architecture

This release introduces a new architecture for the core engine, with numerous modifications affecting the way data is stored and queried. Our custom in-memory indexes have been replaced by persistent ones, based on Apache Lucene. The data representation and caches have been redesigned from scratch. The in-house query execution framework has been modernized with the introduction of Apache Calcite, which features a powerful query optimizer. This new architecture provides the following improvements:

• Supports larger volumes, by offloading the bulk of the caching duties to the filesystem cache of the OS.

• Improves search capabilities through text/fuzzy search.

• Enables support for arbitrary SQL within 'SELECT' queries.

• Improves availability by reducing the startup time of the application server.

The new architecture also improves transaction management: more concurrency and a better isolation of data. This release bears some limitations, described in the section Limitations for this release.

User interface

• Search and sort improvements: The new 'Quick Search' component and the 'Advanced Search' pane provide a better, more robust search experience. 'Find in hierarchy' has been moved from the filter pane. Sort by relevancy is available with quick search, and you can also now sort on hierarchical views. Text sort now better follows language alphabetical order.

• More customizability: You can now use the web user interface to create customized forms with a graphical editor. You can now launch a workflow from any data context.

• Workflow editor: You can now completely edit a workflow model from the workflow diagram editor.

• Browsing support: The Microsoft Edge versions based on Chromium are now supported.

User experience

Quick Search

A new Quick Search component is available in the tabular and hierarchical views, at the center of the toolbar. Quick Search replaces the 'Text search' pane that was formerly available in the filters pane, offering a more user-friendly and more efficient search. Quick Search relies on a configurable search strategy, defined in the Search extension of the Data Model Assistant. Quick Search also offers Lucene query operators, such as wildcards, fuzzy search, or logical operators (NOT,...).

See Quick Search for more information.

Advanced Search

A new 'Advanced Search' pane is available at the top of the workspace area, for tabular and hierarchical views. This pane is hidden by default. It can be expanded and collapsed with a new dedicated button, which is located to the right of the Quick Search. The Advanced Search pane replaces the 'Search' pane that was formerly available in the filters pane, offering a more user-friendly and efficient search.

Find in hierarchy

On the hierarchy view, the 'Find in hierarchy' feature has been moved from the filter pane on the right to the bottom bar of the workspace area. This bar is hidden by default. You can expand or collapse the view with a new dedicated button at the bottom left of the workspace area.

Sort by relevancy

When a text search is applied on a tabular view, by default, this view is sorted by score, with the most relevant records displayed first. Sort by relevancy is available for the results of a quick search, potentially in combination with more predicates (from advanced search). You can disable this type of sort and restore the previous sort criteria before the search. Note that sort by relevancy is supported on inherited datasets, history or mapped tables.

Text sort

When you sort on a string column with a Name or Text search strategy, the sort now better follows the usual language alphabetical order. For instance, the letter 'é' is after 'e' and before 'f'.

Request cancellation

When loading a page in a tabular view takes longer than expected, a 'Cancel' button is displayed.

Clicking this button cancels the request and prompts you to define more selective filters or more simple sort criteria.

Breadcrumb

The breadcrumb has been improved to better support many levels, and to facilitate the navigation in these levels.

Maximum three elements in a breadcrumb	The number of displayed elements is reduced to three to provide faster access to the breadcrumb items.
Menu button	All non-displayed items are available in a drop-down menu.
Close button	A new 'Close' button closes the last item of the breadcrumb (the current page).
Relationship labels	When the record is opened from a foreign key or an association table, the table label is replaced by the relationship label.
Display of optional information	The display of a table or relationship label preceding the record label is now optional, and can be configured directly on the node in the Data Model Assistant.

New toolbar for view menu

The 'Edit', 'Duplicate', 'Delete', 'Define as favorite' actions are now accessible on the first level of the view menu. The toolbar containing these actions is displayed on mouseover of any view menu item.

Built-in service 'Edit link record'

The built-in user service "Edit link record" was available only from the record form. You can now open the link record directly from the hierarchy view. In the hierarchy tree, this service is now available directly on the concerned hierarchy node.

Dataspace management new features

Dataspace tree results are now paginated to handle larger volumes, and a search bar has been added to help you quickly filter dataspaces. The search query is performed on the key and the localized label, and as you finish typing, the search results are refined seamlessly.

In the administration pane, the default view of the dataspaces table has been changed to a hierarchical view, which better matches the logical representation of dataspaces.

Message box

The message box is restored to the state of versions prior to 5.9. It displays only messages related to the current page. Every time a new page is displayed or closed, the content of the message box is automatically updated. Hence, old and new messages do not build up in the message box, and the same message is displayed only once.

User interface customization

Custom layout editor

From the Web user interface, you can now create customized forms with a graphical editor. No Java programming is required. This feature is available as an extension in the Data Model Assistant.

See Custom forms for more information.

Workflow Launcher

To launch any workflow with any data contexts, you can use the new user service. Workflows are henceforth available directly on the toolbar of tables, records, or hierarchies.

To display a workflow on the toolbar, first create and configure a workflow launcher from the "Workflow launchers" section. This section is available in the "Administration/workflow management" area.

This new feature offers several advantages, including improving the user experience by saving the user from shifting attention between the data view and the workflow view. It also supports launching the same workflow from any data context.

See UI – Workflow launcher for more information.

Sort on hierarchical views

You can now sort every level of a hierarchy view. In the configuration of the hierarchy view, the previous 'Ordering field' option has been replaced by the new 'Sort strategy' setting. This setting allows choosing between the strategies 'Default', 'Sort by columns', or 'Sort by ordering field'.

See Viewing data for more information.

Ergonomics and layout new parameters

In "Administration > [any perspective] > Ergonomics and layout", three new parameters have been added.

See Ergonomics and layout for more information.

Maximum length of text areas	The maximum number of characters allowed for text area fields in forms. You can disable this parameter by entering a negative number.
HTML editors spellchecking	Spellchecking is provided by the browser spellchecker. After spellchecking is activated, the browser user interface dictionary is used by default. This setting is customizable if many dictionaries are installed. If HTML input is read-only, spellcheck is disabled.
Default display density	You can set the default density for all users. The default display density is 'Compact'.

Association tables

The Data Model Assistant now contains options to hide toolbars of association tables. On association nodes, the options are located within the setting group "Rendering in forms". This new feature is also available programmatically from the Java class UITable.

Form node label

To simulate whether a form node is mandatory, you can now add or remove a 'mandatory' indicator (*) next to the label of the form node. Use the JavaScriptCatalog.ebx_form_setMandatoryIndicator in JavaScript or UIFormWriter.addJS_setNodeMandatoryIndicator in Java.

Built-in service 'Profile settings'

To modify your profile, you can use the built-in service 'Profile settings'. This built-in service is available with the default directory. It is deactivated when EBX® is configured to use a custom directory. If a custom directory extends the default one, the user service can be reactivated by setting ebx.myProfile.enabled to true in the EBX® main configuration file. The built-in service 'Profile settings' key is ServiceKey.MY_PROFILE.

New Java API related to perspectives

The new package Perspective provides classes for retrieving all relevant information about the preferences and menu for the perspectives.

To get the current perspective from a context, use the new method getCurrentPerspective. This method has been added to the following classes :

• UIFormContext. See UIFormContext.getCurrentPerspective

• UILabelRendererContext. See UILabelRendererContext.getCurrentPerspective

• UserServiceRequest. See UserServiceRequest.getCurrentPerspective

• WidgetDisplayContext. See WidgetDisplayContext.getCurrentPerspective

New method in ToolbarButton API

The new method ToolbarButton.setHighlighted has been added to the ToolbarButton API. It allows defining whether the current button should be highlighted by default.

New scripted features in Data Modeling

Scripted computed fields (limited to the current record)

To specify fields that are computed at runtime, you can use the new scripting language. In this release, the computation can depend only on the data of the current record.

Compared to a "value function" written in Java, the advantages include the following:

• Scripting is less error-prone, due to tighter integration with EBX®.

• Indexing of the computed value allows efficient search and sort on the field.

• Does not require advanced programming skills.

• Does not require redeploying EBX®.

A rule editor, with syntax highlighting and contextual completion, is available in the Data Model Assistant.

See Function field for more information.

Scripted permission rules

To specify record-level access rules on a given table, use the new scripting language. Any permission rule must return the access ("read-write", "read-only", or "hidden") for the current context: the current record, the current user session (including user roles), and so on.

Compared to "access rules" written in Java, an advantage is that scripting significantly improves performance, especially for large volumes of data, by leveraging the new EBX® SQL engine.

A rule editor, with syntax highlighting and contextual completion, is available as an extension in the Data Model Assistant.

See Record permission for more information.

Data modeling

Indexing and search strategy

When you edit a field in the Data Model Assistant, you can now define its indexing. This new option is available under the tab 'Extensions', 'Search' section. In particular for string fields, this new option supports defining how the Quick Search is performed.

Also, to implement advanced and efficient search strategies, use a new service to define custom indexing. This new service is available in the left navigation pane, under 'Extensions > Search'.

See Search for more information.

Linked fields

You can now define a new kind of field that refers to another field in the target table of a relationship. Use this new feature to simulate a multi-table view by aggregating the fields of a main table with some fields of another table.

See Linked fields for more information.

Uniqueness constraints on aggregated lists

You can now specify that a value must be unique in the scope of a simple list (fields that define maxOccurs > 1)).

See Uniqueness constraint for more information.

Associations by foreign key

You can now define associations 'by foreign key', based on an existing foreign key field that is multi-valued (defines maxOccurs > 1).

Data model introspection Java API

Facet on mandatory field	A new interface to retrieve the cardinality constraint specifying that a field is mandatory (minOccurs="1" and maxOccurs="1"). See SchemaFacetOnMandatoryField for more information.
Retrieving the node associated to a facet	A new method to retrieve the node that is associated with a facet. See SchemaFacet.getSchemaNode for more information.
Facet validation message	A new interface to retrieve the validation message defined by a facet. See SchemaFacet.getFacetValidationMessage and SchemaFacetValidationMessage for more information.
XPath filter on foreign key	A new method to retrieve the XPath filter defined on a foreign key constraint. See SchemaFacetTableRef.getXPathFilter for more information.
Default view properties	A new method to determine if a node is hidden in all views of a table. See SchemaNodeDefaultView.isHiddenInAllViews for more information.
Association consistency	A new method to determine if an association is consistent in the context of a specific dataset. See AssociationLink.isConsistent for more information.

Workflow

Workflow editor

The workflow modeling has been enhanced with a full graphical editor.

This editor is based on the graphical view already available in the previous version. This diagram has been enriched to support all of the creation, modification, and deletion operations. This editor provides more readability--in particular, in the case of redirects to existing steps.

All task actions are available from the diagram.

The [+] button, which inserts a task, is displayed on each selected link.

In addition to being more readable, the new editor offers additional features:

• Diagram export to a PDF document: this feature is useful for printing a workflow model.

• Overview panel: this panel can be collapsed, expanded, and dragged, providing a high level view.

  

• The data context wizard is also provided for custom library items (scripts, user tasks, and conditions).

• The declaration of a static sub-workflow has been improved in the user interface: both input and output parameters now have either a select list or data context wizard to help.

• Explicit actions to zoom in and out have been added in the toolbar of the diagram.

  

By default, the old hierarchical view is no longer offered. You can reactivate it in read-only mode by using a dedicated property. See Backward compatibility for more information.

For more information about the workflow editor, see Editing the workflow diagram.

Sub-workflow progress strategy

The progress strategy is useful for performing several tasks consecutively, without returning to the inbox. This feature is now available for the sub-workflows.

A new concept has been added to implement this feature for sub-workflows: if there are several sub-workflows, you must define which one is the "foreground sub-workflow". Only in this foreground sub-workflow can you exploit the progress strategy.

The user interface now includes a field for static sub-workflow to define the foreground sub-workflow.

The method SubWorkflowsCreationContext.setForegroundSubWorkflow has been added to set which sub-workflow should be considered in the foreground, for dynamic sub-workflow. And the method ProcessLauncher.isForegroundSubworkflow has been added to know if a launcher is associated to a sub-workflow in the foreground.

For more information, see Sub-workflow invocations.

Workflow diagram available in the Data Workflows section

In addition to the progress view, the workflow model diagram is now available in the Data Workflows section (inbox, Monitoring-Work items, Monitoring-Active workflows, and Completed workflows).

The progress view provides a view of the workflow execution history, whereas the diagram offers a more functional view of what the workflow does as a whole.

Ability to deactivate the validation report copy for the create snapshot script task.

You can now deactivate the validation report copy while creating a snapshot via built-in script task. An additional property is available in the script task definition.

Publication service updated according to the new Workflow launcher feature

The publication service has been updated. At the last step, when the selected workflow models have been published, a new tab is available to list the potentially affected workflow launchers.

Workflow new Java API

Workflow input parameter	A new feature to test whether a variable name is defined as an input parameter of the data context. See ProcessLauncher.isDefinedAsInputParameter.
Hidden step in progress view	A new method determines if a step is defined as hidden in the progress view. See ProcessInstanceStep.isHiddenInProgressView.

Permissions

Scripted permission rules

See Scripted permission rules section above.

Permissions cache extension

To improve performance, permissions caching has been enhanced, and now also handles the following permission resolutions:

• Low-level actions on a table (for example, create, delete, overwrite, and occult).

• All kinds of access rule results: AccessRule, AccessRuleForCreate, and scripted record access rules.

Data validation

Persistent validation reports

Validation reports are now persisted in the EBX® repository. That is, validation reports are no longer kept in memory, but are stored in technical datasets of dedicated dataspaces. Storing validation reports in an EBX® repository addresses memory-related issues when datasets have a large amount of validation messages. Also, validation reports are no longer lost when the application server restarts.

Declaring modifications on programmatic constraints

You can now indicate that the implementation of a programmatic constraint has been changed, by associating a version number to the implementation of a JavaBean. This version number is used to indicate that existing persisted validation reports must be updated by revalidating the constraints that have been modified. As a consequence, you must define a version number of each of the constraints defined in a data model to keep existing persisted validation reports consistent when the implementation of these constraints is modified.

See JavaBeanVersion for more information.

Validation message threshold

You can now specify a validation message threshold at the data model level. This new feature replaces the property ebx.validation.constraints.messages.threshold that is defined in the EBX® main configuration file (ebx.properties).

See Validation message threshold for more information.

Validating new records and updated values

You can now validate a record that has not yet been created, and you can validate the updates performed on a record or on dataset nodes that have not yet been committed.

See ValueContextForUpdate.getValidationReport for more information.

Retrieving the constraint related to a validation message

To retrieve the facet associated with a validation item, use the new method ValidationReportItem.getFacet. That is, this new method supports retrieving the facet that has added a validation message.

See ValidationReportItem.getFacet for more information.

Retrieving the blocking policy related to a validation message

To retrieve the blocking policy associated with a validation item, use the new method ValidationReportItem.getBlocksCommit.

See ValidationReportItem.getBlocksCommit for more information.

Closing a ValidationReportItemIterator

To release underlying resources, use the new method ValidationReportItemIterator.close.

See ValidationReportItemIterator.close for more information.

Changes regarding the refresh policy

Requesting a validation report with the refresh policy ValidationSpec.RefreshPolicy.AS_IS returns only the persisted validation report. The validation report is empty if the dataset, the table, or the record has not already been validated.

See ValidationSpec.RefreshPolicy for more information.

Limit validation messages in logs

To specify the maximum number of validation items to display in the logs, use the new property ebx.validation.report.maxItemDisplayedInLogs, from the EBX® main configuration file (ebx.properties). By default this property is set to 100. See Backward compatibility for more information.

Associations with explicitly defined dataspaces

When using an association defined with a foreign key or link table that is in an explicitly defined dataspace, the dataset will now be in severe error if the current dataspace does not match the association reference dataspace. See Association.

Data services

REST Toolkit APIs

Procedure

A new interface and built-in query parameter have been added to ease Procedure execution through automatic pre-configuration.

See ProcedureUtility and BuiltinQueryParam.COMMIT_THRESHOLD for more information.

URI builders for built-in

URI builders have been reworked to do the following:

• Add new methods for query parameters management.

• Enhance modular, generic, and intuitive developments.

• Comply with the addition of new REST categories.

• Support the new beta feature on dataspace or snapshot operations.

  See Dataspace operations for more information.

See URIBuilderForBuiltin and URI builders for more information.

Serialization

Built-in serializers and deserializers are available to handle table records.

Built-in deserializers

To automatically deserialize a table record, use a DTO's attribute of ContentHolder type and annotated with Table.

Built-in serializers

To automatically serialize a table record, use a DTO's attribute of ContentHolder type and optionally annotated with ExtendedOutput. A table record serialized through this mechanism gets the same JSON content structure as the ones from the Built-in RESTful services. However, a new compact format has been added to ease interaction for simple use cases. The absence of the ExtendedOutput annotation on the ContentHolder attribute implies the use of the compact format of the JSON content structure. For the opposite way, the presence of the ExtendedOutput annotation implies the use of the extended format of the JSON content structure. The ExtendedOutput annotation and Include enumeration determine the content of the generated JSON structure of the table record.

See Serialization of a table record for more information.

Built-in RESTful services

The built-in RESTful services have been improved on the following topics:

Data views operation	A new operation from the 'data' category is available to look up the published views of a table. See Look up table views operation for more information.
Data operations	Two new prepare operations with selector values are available to create and initialize a transient record for creation or duplication. Several new properties have been added to fields' metadata, namely: autoIncrement, association, checkNullInput, constraints, linkedField, readOnly, and enumerationDependencies. On select operations, the sort features have been improved, allowing the results to be ordered: by relevancy with the new sortByRelevancy parameter, by overriding the default priority of sort groups with the new sortPriority parameter. See Prepare operations, includeMetamodel, and Linked fields parameter for more information.
Form data operations	A new 'form-data' category is available. It is intended to be used to insert or update data on user form. This category validates the incoming data before performing the requested action and returns a report. See Form data operations for more information.
Compact format serialization	New categories with a 'compact' suffix are available for data and form data operations. They are very similar to the root ones but they are intended to be used to compact the JSON body content structure to ease interaction for simple use cases. All formats are based on the new REST Toolkit's Built-in serializers. See Operation category and REST Toolkit serialization for more information.
OpenAPI operations	A new 'open-api' category is available. It is intended to generate OpenAPI documents for table and schema nodes of the REST built-in 'data' category operations. See OpenAPI operations for more information.
Beta feature: Dataspace or snapshot operations	Several new operations from the 'data' category are available, as beta feature, to manage dataspace or snapshots. See Dataspace operations for more information.
Token operation	The built-in REST token authentication create operation has been improved: It can now be performed against the HTTP request. See Token authentication operations for more information.
Common parameters	A new 'ebx-channel' query parameter is introduced to apply visibility configuration, defined in the data model, according to a table, a field, or a search. The content also corresponds to the current view (tree view, tabular, custom, form, and so on). A new 'ebx-indent' query parameter is introduced to indent the response body and be more readable for a human. The 'indent' query parameter has been deprecated. See Session channel, ebx-channel and ebx-indent for more information.

Application server deployment

Data compression library

The LZ4 compression algorithm is now used internally by the EBX® engine. You must deploy the dedicated library. Also, the way you deploy this dedicated library has a global impact on performance. See Data compression library for more information.

Speed of server startup

To speed up the startup of the web applications server, configure the JAR files scanner.

XML Catalog API

The upgrade of the Xalan and Xerces packaged libraries introduces new dependencies toward classes provided by the Java runtime environment 9. You can add these classes without resorting to the Java runtime environment 11 LTS by installing the new bundled jar archive called xml-apis-1.4.01.jar. If a Java runtime environment 11 LTS, or above, or a WebLogic 14c application server is used to run the application, then no additional installation steps are required. See Installation notes for more information.

Core Java API

SQL queries

New APIs have been added that allow 'SELECT' queries to use arbitrary SQL, notably with joins over tables located in different dataspaces and datasets.

The main entry point is the Query interface. For more information on SQL queries, see the respective chapters in the Developer's guide.

Triggers on datasets

Triggers defined at the dataset level have been enhanced to perform operations:

• After duplicating a dataset.

• Before updating a field outside a table.

• After updating a field outside a table.

See InstanceTrigger for more information.

Request and query optimizations

Several features offer better performance. In particular:

• A request cache has been introduced that caches the transformation from requests to queries, when executing a Request. See Request cache activation for more information.

• RequestResult size related methods have a better performance if executed multiple times.

• Requests that are based on primary key values will be executed using a more optimized implementation of RequestResult.

Limitations for this release

Query limitations

The performance of these features is not yet optimized:

• Navigation on paginated views (next/previous page) with sort on foreign key fields having a custom display pattern, if this pattern contains fields outside the primary key of the referenced table.

• Advanced search filtering on foreign key fields using string-based operations (e.g. starts-with, contains, matches, etc.).

• Navigation on paginated views (next/previous page) with sort on linked fields.

The following cases do not rely on Lucene indices. This has two consequences: first, performance is not optimized; second, Quick search behaves like a 'contains', rather than a full-text search:

• computed fields with non-local dependency;

• inherited fields;

• all fields in an inherited dataset;

• all fields in table history views.

More specific cases not benefiting from index optimization are described in the following section.

Value-labeling limitations

Search and sort using the value-labeling feature have a few limitations described in section Value-labeling.

Hierarchical views limitations

• The container node used for grouping orphan nodes appears to be openable, even when the container node is empty. Detecting whether the container node is empty is currently disabled because it is not optimized.

• Hierarchy dimension based on dependency link field which is a list or under a list is not allowed.

• While building a hierarchy dimension, the same table (or same table path) cannot be used in more than one hierarchy dimension level.

• While building a hierarchy dimension, the same table cannot be used both as a target table and a link table.

• Selection in hierarchy is not cumulative: changing a user filter during a selection session may lead to unwanted results.

• Specific parameter: the property 'Remove recursive root leaves' and 'Display non-matching children' is not implemented.

Data validation limitations

• It is not possible to validate a dataset or a table within a Procedure.

Linked fields limitations

• Targeting a node which is an enumeration is not fully supported yet. Currently, only the raw value will be available for display and search, instead of the enumeration label. Until this limitation is removed, the workaround is to declare the same enumeration in the target node and in the linked field itself.

Version upgrade

In addition to the information provided in this release note, the chapter 6.0 Java API changes provides an exhaustive list of the new Java APIs published since the release of the 5.9 version.

Migrating an existing EBX® 5 environment

To migrate from an existing EBX® 5 environment, follow the dedicated migration guide.

Changes to supported environments

The chapter Supported environments has been updated to reflect the necessary upgrades detailed below.

Browser support	The Microsoft Edge versions based on Chromium are now supported. For Mozilla Firefox, the support of ESR 60 has shifted to ESR 68. For Microsoft Edge, the support of 42 has shifted to 44.
Oracle Database support	Minimum supported version of Oracle is now 12c (but excluding 18c).
PostgreSQL support	Minimum supported version of PostgreSQL is now 10.
Microsoft SQL Server support	Minimum supported version of SQL Server is now 2012 SP4.
Java web application server support	The Java web application server must support the JSON Processing 1.1 (JSR 374) or allow the use of that embedded in the ebx.jar library.
WebSphere Application Server support	WebSphere Application Servers which use a Java SDK below version 8.0.4.10 are incompatible and consequently not supported. It is highly recommended to use the latest Java SDK available and compatible with the application server. See Supported application servers for more information.

New transaction management

The improved implementation of transactions improves concurrency and data isolation.

Concurrency	A running Procedure no longer excludes multiple concurrent ReadOnlyProcedures. For more information see Concurrency.
Snapshot isolation	While the isolation level was READ_UNCOMMITTED by default, accessed data are now "frozen" on a committed transaction. For more information, see Query snapshot isolation.
Adaptations are now immutable	Instances of the Java interface Adaptation are now immutable and reflect the persisted state of the record when it was retrieved. They are never affected by subsequent updates on the record (and in case of dataset inheritance, they are never affected by an update on their parent, recursively). For more information see Adaptation objects. As a consequence, the method ProcedureContext.doModifyContent never modifies the Adaptation object passed as an argument. If access to the modified record or dataset is needed, then you must use the Adaptation object returned by the method. Note: In previous versions, the documentation of this method already stated this rule. However, neglecting to do so only had consequences in rare cases. In this version, the immutability is always preserved.

Removed features

Some features are no longer necessary in the new architecture.

Dataspace loading strategy	Formerly, you could define a custom loading strategy for a dataspace, namely: a 'forced loading' strategy for loading the tables and indexes as early as possible in the background at startup; and a 'forced loading and prevalidation' strategy for also pre-computing the validation reports. This new version suppresses the need to load the entire table and compute its indexes when accessing it for the first time. It also offers persistent validation reports. Consequently, all dataspaces are loaded on-demand.
Indexing	Formerly, you could define indexes on tables, aiming at speeding up requests. This new version makes obsolete the in-memory indexes previously used, as well as the need to define a custom indexing. All fields are indexed persistently.
License keys	EBX® 6.0 and its addons no longer require license keys. You can remove the ebx.license key from the ebx.properties main configuration file.

Backward compatibilities

Relational mode is no longer supported

The new architecture for persistence and the new query engine has made relational mode redundant and cumbersome to support. It is no longer supported. The primary justifications of relational mode were:

• Performance on large volumes: this is now provided by a more efficient persistence format, and by the Lucene search library.

• Using SQL from inside EBX®: this is now provided through the Query API.

• Offering data connectivity with third-party systems. We propose the following to replace access to a relational table: for read-only access, you can activate the replication feature in the data model. For read-write access, which was allowed in restricted conditions, you can use other integration features (such as Java API, data service, and Data Exchange).

On upgrade to version 6, the data existing in relational tables is automatically transferred into newly-created semantic tables, and then the relational tables are deleted from the database.

This conversion also means:

• Limitations due to the relational model no longer apply. For example, you can create snapshots and child dataspaces.

• The table can now be replicated.

• Previous structural constraints are no longer structural if the data model enables neither history nor replication.

• Blocking foreign key constraints are no longer blocking.

XPath and Java Request API

Due to the overhaul of the indexing and querying engine, some XPath functions are no longer supported, and sorting is impacted.

Numeric fields	Invoking the XPath functions contains, starts-with, ends-with, and matches on a numeric field is no longer supported. For example, contains(nCol, 123) now generates an error.
matches	Previously, functions of the type matches ignored a parsing error silently when malformed, and never matched a record; the error is now generated explicitly.
Sorting on text	It is not possible to sort according to a text field. This concerns osd:text fields by default, and fields having an explicit 'text' search strategy. You can use the 'name' strategy in cases where sorting is required. When sort is done on a field of type xs:string, it now better follows the usual language alphabetical order. For example, the letter 'é' is after 'e' and before 'f'.

The Java interface RequestResult has several changes.

Method getAdaptation(int)	The method RequestResult#getAdaptation(int anIndex), which had long been deprecated, is removed from the public API.
Closing request results	In previous versions, the documentation of RequestResult stated that it was mandatory to invoke its close method after invoking RequestResult.nextAdaptation(). However, neglecting to do so had no consequence for tables in semantic mode. In this version, the request result opens system resources that must be released after the request completes. Consequently, it is now crucial to abide by the recommended protocol (see below). Also note that the method has been slightly modified, and creating the request result (RequestResult.execute) now also opens some resources. RequestResult has been modified to implement the AutoCloseable interface. This change supports tracking incorrect uses in your development environment. The coding pattern should now be: try (RequestResult result = request.execute();) {...} .
Method isEmpty()	The documentation of this method has been clarified: invoking it has an inherent cost, it should not be called if the request result is then traversed.
Methods refresh() and refreshIfNeeded()	These methods are deprecated and do not do anything.

Restricted functionality on osd:text

In EBX® 5, osd:text differed from xs:string only by offering a default editor with several lines for input and display. In EBX® 6, the type osd:text is automatically assigned the 'Text' search strategy, and this assignment cannot be modified. As a consequence, this type:

• cannot benefit from sorting,

  cannot be used in the label of table records, or of foreign key constraints.

Therefore, osd:text should now only be used for very long texts. We advise to change the type of osd:text fields to xs:string if these fields are not intended to hold long strings. It is still possible to use an editor allowing several lines for input and display, by customizing the default built-in text box widget as follows:

<xs:element name="field" type="xs:string" minOccurs="0" maxOccurs="1">
   <xs:annotation>
      <xs:appinfo>
         <osd:defaultView>
          <widget>
            <textBox>
                <multiLine>true</multiLine>
            </textBox>
          </widget>
        </osd:defaultView>
      </xs:appinfo>
  </xs:annotation>
</xs:element>

Other backward compatibilities

XML audit trail becomes deprecated	The XML audit trail functionality becomes deprecated, and will be suppressed in an upcoming version; it is now disabled by default. The verbose XML format is a hindrance when used with the large volumes which this version addresses: it requires large disk space, and negatively impacts the performance of bulk updates.
HTTP Security Headers	Using a web server in front of EBX® is recommended to handle HTTP security headers. Therefore, by default, EBX® no longer positions any HTTP security header: the property ebx.security.headers.activated is now false by default. If it were unset in 5.9, and if HTTP security headers are not handled by a Web Server, then to keep the former behavior of 5.9, it must be set to true.
Incremental validation behavior	Incremental validation is no longer performed while executing a procedure and when performing updates in the user interface. That is, it is now required to explicitly revalidate a dataset or a table to validate the updates that have been performed since the last validation. For this reason, the method ProcedureContext.disableDatasetValidation is now deprecated. Because validation reports are now persisted in the EBX® repository, the property ebx.validation.report.keepInMemory defined in the EBX® main configuration file is deprecated and is no longer used.
Validation report API	It is now required to close the ValidationReportItemIterator used to iterate over a validation report, in order to release the underlying resources. As a consequence, the method ValidationReportItemIterator.close has been added to the API.
Validation message threshold	The property ebx.validation.constraints.messages.threshold, which is defined in the EBX® main configuration file (ebx.properties), is deprecated and is now ignored. The validation message threshold can now be specified at the data model level, and it is set by default to 1000. The validation message threshold limits the number of validation messages allowed per constraint. Therefore, the method ValidationReport.hasTooManyItemsOfSeverity is deprecated and always returns false.
Data validation	As a consequence of the bug fix CP-19407, the number of validation items displayed in the logs is now limited by default to 100. For more information, see Validation report logs.
Association using explicitly defined dataspace	As a consequence of fixing bug CP-17968, the dataset using an association defined over a foreign key or a link table in an explictly defined dataspace will now be in severe error if the dataset's dataspace does not match with the association reference dataspace. For more information, see Association.
Improved permission checks when executing queries	When a Session is specified, all fields used in the query are checked, and a QueryPermissionException is thrown when a user has no access to one or several fields. Currently, this behavior does not take into account record-dependent access rules set on fields. See Limitations of the permission checks in Query API for more details. As a consequence, using the permissions only to hide in the UI non-sensitive information is highly discouraged, especially if this information is likely to be used for filtering, joining, or sorting in a request. See Using permission for hiding information in the UI for more details.
Legacy user service API is now deprecated	The legacy user service API (ServiceContext) is now deprecated. Please use UserService API.
JavaScript API	The method ebx_getContainerWindow threw an error when EBX® was embedded in a frame on a different domain. Now, it returns the root EBX® frame.
Java interface UserServiceRootTabbedPane	The Java interface UserServiceRootTabbedPane has been replaced by UserServiceTabbedPane: a bug prevented setting two tabbed panes above a user service. UserServiceDisplayConfigurator now returns a UserServiceTabbedPane.
Workflow - Permissions evaluation for workflow launching has been improved	Formerly, permissions for launching a workflow were evaluated when the workflow was launched by the user interface, in Data workflows section, and by data services. Permissions were not evaluated when launching by web component and API. Now, permissions are checked for the following cases: When launched through the UI (including web components). When launched through the SOAP data services.
Workflow - Hierarchical view of a workflow model	By default, the hierarchical view is now hidden. However, you can reactivate it in read-only mode by setting the property ebx.manager.workflow.hierarchyPlanView.hidden to false. See TIBCO EBX® main configuration file for more information. If the hierarchical view is enabled, then it is available in the workflow diagram, from the View menu (Plan View > hierarchy).
Workflow - User task with multiple work items	By default, the service for creating a user task with multiple work items is now hidden. However, you can reactivate it by setting the property ebx.manager.workflow.legacy.userTaskMode to true. Consider using the recommended sub-workflow feature instead. See TIBCO EBX® main configuration file for more information.
The display mode 'link' is no longer supported for group nodes.	For compatibility reasons, groups that were configured as 'link' are now considered 'collapsed'.
Data services pagination	The select operations no longer accept an unlimited page size: In SOAP select operation, the default page size is specified by the property ebx.dataservices.pagination.pageSize.default; it is used when the pagination XML element is not present. The maximum page size is specified by the property ebx.dataservices.pagination.maxSize.default; it is used when the XML element <pageSize>0</pageSize> is set. In REST built-in select operation, like in SOAP, the maximum page size is specified by the property ebx.dataservices.pagination.maxSize.default; it is used when the value 0 is set to the query parameter pageSize. Those mentioned properties are added to the EBX® main configuration file ebx.properties. For more information, see REST built-in select operation and SOAP select request on table.
Foreign key label and sortable search strategy	The custom display for presenting a foreign key must now be sortable: thus, the fields that are part of the custom display must have only sortable search strategies. This also means that previous models that define foreign key display using non sortable fields such as osd:text fields will be in error. Note that this applies to both defaultLabel and pattern elements.
REST built-in select operation with body	To invoke a select operation from HTTP POST with a body, you must use :select URL extended action; otherwise, the insert operation is used. See Select operation for more information.
REST built-in Exception handling	The key severity is replaced by the key level in the JSON exception response, and the value becomes human readable. See Exception handling for more information.
REST Toolkit URI default JSON serialization	The previous URI default JSON serialization suffered from asymmetrical representation. URI are no longer serialized as JSON objects but as String values. This change restores a uniform behavior when URI are: Used as Java method returned value. Held in collections. Stored in DTO objects.
REST Toolkit URI builders for built-in	The URI builders API for built-in have been reworked, leading to multiple methods deletion. The builders' API are more in line with the Java builder design pattern, providing enhanced modularity, genericity, and intuitive developments. See URIBuilderForBuiltin and URI builders for more information.
Default directory	The "Job Title" field is no longer automatically formatted to the proper case.
Association with inconsistent definition	As a consequence of fixing bug CP-22691 , if an association is not consistent in the context of a specific dataset then: It is condisered as hidden from a permission perpective. That is, the association is hidden in the user interface and in dataservices as well. It is excluded from the validation of the container dataset, table or record. An exeption is raised in the existing API that compute the result of the association. For more information, see Association.

Cloud Software Group, Inc. Support

The EBX® Support can now be reached from the https://support.tibco.com website (an account is mandatory in order to access, follow the 'Register' process). Visit the About Support section of this page for more details.

6.0.13 Closed issues

Container Edition

• CP-24408 A stopped EBX Container Edition container may fail when started again

  A stopped EBX Container Edition container may fail when started again.

Core engine

• CP-20968 The loading screen displays for a long time when applying the <emphasis role="bold">Apply last modifications</emphasis> service.

  This occurs when the service is run on a child dataset that contains 100,000 records.

• CP-23572 The "present in" operator is not available on an enumeration field of the HTML data type.

• CP-24257 A NullPointerException occurs when levels for logging categories are defined by a variable.

• CP-24290 Error importing a perspective archive when a perspective has been duplicated.

  An error occurs when importing a an archive contain a perspective that was duplicated.

• CP-24345 NullPointerException when exporting a record, table or dataset created in severe error

  A NullPointerException occurs when exporting a record, table, or dataset that was created in a state of severe error.

• CP-24346 ClassCastException when calling SchemaFacetTableRef.getLinkedRecords() in a Widget.

  A ClassCastException occurs when calling SchemaFacetTableRef.getLinkedRecords() in a widget.

• CP-24379 Repository purge operation fails at validation reports cleaning step.

  A repository purge operation fails at the validation report cleaning step.

• CP-24405UIDropDownList on a foreign key does not display elements.

• CP-24579 An error occurs when rebuilding indexes in a child dataspace

  An error occurs when rebuilding indexes in a child dataspace.

• CP-24626 Default value is set on existing records in history tables when adding a new column

  When adding a new column in a historized table its default value is added to the existing records in the history.

• CP-24706 Editing Greatest Value set field to a greater value doesn't affect Auto-increment when creating a new record

• CP-24763 Missing com.orchestranetworks.service.directory.EncryptPassword class.

  The com.orchestranetworks.service.directory.EncryptPassword class mentioned in the main configuration file documentation is missing in the product package.

• CP-24895 Values from uncommitted transactions can corrupt table size estimation.

• CP-24944 When extending SchemaDocumentation, displayed links are not clickable.

• CP-25032 Log failure can occur during index refresh at commit.

  This can disrupt synchronization between the cached index and the table content after commit.

Data modeling

• CP-24350 Incorrect behavior occurs after clicking the <emphasis role="bold">Clear</emphasis> button on a toolbar's service dropdown list.

Data services

• CP-17620 Optional parameters in SOAP header do not work using default namespace without prefix.

  When calling a SOAP operation, adding additional parameters in the header does not work when using the default namespace without prefix.

• CP-24450 Published hierarchy views should be filtered from Look up table views operations.

• CP-24606 Lineage WSDL configuration does not considered UI selected context.

  The lineage WSDL configuration does take the selected UI context into account.

• CP-24653 The REST built-in transient operation should ignore association fields and value function nodes.

• CP-24664 SOAP body validation error for select operation with some activated parameters.

  When a select operation enables exportCredentials with a value of true, the XML body response is invalid.

• CP-24673 NPE when branch does not exists using SOAP operation.

  When using a SOAP operation, a NullPointerException occurs if the defined branch does not exist.

• CP-24722 Association field must be ignored using REST builtin import/update operations.

  When using the REST built-in import/update operations, ignore association fields.

• CP-24961 A PathAccessException occurs when importing XML.

  The exception causes the import to fail when a record contains an empty group that comes directly after a list of groups.

• CP-25013 A REST built-in selector field must not be present for a read-only field.

Data validation

• CP-18807 In a user service, an exception may be thrown on page redisplay.

  The User Service API can be used to customize the form for creating a record. If a record contains a foreign key, an unexpected exception may be thrown, when the user service is redisplayed.

• CP-21583 In a UserService, the record form submission is not blocked when an error message is added in ConstraintOnTableWithRecordLevelCheck#checkRecord.

• CP-23950 The record validation report is computed for each tab to be displayed in a record form

  The header of a record's form displays the up-to-date validation report of the related record. Each tab of the form has its own header. As a consequence the validation report is computed as many times as the number of tabs to be displayed in the form. This can lead to performance issues if the table contains many constraints.

• CP-24171CalciteException Table not found when validating a table with a table reference filter using a table reference node.

  An exception is thrown when a table reference filter uses a node that is a table reference to a table located in another dataset.

• CP-24192NullPointerException in code generated by calcite for a table reference filter validation.

  A NullPointerException occurs in code generated by Calcite for a table reference filter validation.

• CP-24336 xs:pattern facet is not applied on the XML string representation of dates.

  The xs:pattern facet checks the string representation of the Java Date object instead of the XML string representation. 

• CP-24483NullPointerException while updating the persisted validation report.

  A NullPointerException occurs while updating the persisted validation report.

• CP-24526 Slow performance when accessing the persisted validation report

  Performance is impacted when accessing the persisted validation report.

• CP-24777 An IllegalStateException occurs when accessing a validation dataspace after an aborted deletion of a functional dataset.

• CP-24904 Validation fails with a ClassCastException.

  This is caused when PvrActionForTableRef generates a query plan with the wrong FK_AS_STRING parameters

• CP-24923 A ClassCastException occurs for a contextual table reference filter.

• CP-24933 A contextual table reference filter with multi-valued fields returns incorrect results.

• CP-25019 There is a null record in ValidationReportItemSubjectForTable.getRecords().

• CP-25148 A NullPointerException occurs in ValueContextOnCfContent when validating a table reference with a filter.

Data workflows

• CP-24148 No way to clear a profile field in a workflow step.

  There is no way to clear a profile field in a workflow step.

• CP-24578 Workfow complex objects should be expanded by default.

• CP-24769 The documentation for the Xpath parameter of the update a record field value script task is not correct.

Dataspaces & History

• CP-23598 The comparison view should show icon difference changes

  The dataspace comparison view should be improved.

• CP-24934 A NullPointerException occurs during the dataspace compare action.

Documentation

• CP-25026 Incorrect Java EE platform for WebSphere AS 8.5 and 9.0 support.

  The Java EE platform for WebSphere Application Server 8.5 and 9.0 support is incorrect.

• CP-25092 The double quote usage for the Windows path in the Tomcat installation note is incorrect.

Import / Export

• CP-24263 An error occurs when importing the exported XML file on "Perspective menu".

  When users export XML file on "Perspective menu" and then import that file, an error message is displayed.

Infrastructure

• CP-24678 A deadlock happens during server starts with rolling logs.

Queries

• CP-24291 Field trimmer removes all fields from the input to a LogicalAggregate with COUNT()

  A ClassCastException occurs during the field trimming step.

• CP-24319 Result of RequestResult.getSize() might be incorrect and vary from call to call

  The RequestResult.getSize() method might return an incorrect result and vary from call to call.

• CP-24551 Performance problem when setting request log level to DEBUG because of a slow RelNode to SQL translation

  A performance issue occurs when setting the request log level to DEBUG.

• CP-24608 Although set as non confidential, hidden nodes cannot be used in a foreign key filter when under a complex node.

• CP-24631 Sort by resource field (and pagination) when the Lucene index is not available throws an exception.

• CP-24677 An SQL query throws a CannotPlanException due to HepRelVertex in the plan after the Hep phase.

User interface

• CP-2041 [33485] Fields are unexpectly filled automatically while creating a record through an association.

  When creating a record through an association, fields in the creation form should be pre-filled, but not read-only.

• CP-17509 An error occurs when using a web component URL that selects a perspective and a branch.

• CP-21684 The tables under a group still display when the group is set to hidden.

• CP-24131 The highlighted bar is hidden in Data Model and Workflow sections.

• CP-24668 A confirmation message is not displayed for a user service declared on a dataset in a perspective.

• CP-25000 Errors are not clickable.

• CP-25037 Performing scripting in the editor causes double key entry.

Users & Permissions

• CP-24489 The dropdown to set the owner of perspectives is only showing profiles with the built-in administrator role.

6.0.12 Closed issues

Core engine

• CP-24583 An error occurs when rebuilding indexes in a child dataspace

• CP-24634 "CfOccurrence could not be fetched" error.

  Errors with messages "java.lang.IllegalStateException: CfOccurrence with ref (...) could not be fetched from table (...): record have been deleted a transaction" prevent the access to the table contents.

• CP-24657 "java.lang.IllegalStateException: This index is read-only" error while updating a record.

  An IllegalStateException error can occur when updating a record.

Infrastructure

• CP-24586 All handleRepositoryStartup method should complete before opening access to EBX®.

6.0.11 Closed issues

Core engine

• CP-24178 AdaptationHome.getTerminationDate() does not return null on open dataspaces

  The AdaptationHome.getTerminationDate() method does not return null on open dataspaces.

• CP-24381 Repository purge operation fails at validation reports cleaning step.

  A repository purge operation fails at the validation report cleaning step.

Import / Export

• CP-24279 An error occurs when importing the exported XML file on "Perspective menu".

  An error occurs when using the perspective menu to import an XML file that was exported from the perspective menu.

6.0.10 Closed issues

Cloud - Core functionalities

• CP-23530 Useless avatar options on EBX Cloud.

  On EBX Cloud, the user pane and the history are not available, so the two fields 'Avatar displayed in the header' and 'Avatar displayed in the history' are useless and should be hidden.

• CP-23654 Display user email and review info message for 'send email' built-in script task.

  When enabling the 'send email' option in a built-in script task the user identifier is displayed instead of the email address.

Core engine

• CP-20180 Table trigger called when there is no modification on multivalued field.

  When importing a record using the 'Import XML' built-in service, the handleBeforeModify and handleAfterModify methods of TableTrigger may be erroneously called.

  This occurs when the imported record has an empty value for a multivalued field. Although the exact same record is present in the table, the two methods are still called.

• CP-22283 Supplementary Unicode characters are incorrectly serialized and incorrectly exported to XML archive.

  Unicode characters in range [ #x10000 - #x10FFFF] are incorrectly split into two characters, when exported to an XML archive.

• CP-23668 Pre-processing return inconsistent result.

  When running the <emphasis role="bold">Evaluate Matching</emphasis> service multiple times, the pre-processing step returns inconsistent results.

• CP-23778 An error occurs when accessing an association table with non sortable fields used as sorted column.

  An error occurs when accessing an association table with non-sortable fields used as a sortable column.

• CP-23931 When creating a new record containing a multivalued complex targeted by a validation rule, new values are not taken into account

  When creating a new record containing a multi-valued complex field targeted by a validation rule, new values are not taken into account.

• CP-23951 Composite pks are not projected correctly in sub queries.

  When clustering with relationship, the score could be incomplete when the primary key of a linked table is composite. 

• CP-23957 Increase cost/precision balance when clustering.

• CP-24190 Inconsistent index rebuilts.

  After rebuilding an index, some records may be missing and the table in EBX appears to be out of sync with its replica.

Data modeling

• CP-23427 An error occurs when publishing a data model that contains invalid characters on its table/field label.

  An error occurs when publishing a data model that contains special characters in its table/field label.

• CP-23674 Sortable fields are not displayed when defining a tabular view on an association.

  Sortable fields are not displayed in the DMA when defining a tabular view on an association.

• CP-23828 Cannot create a linked field under a group.

  An error occurs when publishing a data model that contains a linked field defined under a group.

Data validation

• CP-23685 ClassCastException happens when validating a validation rule using aggregated lists

  A ClassCastException occurs when a table includes a complex data type where the maximum number of values is unbounded and the table's validation rules check aggregated lists.

• CP-23930 Foreign key filter not applied correctly.

  A foreign key filter ./fkField1 = ${../fkField2}, where both fields point to a table with a composite primary key, does not accept a valid change when creating or editing a record in the user interface.

• CP-24045 Validating a uniqueness constraint on a foreign key node throws a MalformedXPathPredicateException: The specified value: null is invalid.

  A MalformedXPathPredicateException exception occurs if a uniqueness constraint is defined (among others) on a foreign key field to a table whose primary key contains at least a field of certain types (integer, decimal, and possibly others).

• CP-24120 Computed value is unnecessarily recomputed during a revalidation if a constraint defines a dependency on insert/delete/modify to the table that contains it.

  Changes performed since the last validation of a table or a dataset are computed to update the validation report according to them. If a constraint defines a dependency on insertion, deletion and modification to a table that has a computed field (using the method DependenciesDefinitionContext.addDependencyToInsertDeleteAndModify) then the computed value is unnecessarily recalculated when estimating the changes.

Data views - hierarchies

• CP-23711 Pagination issue when closing a dataspace in the dataspace hierarchy.

  When closing a dataspace in the dataspace hierarchy view, the last page in the view displays instead of the first page. 

• CP-23803 A ragged hierarchy is not pruned correctly.

  On a ragged hierarchy does not displays correctly when searching for a target leaf (or setting the pruned mode) on deep (more than 2 levels) recursive branch. 

• CP-23891 Cannot create a hierarchy on a list field.

  Hierarchies cannot be created on list fields.

• CP-23908 A sort exception might occurs on some hierarchies.

  A sort exception can occur on certain hierarchies.

• CP-23912 The open/close state of the hierarchy members is not correctly managed.

  A hierarch is not collapsed when a filter is removed and this can extend the time required to display the hierarchy.

• CP-24075 The records are displayed incorrectly after the pruned mode on the hierarchy view has been disabled.

  When users first run the 'Enable pruned mode' service then run the 'Disable pruned mode' service, the records are displayed incorrectly.

Data workflows

• CP-23426 A fatal error occurs when publishing a workflow model containing an invalid characters in its label.

  When users create a workflow containing special characters in its label then publish it, a fatal error occurs. After restarting the server, a technical error occurs.

• CP-23801 Workflow Modelling Defaults Not Used.

  Default values set for workflow modeling are not used.

• CP-23970 Work Item URL is not working as expected.

  When using the built-in data context variable 'workflow.workItem.link.AllocateAndStart' in the message template used for sending emails to offered users, after the user logs in, the workflow graph view is displayed instead of the work item page.

Dataspaces & History

• CP-24051 Unclear error on access to a deleted dataspace.

  When a dataspace is deleted, then the method getDescription() of AdaptationHome throws a NullPointerException. It should throw a clearer exception such as IllegalStateException specifying the dataspace identity.

Documentation

• CP-21198 Incorrect directory users XPath from documentation samples.

  The documentation samples include the wrong directory in an XPath sample.

• CP-23873 Incorrect title in documentation index page.

  The title that displays in the documentation index page is incorrect.

• CP-23980 The addons deployment documentation contains inconsistencies.

  The addons deployment documentation is not correct for Weblogic and Websphere; as the addons web applications are now deployed in the EBX EAR, the configuration is done at the EAR level and not the web application one.

• CP-24109 Incorrect documentation for ProcessLauncher.

  In the developer guide, chapter 'Workflow model', the sample in the last section is incorrect. It does not use the public Java API to get a ProcessLauncher instance.

• CP-24186 Incorrect H2 database version in Tomcat and WebLogic documentation notes.

Filter

• CP-23771 An error occurs when using osd:is-empty or osd:is-not-empty on a list field.

  A MalformedXPathPredicateException occurs when applying osd:is-empty or osd:is-not-empty on any list field.

Java EE integration

• CP-23953 Inconsistency in the documentation for the JBoss application server.

  The supported environments section of the documentation specifies that JBoss EAP 7.4 or higher are supported, whereas the notes describe a deployment for JBoss 7.1.

Miscellaneous

• CP-24113 On application server startup, task scheduler initialization throws a NullPointerException.

Queries

• CP-23817 SQL query with UNNEST operation on multivalued complex field throws exception on query optimization (field trimming step).

  An SQL query that includes an UNNEST operation on multivalued complex field throws exception on query optimization (field trimming step).

• CP-23851 An error occurs when applying an advanced search on the 'Component library' table of a data model.

  When users apply an advanced search combined between the 'matches (case sensitive)/ matches' and 'is empty' operators, an error occurs.

• CP-23876 An error occurs when applying both quick search and advanced search on the hierarchy view of a table with a record permission.

  When record permission is activated on a table's hierarchy view, an error occurs when applying a quick or advanced search.

• CP-24023 LuceneFieldScan on sort field with null values can return incorrect results.

  A LuceneFieldScan operator using a sortField, for example reading values from a Lucene field used in ORDER BY can return incorrect results when the field contains null values.

User interface

• CP-1503 [36316] Permalink to a dataspace in node scope leads to "Page not found" error page.

  Accessing a permalink (generated in the data tab) to a dataspace in the scope of a node displays a "Page not found" error.

• CP-21157 Confirmation message in UserServiceDeclaration.OnDataset is not working

  The confirmation message in UserServiceDeclaration.OnDataset does not work.

• CP-22129 Normal Redirection Exit Policy URL

  From EBX® version 5.9.13, special characters must be escaped in the URL value set for the Normal Redirection Exit Policy.

• CP-22295 [All browsers] The scrollbar is not displayed on 'Refresh all data models'.

  When invoking 'Refresh all data models', the scrollbar is not displayed in the pop-up.

• CP-23335 The create and duplicate services are not displayed in the grid edit of an association.

• CP-23812 Performance issue with Chromium-based browsers.

  There is a performance issue when opening pages that contain FormPaneWithTabs when using Chromium-based browsers on an external monitor.

• CP-23982 Perspective Menu: Action record should not be able under root.

  When creating a perspective menu in a custom perspective, a child record of the <emphasis role="bold">Action</emphasis> type can be created.

Users & Permissions

• CP-23510 Limited permission checks for associations using XPath predicates.

  Permissions are not checked for contextual fields used in the XPath predicates of associations (filter and/or definition).

• CP-23838 Associations with predicates that include hidden fields throw an exception.

  When a hidden field is used in an association's predicate, a QueryPermissionException is thrown. In such a case, the association should be hidden.

• CP-23917 ClassCastException on new audit trail logging.

6.0.8 Closed issues

Container Edition

• CP-23707 Java memory settings may be incorrect in Container Edition.

Core engine

• CP-23492 Search strategy association table is not validated when a field is modified into a linked field.

  Search strategy association table is not validated when a field is modified into a linked field. When modifying an existent field associated with a search strategy into a linked field, the record holding the field is marked as in error while the entire table is fine. 

• CP-23601 Data model compilation takes a very long time to update localized validation messages.

  Altering the property ebx.locales.available in ebx.properties causes a very long startup time.

• CP-23709 Run match miss matches when using synonyms/phonetic.

• CP-23744 Slow dataspace merge and on-commit replication.

  The on-commit replication is slow because it triggers a query to the database for each create, update and delete on EBX® record. These statements must be batched. Additionally internal primary key comparisons can be optimized, which also improves the performance of dataspace merge.

• CP-23819 Failover backup server does not start up.

• CP-23858 Migration fails if a bloc contains supplementary unicode characters.

  EBX®6 is unable to migrate characters containing supplementary Unicode characters (encoded on more than 2 bytes).

Data modeling

• CP-22209 The constraint description from a foreign key used in an association does not resolve properly.

  This issue happens when using an association which relies on a foreign key that targets a specific dataset.

• CP-22690 Restrict access to associations when there is an inconsistency in their definitions

  In CP-17966, we have imposed some restrictions about the definition of associations regarding child dataspaces. That is, when an association is based on a foreign key or a link table in another dataspace, the definition of the association becomes inconsistent in the children of the dataspaces involved in the relationship. In this case related datasets are considered in severe error. As this behaviour is considered too strict we have decided to relax this constraint and to only restrict the access to these associations.

• CP-23786 In TIBCO EBX®, cannot import a data model with a scripted validation rule.

Data services

• CP-23747 Incorrect HATEOAS URLs for hierarchy views in the REST 'look up table views' response.

  When data lineage base URL is overridden, the returned JSON detail URLs are incorrect.

Data views

• CP-23344 [WCAG 2.0 - 2.1.1] The tree path wizard is not keyboard accessible.

  The hierarchy node labels may be customized in the hierarchy view configuration. The label entry can be made easier for the user with the wizard button. However, this button and the popup used to edit the label are not accessible with keyboard.

Data views - hierarchies

• CP-23717 Error when running 'Validate' or 'Move' on a hierarchy view with filtered records.

  A user first applies a validation search then selects all records. Then a fatal error occurs when running the 'Validate' or the 'Move' services.

Data workflows

• CP-20881 Workflow model publication is too slow.

  It takes too much time to validate each workflow model for publication service. We need to improve this: by improving validation or by validating workflow models only in a second step (after a first selection). Only the current workflow model (by default selected), could be validated in the first step.

• CP-21358 After closing a snapshot in Publication history service, user is redirected to a blank workspace.

  In the workflow publication history, when closing a snapshot, the user should remain on the current service.

Dataspaces & History

• CP-19939 Dataspace is still displayed even though it is already closed.

  The child dataspace is still displayed when users close it under the 'Actions' menu from the dataspace area.

• CP-23649 Dataspaces in administration can not be accessed when the content is in an inconsistent state.

  When a dataspace record exists in the dataspace table but the corresponding dataspace is missing, a NullPointerException occurs when accessing the 'Dataspaces' section under 'Administration > Repository management'.

  Launching 'Execute purge' under 'Administration > Clean up tools' is supposed to delete the records in an inconsistent state. However, the purge procedure fails to perform the cleanup.

Documentation

• CP-23689 Error on documentation indexing when a custom language is set as default.

  When a custom language is set as default, an error is written in the log file on application server startup: "java.lang.IllegalArgumentException: No analyzer implementation declared for this language".

• CP-23810 Limitation on history access is not always documented correctly.

  There is a limitation on history access when a record permission scripted rule is set on the functional table. In the documentation, it is correctly documented in the 'Permissions' chapter but not in the 'History' chapter.

Queries

• CP-23623 Incorrect NULL, NOT NULL counting with multi-value password field.

User interface

• CP-20524 'Page not found' after logout and login, using permalink.

  When a user accesses a page using a permalink, then logs out and logs in, a blank page is displayed with the error message 'Page not found'.

  On IE 11, after changing the 'Close button' to 'None(default)' and accessing the link, the close button still exists.

• CP-21749 The 'Request cancelled' message is not displayed correctly in an association table.

• CP-22394 The 'Accept' button is unavailable when custom form is enabled in the workflow context.

  The 'Accept' button is unavailable in a work item when using the 'Duplicate' a record built-in service, if a custom form is enabled in the data model extensions.

• CP-22767 A snapshot disappears on the dataspace navigation pane when closing its 'Information' page.

  When users access the 'Information' page of a snapshot then close this page, the snapshot is not displayed on the navigation pane.

• CP-23293 The dataset creation page should show the exact error cause.

• CP-23620 Error when expanding a node in the 'Dataspaces/snapshots' hierarchy view.

  When users create and apply a hierarchy view with the dimension 'Dataspaces/snapshots > childBrandCreationAllowed', and expand any node on this view, an error message occurs.

• CP-23749 Submit button in legacy user service redirects to the login page.

  Clicking a submit button in a legacy user service redirects to the login page instead of the running user service.

• CP-23814 Rescue mode URL is not working.

  When the "rescue mode" is enabled by setting the property frontEnd.rescueMode.enable to true, the URL specified by the product documentation does not work.

  The page is blank.

Users & Permissions

• CP-23644 NPE using the default directory and an empty login value.

  A NullPointerException is thrown when the login value is empty using directly or indirectly the Repository.createSessionFromLoginPassword(login,password) API. This impacts the programmatic service, the SOAP/REST built-in, the webcomponent.

6.0.7 Closed issues

Core engine

• CP-21254 InvalidPrimaryKeyException occurs when validating a dataset child.

  The error InvalidPrimaryKeyException occurs when validating a dataset child, after the primary key of a table has been changed.

• CP-22252 On the workflow diagram, the steps are not displayed.

  When a user modifies the ebx.properties / server.xml files by adding a custom path, the steps on the workflow diagram screen are lost.

• CP-22278 During matching, records that do not match with the center must not be marked as processed.

  During matching, when a record is in the inner ring of a cluster (very similar to the center) but does not match the center, setting it as processed sometimes causes missing results.

• CP-22410 Do not restore the old data after changing the type of a primary key.

  Type conversion should not be supported for a primary key field. Thus, the content of the table must always be reset to empty after changing the type of a primary key field, even if the data type was used previously. The product documentation on data model evolutions must also be changed accordingly.

• CP-23035 Clustering query throws exception when sorting by a foreign key.

• CP-23187 Matching does not return an appropriate result for a foreign key matching field, if this field is in a terminal group.

• CP-23443 Issues with the Java method SessionPermissions.getNodeActionPermissionForService.

  The method SessionPermissions.getNodeActionPermissionForService can throw a NullPointerException. This behavior is different than that of the version 5.9 and it is not specified in the Javadoc.

• CP-23465 XML and CSV import services are available on the data model descriptors dataset.

  Services for importing an XML or a CSV document are available on the technical dataset that contains the serialization mappings used by data models. These services should not be available since this dataset contains sensitive data.

• CP-23469 NullPointerException occurs when building a descriptor

  On data model compilation, when descriptors are being built, a NullPointerException is thrown.

• CP-23473 On server startup, the 'kernel' log has an empty 'EBX(R) CONFIGURATION'.

• CP-23476 Infinite loop with high GC activity when an index file cannot be deleted.

  When the application tries to remove an index, it falls into an infinite loop triggering a high CPU consumption, due to the Garbage Collector, if one of the local files cannot be deleted (whatever the reason).

• CP-23484 NullPointerException while deleting and closing Dataspaces recursively.

• CP-23507 Error on opening a foreign key combo-box.

  When opening a foreign key combo-box, an error occurs. The log displays the following entry: "IllegalArgumentException: Specified context must be in the same data model as the field node holding this foreign key constraint."

• CP-23546 Useless validation by the scheduler

  The execution of a scheduled task triggers an explicit validation of the result table. This is unnecessary while increasing the repository size.

• CP-23562 Quick search fails when operating with a foreign key composed of a 'Boolean' or a 'Locale' field.

• CP-23573 For foreign keys, custom search strategy of underlying primary key fields is ignored.

  It is possible to define custom search strategies on primary key fields. However this customization is ignored when performing a quick search on a table that defines a foreign key without a custom label and which targets a primary key field with a custom strategy. As a consequence the result of the quick search can be incorrect.

• CP-23585 The bloc size drifts.

  The size of the block increases by two bytes with each round trip readBloc() - flush().

• CP-23590 "IllegalStateException: This index is read-only" caused by a specific kind of merge and concurrent writes.

  A dataset has been created in a child dataspace. When this child dataspace is being merged, this dataset is not protected against concurrent writes. This can lead to the exception "java.lang.IllegalStateException: This index is read-only". This kind of scenario may occur when validation reports are copied on a dataspace creation.

• CP-23600 Cannot merge dataspace, even after full index rebuild.

  Happens when copying the validation report on branch creation. This has probably the same cause as CP-23590.

• CP-23608 A QueryParseException 'column X not found in table T' occurs when validating a uniqueness constraint containing an ambiguous field referring to another table.

• CP-23630 Multi-occurenced matching fields are never disabled.

  Multi-occurenced matching fields are never disabled. Under certain conditions, it causes an exception during matching: "IllegalArgumentException: At least two boolean clauses must be specified."

• CP-23715 Records appear as if they were duplicated.

  When accessing a table, all records appear as duplicates. This issue appears for a table migrated from EBX 5 and then the index was not committed but recovered (for example, because of a brutal shutdown of the server). This scenario leads to a duplication of the entries in the index.

Data modeling

• CP-23583 When the user performs a search on the 'Data types' table of a data model, an error occurs.

• CP-23596 An errors occurs when defining a uniqueness constraints on several fields under a reusable type.

  In the data modeling assistant, an error occurs when defining a uniqueness constraint on several fields under a group that reuses a complex type from another data model.

• CP-23681 The 'Data model descriptors' entry in the Administration is not localized.

Data services

• CP-23575 The REST select operation returns '400 Bad request' if filter expression contains an encoded '&'.

• CP-23632 A null body of a REST built-in PUT request causes a server error.

  A 500 error is returned when running the request PUT with a null body. The response should be a 400 with a message that describes the error. 

Data validation

• CP-23198 Validation does not stop when validation message threshold is reached, if a constraint has a try-catch block.

• CP-23415 Validation causes a PathAccessException for a specific kind of foreign key.

  If a foreign key field is not Lucene-optimized (for example it is an inherited field) and it refers to a table that it is in a different dataset, then a PathAccessException is thrown on validation.

• CP-23452 An informal report returned by Adaptation.getValidationReportForNode(...) does not contain the persisted messages.

• CP-23456 Subjects of a validation report of a child validation dataspace are from the initial snapshot, but should be from the branch.

• CP-23486 Query optimization rule can throw a ClassCastException when trying to remove a Sort operator on top of a MergeJoin.

Data views

• CP-23594 Tracking information is null in a sub-session.

Data views - hierarchies

• CP-23421 Hierarchy views pagination does not work properly with the 'Find'.

  In a hierarchy, a dimension dependency is based on an enumeration. In a 'Find', to display the next found member, the pagination fails.

• CP-23578 The 'ebx-validation' records in the hierarchy of dataspace in the 'Admin' panel should be hidden.

Data workflows

• CP-20100 In the workflow publication history, the export service redirects to the tabular view.

• CP-20505 The 'Workflow model steps' table is displayed after creating a snapshot on the workflow.

  When users create a snapshot on a workflow model, the system shows the 'Workflow model steps' table on the workspace of the existing workflow.

  This bug also occurs when exporting a snapshot of a workflow model.

• CP-20831 In the workflow model, when closing 'Validate service' or 'View validation report', user is redirected to a blank workspace.

Dataspaces & History

• CP-22125 Error when anonymizing foreign key fields in history

  If we port to trunk the changes from CP-22125, the test TestHistoryAnonymization fails because of field

  ANONYMIZED_VALUES_BY_PATH.put(F_REF, "0|0"); (it passes OK on branch 5.9).

  The reason is that the tableRef field has a TNode#serializationDescriptor = null because we don't generate descriptors for history schema in EBX 6.  So, it fails in _IccRecordKey_Composite#toInternalValue(TNode,Object))_ when copying TxChanges.

• CP-23023 An exception occurs on boot if 'ebx-directory' is not owned by an administrator.

Java EE integration

• CP-23665 Startup fails if Apache Log4j 2 is deployed twice on the server.

Queries

• CP-23568 Out of memory error when executing a Matching query.

  This bug is caused by a wrong hint in a SQL query used for the Matching.

User interface

• CP-20836 Remove service "Clean up unreadable records" from dataspace administration.

  The service "Clean up unreadable records" should be removed from dataspace administration since it is now useless.

• CP-22555 In a user service, registering a table from a different dataset for creating new records does not work.

• CP-23030 Make wizard panel keyboard accessible (WCAG).

  Accessibility issue: when the wizard panel is displayed, there is no way to navigate or select content with the keyboard. Moreover, when the tab or shift+tab key is pressed, the focus may go on user interface components that are outside the wizard panel. In addition, a tab key press on a tree inside the wizard, will set focus on all nodes. This behavior is not compliant with WCAG standards.

• CP-23159 The old filter button in the Grid Edit view is still displayed.

  This button should not be displayed if it is not displayed in the non Grid Edit view.

• CP-23298 The 'Record form: rendering mode for nodes' has an invalid value by default.

  EBX offers the possibility to configure the display mode of 'Record form: rendering mode for nodes' of non-terminal nodes in record forms. This option can be accessed in the Administration area for configuring the Ergonomics and Layout of a given perspective, from the Administration > User interface > User interface configuration section.

  In EBX 5, this option allows you to choose one of the following three display modes: (1) Link, (2) Expanded, and (3) Collapsed. In EBX 6, the 'Link' mode has been removed. However, it seems that it still used as default value for this option. When changes are made to another field and the form is saved, the option 'Record form: rendering mode for nodes' displays that the current value (the default value) is no more valid.

Users & Permissions

• CP-3935 [21388] A field with a default view defined as hidden should not be exported in CSV or XML exports.

6.0.6 Closed issues

Core engine

• CP-22742 In validation contexts, Adaptation.getHome() returns a technical snapshot , instead of the concerned dataspace.

• CP-22893 Version v2.x of H2 database is not supported

  An error occurs with the message: org.h2.jdbc.JdbcSQLSyntaxErrorException: Table "CONSTRAINTS" not found

• CP-23132 A read-only index error occurs while modifying a record.

  Modifying a record could fail with a "java.lang.IllegalStateException: This index is read-only" error.

• CP-23148 CfOccurrence could not be fetched.

• CP-23154 Error on index rebuild, after deleting a parent index.

• CP-23156 A deadlock can occur when a dataspace creation and an index rebuild occur simultaneously.

• CP-23193 Triggers handleAfterCreate and handleAfterModify slow down a mass-update.

  If TableTrigger.handleAfterCreate is implemented and calls aContext.getAdaptationOccurrence(), then a mass-creation of records becomes extremely slow. The same happens for TableTrigger.handleAfterModify.

• CP-23237 AlreadyClosedException on matching merge, even after index rebuild.

• CP-23239 An existing record is not found.

• CP-23309 An error occurs when refreshing replicated tables during the compilation of data models.

  An error occurs during the update of replicated tables when data model evolutions have been done on data models used by some datasets in the same dataspace, if related indexes must be rebuilt. This issue occurs if the other modified data models define  replicated tables that must be refreshed in the same dataspace.

• CP-23358 An error occurs when using the ReverseRelationshipInterDataset.getReferringRecords(Adaptation) API, if a referred dataset has severe errors.

  A NullPointerException occurs when calling the method ReverseRelationshipInterDataset.getReferringRecords(Adaptation) if a dataset that has a relationship to container table of the specified record has severe errors.

Data modeling

• CP-22344 Incorrect error message when publishing a schema to a read-only module.

  When users import a file from a module into the DMA, then publish that model, the message "_unexpected error publishing data model. Please contact your administrator_"  is displayed if the module folder is read-only, i.e. EBX has no access permissions to the schema publication location folder. 

• CP-22954 Startup hangs after deploying data model changes.

  Startup hangs after deploying data model changes because of daily log rotations.

• CP-23033 An error message is displayed when selecting the 'Not defined' value for the 'Configuration' field on the 'Component library' table.

  When users select the 'Not defined' value for options under 'Rule' tab of the Component library table, an error message is displayed.

• CP-23242 The 'data type' property of fields using a type from an included data model is not displayed.

  In the data model assistant, the data type of a field that uses an included type is not displayed in its form.

Data services

• CP-22397 The stack trace is included in the SOAP response.

• CP-23043 The REST built-in 'insert' operation does not support multivalued primitive fields or foreign keys in the compact format.

• CP-23300 Implement and document the zero page size behavior for REST and SOAP operations.

Data validation

• CP-23087ClassCastException when validating a table reference filtering on a multi-valued field in the referenced table.

• CP-23197 Error validating a table reference with a filter on a multi-value field.

• CP-23289 StackOverflowError on data validation, if dataspace has many snapshots.

• CP-23376 The technical field "Validation Dataspace" should be hidden, in the Dataspace information service.

Data views - hierarchies

• CP-21417 The number of selected record is wrong when choosing the filtered record, then selecting all records in the hierarchy view.

  In the hierarchy view, when users apply quick search, tick on the filtered record then select all records, the number of selected record is wrong.

• CP-22654 A level predicate set on a recursive hierarchy greatly hinders performance.

• CP-22883 Hierarchy branch selection does not work if a search is done.

  If a search is done, the "Select all children" on a member does not consider the search filter.

• CP-22915 A new record should be checked after creation from a parent node in a Hierarchy view.

  A new child node is not selected after using the 'Select and display' > 'Select records' > 'All' service in a hierarchy view.

• CP-22931 The value of 'selected' is incorrect after running the 'Move' service in a hierarchy view.

  When the user moves 2 records to another position in a hierarchy view, then unselects them and moves other records: the value of 'selected' is wrong. If the user unselects again, a negative value is displayed.

• CP-22996 A hidden record is still selected on a hierarchy view.

  A hidden record is not displayed on a hierarchy view, but is still selected when executing the "Select All" action on that view.

• CP-23003 Hierarchy members are not selected if the primary key of the target table is composite.

  When running the "Select all children" service in a hierarchy view, a checkbox is not checked.

• CP-23011 After deleting a record in hierarchy view, the number of selected records is not updated.

  On hierarchy view, after selecting a record then deleting it: the number of selected records on the top right is not updated.

• CP-23020 The search result is not updated after changing the search value on a hierarchy view.

  When users search a value of a parent node on a hierarchy view, then execute a different search on a child node then apply, the search result is not updated.

• CP-23032 A blank screen is displayed when choosing 'Select all children' on the hierarchy view of table 'Extensions'.

• CP-23038 The quick search is not applied when searching on a hierarchy with an enumeration date dimension.

• CP-23117 The hierarchy action 'Expand 2000' expands too much.

  The hierarchy action 'Expand 2000' should stop expanding the page around 2000 members displayed, but it continues.

Data workflows

• CP-23083 The 'Monitoring' label in Data Workflows is not externalized.

Dataspaces & History

• CP-22134 Error when anonymizing lists in history

Filter

• CP-23036 A permission rule combined with a quick search leads to 'CannotPlanException'.

Java EE integration

• CP-22983 JDOM 2.0.6 library should be upgraded to version 2.0.6.1.

Queries

• CP-22886 Exception in query with parameters.

• CP-22944 QueryPermissionChecker does not check fields under certain operators.

User interface

• CP-21186 The 'Create' button is missing for association tables of newly created records.

  In case a record was created using the 'Create' button of a foreign key selector, some actions are temporarily missing from the toolbars of its association tables.

• CP-21562 [WCAG 2.0 - 2.1.1][Accessibility] It is impossible to open a record when using the keyboard.

• CP-21619 [WCAG 2.0 - 3.2.1] It is impossible to expand the list of messages at the top-left of the record view, using the keyboard.

  When users press the Tab key on a record form, the cursor is not focused on the Expand button on the list of messages.

• CP-21626 [WCAG 2.0 - 3.2.1] The buttons in the toolbar of the HTML field are not accessible from keyboard.

  When users navigate inside a record form which contains a field of type HTML, none of the user controls of that field are accessible via the keyboard. For instance, the focus cannot be moved on the 'Expand Toolbar' of the HTML field's text box.

• CP-22307 The 'Select' button overlaps the node's icons in the hierarchy selection view, when using the French language.

• CP-22419 The wizard button of the wizard widget is not accessible with keyboard.

• CP-22791 [WCAG 2.0 - 2.1.1] Buttons in a toolbar can be focused with TAB key.

• CP-22794 [WCAG 2.0 - 2.1.1] Control buttons in dual lists can be focused with TAB key.

  These buttons should be contained in a toolbar that will be accessible and that handle the focus management nicely.

  In addition, the first element in each list will be active by default and the buttons acting on those active elements will be active too.

• CP-22898 Opening a work item can fail when a data context variable is too large.

• CP-22899 A record is not automatically selected in the combo box after its creation through the 'Create' button of a foreign key's combo box.

  This happens when the creation form is a user service declared by UserServiceRecordFormFactory#newUserServiceForCreate(ForCreate).

  To overcome this issue, services returned by this method should implement the new API UserServiceForCreate.

• CP-22917 [WCAG 2.0 - 2.2.1] Screen reader does not provide enough info to use dual list component.

• CP-23022 A JavaScript error is thrown on login, when the application is embedded  in a portal.

Users & Permissions

• CP-1876 [29663] @29660 Unclear error when calling Session.isUserInRole() with a null parameter.

  Calling Session.isUserInRole() with a null parameter leads to an unclear 'Unexpected error in request computation.' error. Error management in such a case should be improved.

• CP-21203 A blank workspace appears when accessing the 'Information' service in a dataset as a user with hidden permission.

• CP-22604 An error occurs when an Access Right has a null node default value.

• CP-23099 A procedure executed by 'ebx-system-user' should not trigger any access rules.

  As mentioned in the documentation: "A System session has no associated roles: internal permission resolutions bypass role-based permission rules and more generally any permission rule.".

6.0.5 Closed issues

Core engine

• CP-2477 [30363] Memory leak due to redundant internal objects that remain in the cache.

• CP-21256 Incorrect request result after changing the PK of a table.

• CP-22175 Cannot retrieve occurrences that are deleted between two transactions when using Diff API.

• CP-22296 Converting null internal values to external values throws NullPointerException.

• CP-22302 Inaccurrate similar query limit can cause incomplete clusters.

• CP-22385 The TestHelper API does not close the datasource when shutting down a new repository.

  When a repository is created by using TestHelper.getNewRepository(java.io.File, boolean), the datasource is not closed on shutdown. As a consequence, the datasource related files may remain locked on the operating system even after the repository has been shutdown.

• CP-22454 Inter-database migration from HANA fails.

  Migrating from HANA to any database fails with this message:  Caused by: java.lang.IllegalArgumentException: Unsupported type: 2011

• CP-22482 Cannot run match on small selection when having isNull exclusion.

• CP-22556 TableTrigger handler is called more than expected.

  When a TableTrigger with a *beforeTransactionCommit* or *beforeTransactionCancel* handler is used, these handlers are called more than expected.

• CP-22657 Lookup on PK with null fields causes NullPointerException.

• CP-22790 A ClassCastException occurs while building a parameterized predicate.

  While programmatically building a XPoStringEqualsCaseInsensitive predicate with a parameterized literal value, a ClassCastException occurs.

• CP-22839 An incorrect result is returned when applying an advanced search on the "Label" field with the operator "is empty or null".

  When users apply an advanced search on the "Label" field with the operator "is empty or null" on the "User filters" table, all records with "Label" are returned.

• CP-22877 StackoverflowException during incremental index rebuild in certain conditions.

  The table revision chain required to rebuild an index can be long, driving to a StackoverflowException.

• CP-23004 'No dynamic parameter in position 0' error during matching.

  When matching a selection of records, the process fails with the exception "No dynamic parameter in position 0".

• CP-23016 An exception occurs when rebuilding the indexes of a dataset prior to the compilation of data models.

  An exception is raised when initializing a specific directory that looks up a specific dataset in the repository. 

• CP-23017 Replace method File.lastModified, which is buggy for some JREs version 8.

  If files are changed very quickly (within one second, probably via automated means), it can cause some changes to be missed.

  See [https://bugs.openjdk.java.net/browse/JDK-8177809.]

Data modeling

• CP-3087 [23461] UIBeans on fields of type "osd:password" are not imported when importing an XSD in the data modeler.

• CP-21387 The value context is null in the ConstraintOnTable#toUserDocumentation method.

• CP-22527 The dropdown list displaying the library components should be sorted

Data services

• CP-22634 The REST built-in insert operation used for update must consider the 'byDelta' parameter.

  The PUT update operation for a single record manages the 'byDelta' parameter, but this parameter has no effect when using the POST insert operation with 'updateOrInsert=true' parameter.

Data validation

• CP-22533 Validating a table with fields that only differ by their case leads to a QueryParseException "Column 'ABC' is ambiguous".

• CP-22828 Quotes missing in SQL queries for enumeration constraints.

• CP-22913 NullPointerException when validating a table that references another table with minOccurs > 0 or maxOccurs < unbounded.

• CP-22952 NullPointerException because of leftover threshold validation items after removing a constraint from a data model.

• CP-22984 Validation query causes ClassCastException due to wrong table parameter on FK_AS_STRING function.

Data views - hierarchies

• CP-19499 A leaf node is not moved to the unset node after running the 'Detach from parent' service on the hierarchy view of a child dataset.

  On the hierarchy view of a child dataset, when users run the 'Detach from parent' service on a leaf node, this node is not shown under the unset node, and it disappears.

• CP-22197 Cycle detection on recursive hierarchy takes too long.

• CP-22261 Hierarchy view shows hidden records.

• CP-22331 Hierarchy request does not filter out occulting records.

• CP-22552 An error occurs when applying the prune mode on a hierarchical view with multi-level recursive dimensions.

• CP-22553 An "IllegalArgumentException :field not found" error occurs when using a hierarchy view.

  An "IllegalArgumentException: Field not found" error occurs when using a pruned hierarchy view.

• CP-22846 Branch selection in a hierarchy takes too much time.

  In the dataspace tree of the administration panel, the action 'Select all children' on a branch is too long.

• CP-22848 IllegalArgumentException occurs when selecting in a recursive hierarchy.

  1) If a branch is selected in a recursive hierarchy and another branch is selected, an IllegalArgumentException occurs.

  2) If a branch is selected in a recursive hierarchy, and a single select and unselect is done on another member, an IllegalArgumentException occurs.

• CP-22892 Selecting a branch in the hierarchy of dataspaces takes too long.

• CP-22904 PathAccessException while doing a CSV export of a hierarchy.

• CP-22972 Exception occurs after updating a recursive hierarchy member.

  If there is a combination of filters on the recursive level (link table dependency) and also a user search filter, then the update/close of a hierarchy member generates an exception.

Dataspaces & History

• CP-19345 A blank page is displayed when importing an updated dataspace.

  When users import a dataspace that includes some updates, a blank workspace appears.

Distributed Data Delivery (D3)

• CP-22314 Forbidden to close snapshot, with no reason message.

  When a snapshot corresponds to the current D3 snapshot for a delivery dataspace, the access permission is hidden and no reason message is displayed.

• CP-22351D3Broadcast.getDeliveryModes() returns an empty list if there is no registered replica.

  Even if the delivery mode is defined in each delivery profile, an empty list is returned when the registered replica table is empty.

Documentation

• CP-22369 The documentation should specify that it is not possible to use a workflow launcher in a 'table row' toolbar.

Import / Export

• CP-21561 [WCAG 2.0 - 3.2.1] The input field beside the 'Other' should not be on the TAB path of the focus until the 'Other' value is checked.

  When users press the Tab key on the list of radio buttons of 'Field separator' on the 'Export CSV' screen, the textbox of the 'Other' field in that list is always selected while the previously selected button has been removed.

Queries

• CP-21384 Query on child dataset does not filter out occulting records.

• CP-22493 The combo-box for searching on table reference does not work on a table reference without display pattern, and target table with programmatic label.

• CP-22520 Pagination on linked field by inverse table reference throws an error.

• CP-22565 CyclicMetadataException when opening a perspective.

  CyclicMetadataException when opening a perspective

REST Toolkit

• CP-21779 REST Toolkit application throws NullPointerException during providers registration.

• CP-22290 InaccessibleObjectException on Java 17 LTS.

  The JSON library used by the REST components has pending issues on Java 17.

User interface

• CP-17949 The number display policy should be applied onto messages.

• CP-18798 An error occurs when accessing history from an association targeting a table in another dataset.

• CP-19653 In a paginated tabular view, the grid edit comes back to the first page.

  In a paginated tabular view, when the user navigates several pages and clicks the grid edit option, the first page is displayed.

• CP-21553 [WCAG 2.0 - 2.4.3] Some controls of the date widget are not accessible using the keyboard.

  When pressing the Tab key throughout an EBX Form, some controls of the date widgets get skipped.

• CP-21558 [WCAG 2.0 - 2.1.1] It is impossible to select a service under the 'Actions' menu of a dataset using the keyboard.

• CP-21617 [WCAG 2.0 - 2.1.1][Accessibility] It is impossible to delete an occurrence using the keyboard.

• CP-21618 [WCAG 2.0 - 3.2.1] It is impossible to move a field from the Sortable column to the Sorted column on the Sort criterion pane.

• CP-22064 Erroneous scope of the comparison service in a workflow task.

• CP-22127 Message synchronization issue in User Services.

  In a User Service the validation messages may not be correctly synchronized on submit, leading to fields with actual errors not displayed or fields with obsolete error messages.

• CP-22349 [All browsers] A JavaScript error occurs when expanding a computation rule.

  When users click on the "Expand" button on "Computation rule", a JavaScript error occurs.

• CP-22361 [WCAG 2.0 - 2.4.3] 'Clear' button in radio button group is not accessible from the keyboard.

  When the user presses the Tab key throughout an EBX Form, the 'Clear' button of the radio button group gets skipped.

• CP-22420 The control which sets the default value of a field is not accessible from keyboard.

• CP-22453 The language selector is broken, when using Internet Explorer.

• CP-22650 It is impossible to select a perspective from the header when using the keyboard.

• CP-22662 It is impossible to select a language from the login page.

• CP-22684 Error on association with XPath Expression.

  An error occurs while displaying an association defined with an XPath predicate.

• CP-22808 Upgrade third-party library CKEditor to version 4.17.1.

• CP-22941 NullPointerException when creating load menu item on advanced search.

6.0.4 Closed issues

Core engine

• CP-22618 Migration fails in some conditions when a data model defines replication units

  A data model compilation error occurs regarding replication units, when migrating from 5.9.X to 6.0.3.

• CP-22694 Repository migration fails when a table is both historized and replicated, and a column is disabled for history

Data modeling

• CP-22692 Remove restrictions related to inconsistencies in associations.

  Some restrictions had been set about the definition of associations regarding child dataspaces. That is: when an association is based on an foreign key or a link table in another dataspace, the definition of the association becomes inconsistent in the children of the dataspaces involved in the relationship. In this case, the related datasets are considered in severe error. Since this behavior is considered too strict and creates blocking issues, these changes must be rolled back. 

Data validation

• CP-22738 ClassCastException when validating a table reference filter having both contextual and non-contextual terms

  An fatal error occurs when validating a foreign key constraint with an XPath filter that contains several conditions.

User interface

• CP-19856 The page for selecting backed-up user does not display in some cases.

  If the backed-up user feature is enabled in a custom directory, the page for selecting a backed-up user does not display when using either of:

  * a custom login page,

  * a URL with parameters: example: [http://localhost:8080/ebx/?login=testAdministrator&password=testAdministratorPassword]

  * SSO

Users & Permissions

• CP-22531 The new permission framework is not always thread-safe.

  Under some circumstances, the permission framework can trigger a ConcurrentModificationException.

6.0.3 Closed issues

Core engine

• CP-21787 Upgrade Commons IO 2.5 library to version 2.11.0.

• CP-22086 Clustering queries may return incomplete tuples, when matching with relationships.

  When matching with relationships, if a record does not match the center on the main table matching fields, it is not returned, even it if matches the center on several support tables matching fields.

• CP-22110 An exception is thrown when a foreign key defines a display pattern involving a password node.

  The password field should be prevented from composing the label, since its search strategy is Void.

• CP-22118 Cannot access a dataset in the procedure that created it through a merge.

  After a ProcedureContext.doMergeToParent(AdaptationHome) in a Procedure that creates a dataset, accessing the newly created dataspace in the parent dataspace results in an incorrect behavior:

  * Accessing the dataset metadata throws an assertion error because there is no revision associated with the root record of the header content type.

  * All tables appear empty, even though they had values in the merged dataspace.

• CP-22259 Clustering with the same data can lead to differing results.

  The matching results can vary for the same data and configuration in the following cases: 

  * when using the TeseLevenshtein search strategy, and the column contains more than 50 words within the distance of the matched word 

  * to a minor extent, when the frequency is very different for some terms in a column

• CP-22293 An error occurs during an archive import when a model defines an association to another dataspace.

  An errors occurs when importing an archive into a dataspace if the archive contains a datasets that uses a data model that contains an association refering to a table in another dataspace.

• CP-22298 NullPointerException when displaying the 'Module and Schema' page, if a subscriber description is null.

• CP-22306 A blank page appears when filtering on the "URI" field with "contains whole word (case sensitive)" or "contains whole word".

Data modeling

• CP-21150 [IE11] 'No item found' message is displayed under drop-down list on 'Search strategy'.

  On browser IE11, when the user clicks on the drop-down list button on 'Search Strategy' on 'Assigned search strategies' creation for the first time, the message 'No item found' is displayed.

• CP-22077 Control policy on enumerations is ignored when importing an XSD in the DMA

• CP-22225 Deprecated properties of table reference cannot be used.

  The data model does not compile, when declaring a table reference using the deprecated properties: displayKey or labelPaths .

Data services

• CP-18639 API sometimes returns the wrong permission for dataset root node.

  When there is not exactly one access right for a user in a dataset, the API may return a wrong permission for the dataset root node.

• CP-22117 Badly formatted JSON in the body of a REST built-in request causes a server error.

  A 500 response error is returned when the request body contains a badly formatted JSON. The response should be a 400 with a significant message that describes the error. 

• CP-22368 The dataspace tree view is unavailable when a reverse proxy is activated.

  When a reverse proxy is activated, the dataspace tree view is not displayed, and generates browser connection errors.

Data views

• CP-21981 An error occurs when combining the quick search and validation on a table.

  An error occurs when users apply a quick search or text search on the advanced search pane, and run some services on tables and validation reports.

• CP-22104 An unnecessary leave confirmation pop-up displays when deleting a saved filter.

Data workflows

• CP-3594 [22558] The mail content generated by the built-in "Send an email' service may be inserted twice for a locale.

• CP-17424 The preview on the 'Export settings' pop-up when exporting workflow diagram as PDF displays poorly.

  When users export the workflow diagram as PDF, the preview on 'Export settings' pop-up is not fully shown and the 'x' button always has a round border.

• CP-19462 The limit size for profile name in workflow history is not documented.

• CP-22014 [All browsers] An error occurs when deleting a node which is a redirection link from another step.

• CP-22148 [IE11] Users cannot delete a step on a workflow diagram by pressing the 'Delete' button on the keyboard.

  When the user selects one step in the workflow model, then presses the 'Delete' button on the keyboard, the step is not deleted.

Distributed Data Delivery (D3)

• CP-22284 D3 replica node throws IllegalStateException when registering a broadcast trigger on startup.

  The D3NodeAsSlaveOnStartup.registerBroadcastTrigger(AdaptationHome, Class<? extends D3ReceiveBroadcastTrigger>) always throws IllegalStateException with message "This object is not in update mode.".

Import / Export

• CP-3002 [27044] @27042 An exception occurs while importing some special chars using CSV build-in import.

  An exception occurs while importing some special chars using CSV build-in import.

• CP-3442 [27043] @27042 An exception occurs while importing some special characters, using the CSV built-in import.

Infrastructure

• CP-21115 Validation errors are signalled for obsolete addons.

  Validation errors appear in the administration panel when obsolete addon records are present in 'Registered add-ons' table or 'Task Scheduler' tables.

• CP-22359 File i18n_xx.zip of the 'ebx-manager' webapp is corrupted.

Java EE integration

• CP-21852 Upgrade JDOM 1.0 library to version 2.0.6

Messages & i18n

• CP-22322 Build and template language pack (i18n_xx.zip) should not contain *meta.mxml files.

Queries

• CP-22070 A request with JOIN + FILTER + ORDER BY LeftHandSide's FK field does not exploit Lucene index.

• CP-22189 NullPointerException when querying a multi-valued field after a schema evolution.

• CP-22221 ClassCastException when processing a TableRef that points to sub-field inside a group.

• CP-22277 Incorrect field access when translating a predicate with 3 (or more) chained FKs.

• CP-22297 An exception is thrown when searching a label containing a foreign key targeting an external dataset.

User interface

• CP-21896 The 'Sorted by relevancy' icon is sometimes displayed even for non-concerned operators.

  It should only be displayed in the presence of quick-search, or search operators.

• CP-22011 A blank popup is displayed when clicking on the 'Create a record' button on the Permissions page of the 'Dataspaces' dataset under Administration.

  When users access the 'Access rights by profile' page of the 'Dataspaces' dataset under Administration then click on the 'Create a record' button, a blank popup is displayed.

• CP-22041 The URL computing documentation must mention the data lineage configuration when using a reverse proxy.

  When using a reverse proxy (particularly in https), it is necessary to configure the data lineage property. This should be mentioned in the URL computing configuration.

• CP-22073 In custom forms, the label of a field is displayed even if it is hidden by permissions.

  A row defined in a custom form but which is hidden by permission should not be displayed. Currently, its label is still displayed.

• CP-22083 Focused button in dataspace tree has its shadow cut off.

• CP-22094 When a custom form is saved, its name is changed to "name".

  On opening a custom form, its displayed name is "name", and it cannot be changed.

  On saving, a new custom form named "name" is created.

  The previous form remains and is not updated.

• CP-22130 An empty value for workflow action in ebx.properties hides "Data workflows" and "Workflow models".

  As the default value for workflow activation is "true", the "Data workflows" and "Workflow models" categories should be available, even if this property is not set.

• CP-22214 Error when calling UserTaskBeforeWorkItemCompletionContext.reportMessage.

  Clicking on the error should not have any effect, but in this case, it displays a blank page with an unexpected error.

• CP-22223 Displayed dataspace snapshots are not sorted by descending creation date.

• CP-22227 In a dataspace information page, when dataspace has no snapshot, an empty table is displayed instead of a message.

• CP-22343 Left pane does not display correctly when using HTTPS.

Users & Permissions

• CP-19346 A non-admin user cannot delete a dataset that has a fatal error.

  When a non-admin user creates a dataset that contains a fatal error, the user has no permission to delete this dataset.

6.0.2 Closed issues

Core engine

• CP-18301 TablePagedView.getFullSizeEstimation() takes too long.

• CP-21015 Poor performance of the workflow history purge when using a PostgreSQL database.

• CP-21713 AlreadyClosedException when rebuilding index.

• CP-21732 Performance of auto-increments degrades over time.

• CP-21752 Fatal error "LogicalCorrelate cannot be cast to Join" when validating dma/sample-unit4.xsd.

• CP-21758 Error on inter-database migration for mapped table with field of type osd:dataspaceKey or osd:datasetName

• CP-21849 Poor performance of history queries when exporting or calling the REST 'select' operation.

• CP-21856 The default constraint on primary key should be blocking for all operations

  A default pattern is defined on primary key fields if no other patterns are defined in the data model. This default constraint is checked in the user interface, and then blocks any erroneous value but is not blocking on some operations (e.g. during an import). Invalid values can be set during these operations but cannot be corrected.

  This default pattern should be blocking for all operations.

• CP-21870 On migration from EBX® version 5.9, evolutions on indexes of relational tables should be disabled.

  During the migration to v6, DDL statements are issued to suppress the non-pk indexes of relational tables. This is useless, since these tables will be dropped (and their content migrated to semantic tables).

• CP-21871 Poor performance of the history purge when using a PostgreSQL database (technical tables)

  In cases where the 'session' history table is very large, it can cause performance issues during the purge on PostgreSQL.

• CP-21875 Corrupted index because of missing files.

  A com.onwbp.org.apache.lucene.index.CorruptIndexException caused by a java.nio.file.NoSuchFileException is thrown when opening the index.

• CP-21891 Migration to v6 is erroneous if a dataspace holds table revisions belonging to a dataset that does not exist.

• CP-21930 ClassNotFoundException when using the DirectoryDefaultRecoverFactory directory factory.

  A ClassNotFoundException occurs when using com.orchestranetworks.service.directory.DirectoryDefaultRecoverFactory directory factory. This class is missing from the common library jar.

• CP-21971 Missing rows when performing relational matching without criteria on the main table.

• CP-22005 The migration of relational tables fails on PostgreSQL when the commit threshold is reached.

  When using PostgreSQL, a relational table with more than 260,000 records cannot be migrated to v6.

  An exception is thrown: com.orchestranetworks.instance.RepositoryAccessException: org.postgresql.util.PSQLException: ERROR: portal "C_xxx" does not exist

• CP-22023 When computing a single cluster, filtering merged records takes too long.

• CP-22053 When computing a single cluster, filtering excluded records takes too long.

Data modeling

• CP-17966 Inconsistency in association definition when creating a child dataspace.

  Consider a table in a dataspace "A" that has a foreign key to another table in a dataspace "B". The table in dataspace "B" has an association based on that foreign key.

  When creating a child dataspace of B, its association definition becomes inconsistent because it does not comply to the foreign key definition in A. Indeed, that foreign key defines that its target dataspace is B (and not any other dataspace).

  This leads to an unexpected behaviour where all actions made on the association node of a record in the child dataspace are reflected to the association node of the same record in the parent dataspace (B).

  Also, if a record is created in the association of a record in the child dataspace, a validation error message is shown on the record's foreign key because the corresponding record in B is not found (because it is in the child dataspace).

• CP-18103 NullPointerException when publishing a data model with a wrong XPath predicate.

• CP-21622 [WCAG 2.0 - 4.1.2.][Accessibility] The Preview button has no accessible name.

  When clicking on the Preview of some fields, the button has empty accessible name. The preview button must be focusable with a keyboard like the null button, but it must also tell what field will be opened for preview. 

• CP-21623 [WCAG 2.0 - 4.1.2.][Accessibility] The Add an occurrence button has no accessible name.

  When clicking on the (+) button in a multi-valued field, the (+) button has no accessible name.

• CP-21759 In advanced search, on a foreign key field, no results are shown when entering an incomplete word.

  In the advanced search, for a foreign key field, on user input, the drop-down displays results only if a full word is entered.

• CP-21854 Searching on linked fields with TESE causes an error.

• CP-21910 Wrong behavior on search if a linked field refers a foreign key field and redefines the display pattern.

• CP-21959 The validation report is not updated when the definition of association or selection field evolves.

  Currently, the signature of a selection or association field only includes the Path of the node + XSD tag name. Changing the other properties of the association or selection field does not impact the signature of the constraint. Therefore, the validation report will not be updated.

• CP-21973 Data model extensions missing when exporting DMA datasets in an archive.

  Content of data model's extensions are not included when exporting an archive from the DMA branch. The extensions are defined in separate datasets with auto generated names. Therefore it is not possible to indicate explicitly that these datasets must be exported. So these datasets should be automatically exported when creating an archive from the DMA branch.

Data services

• CP-21637 The generated WSDL for getChangesOnDataSet and replicationRefresh operations are incorrect for certain conditions.

  The generated WSDL can contain invalid XML characters (such as space, slash or comma), if the model is packaged in a 'war' file and the webapp is deployed in 'unpack wars' mode. This case happens when the servlet container is unable to translate the given virtual path to a path of the data model.

• CP-21814 REST count operation on association field with :count syntax returns 404.

• CP-21830 The documentation does not mention the limitation for procedure context with the REST built-in JSON compact format.

• CP-21864 Error on UI dataspace tree view.

  If 'Data lineage > Base URL' is not set to the appropriate value: when the dataspace tree view is shown, it displays 'An unexpected error occurred. Please contact your administrator.'

Data validation

• CP-21756 Reaching the validation threshold for table references throws an exception.

• CP-21757 An exception occuring during the validation does not appear in the validation report.

• CP-21782 A NullPointerException occurs when requesting a validation report after deleting a node in the data model.

• CP-21841 An error occurs when purging validation reports.

• CP-22027 The "Validate" service is missing on data spaces containing a dataset with a fatal error.

Data views

• CP-21737 [All browsers] The 'View' button does not work on the 'Permission' table under the Data/ Administration area.

  When users access the 'Permissions' table of a dataset under the Data/ Administration area, then click on the 'View' button, nothing happens.

Data views - hierarchies

• CP-16837 CSV export does not work in hierarchy views.

  When a user exports a CSV file from a node in a hierarchy view, the file has only the title, with no contents included.

• CP-17092 An error occurs after applying 'Enable pruned mode' on a hierarchy view.

  When users apply 'Enable pruned mode' on a hierarchy view, an error or an unexpected behavior occurs after users execute some services on a node.

• CP-17807 Hierarchy nodes and their children are not shown when applying a pruned hierarchy based on a simple-foreign-key recursive link.

  When the user applies a hierarchy based on a recursive link that is a simple foreign key with 'pruned mode' set to 'yes': hierarchy nodes that have children are not shown

• CP-18179 The loading screen persists when users access a hierarchy view through the 'Selector' button.

  If a foreign key has a filter, when the user opens its selector and applies a recursive hierarchy view, then the loading screen persists.

• CP-21672 When applying a quick search on a hierarchy view with the 'Locale' field, an incorrect (unfiltered) result is returned.

• CP-21727 A error occurs when searching a recursive hierarchy for special characters.

  A error occurs when applying 'Find in hierarchy' in a recursive hierarchy, and inputting one of these characters: '*', '+', '-','='

• CP-21835 Hierarchy pagination on date or enumerable page does not work properly.

  If root members are enumerable, i.e. enumeration or date and pagination is needed, the pagination will not work. 

• CP-22009 The 'Delete' service is available under the Action menu of a hierarchy view, even when users have no permission to delete.

Data workflows

• CP-19193 [IE11] An abnormal field is displayed briefly when defining a Sub-workflows invocation.

• CP-20507 A fatal error occurs when executing a user service with an inconsistent selection.

  A fatal error occurs when executing a user service with an inconsistent selection. This happens when the parameters  of a WebComponent are not configured correctly. 

  For instance, when a work item of a workflow uses a custom user service, and the dataspace parameter is missing.

• CP-21149 A non-admin user can close a snapshot without having sufficient permission.

  When an admin user sets "Forbidden" on 'Close snapshot' for a non-admin user, this user can still close a snapshot on an unpublished workflow model.

• CP-21710 The error message when relinking the two condition's links to the same step is not clear.

  When one tries to move a condition's link to the same step as the other link from the same condition, the message box displays an HTTP error. (This is in the case where the ending step was the "final step")

• CP-21740 An exported PDF file of a workflow model is displayed in landscape instead of portrait as expected.

• CP-21742 An error message displays when applying a quick search on the 'Work items' table under Administration.

• CP-21916 Wrong display for 'Foreground sub-workflow' under the advanced search's criterion.

  When users add an advanced search's criterion with "Foreground sub-workflow", its label displays as "isForegroundSubworkflow".

Dataspaces & History

• CP-16707 Cannot access the history record of a merged dataspace.

  From the table history view, when double-clicking a history record of a merged dataspace, an error message is displayed: 'No record found for the predicate [...]'.

Documentation

• CP-22018 The installation note for JBOSS EAP 7.1 specifies the use of an XA data source, instead of non-XA

Queries

• CP-21865 Quick search on a linked field whose relationship is a foreign key linked field causes CannotPlanException.

• CP-21883 Search on tableRef that points to another tabelRef with a display pattern causes CannotPlanException.

• CP-22031 JOIN + ORDER BY LeftHandSide's FK field does not exploit Lucene index.

• CP-22054 AssertionError on Tuple.get(int) with nested complex nodes and computed values.

• CP-22071 LuceneFieldScan retrieves wrong score value.

Scripting

• CP-21872 EBX Script log functions do not set the line number.

  Functions from unit "core.log" do not set the line number.

User interface

• CP-17343 An authentication error is not displayed correctly in the default login page

  When the authenticateUserFromLoginPassword() method of a custom directory throws an AuthenticationException, a "HTTP error 500" message is displayed in the page displayed to the user.

  The message specified when creating in the AuthenticationException should be displayed instead of the meaningless above message.

• CP-18121 A blank pop-up displays when clicking 'Manage recommended views' or 'Publish' at the time a long query is being executed.

  When a long query is being executed on a table, if users click on 'Manage recommended views' or 'Publish', a blank pop-up is displayed. These services should be disabled like others services on 'View' menu.

• CP-19395 [Firefox 68ESR] The last record cannot be closed when users open more than 10 records on the breadcrumb.

  When users open more than 10 records on the breadcrumb, the last one cannot be closed by clicking on the 'x' button or clicking on the previous record.

• CP-21511 Switching between perspectives does not work.

• CP-21575 [WCAG 2.0 - 1.1.1] The user avatar does not have alternative text.

  When users hover/click on the avatar on the user pane, the alternative text of that image is missing.

• CP-21665 [IE11] A JavaScript occurs when deleting the saved filter.

• CP-21673 Complex field incorrectly shows with the "unsearchable icon" in the UI.

  When users apply a quick search in a value under 'Group' that re-uses a complex type, the results are still returned even though when 'Group' is excluded from search.

• CP-21868 UIButtonSpec#setTooltip(UserMessage) behaves wrongly for icon buttons.

  UIButtonSpec#setTooltip(UserMessage) does not behave as described in its documentation when set on a button with UIButtonLayout.ICON_ONLY.

  In this case, the displayed tooltip should use the label if provided.

• CP-21886 The warning 'Changes will be lost...' is not displayed in the presence of validation errors

  After submitting a form that raises a validation error, no warning that changes will be lost is displayed when quitting the form.

• CP-22002 A saved filter is not displayed on the 'Access rights by profile' table.

  When users access the 'Access rights by profile' table and save a filter, the saved filter is not displayed under the 'Load' button.

Users & Permissions

• CP-20536 The owner of a dataset does not have full permission management rights.

  When the owner of a dataset has read-only access rights in the container dataspace, this owner can edit/delete the permission but cannot create a permission for this dataset.

• CP-21288 A NullPointerException can occur when using the dataspace navigation tree.

  When a dataspace permission record is targeting a missing dataspace, a NPE occurs when displaying the dataspace navigation widget. The reproduction steps are unclear.

• CP-21383 The "Close snapshot" action permission is not correctly computed.

  When disabling the "Close snapshot" action permission using a user-defined access permission, the resolved value is still "enabled".

6.0.1 Closed issues

Core engine

• CP-17081 Snapshot and its parent have different content when recreating a dataset.

  In some cases, a created snapshot might be empty. This happens if the following steps are executed:

  - first procedure: create a dataset and a row in the dataset

  - second procedure: delete the dataset, create a dataset with the same name and schema, and call getTable(...)

• CP-17707 A VALUE_FUNCTION_ACCESSOR exception occurs when searching with @default search template.

• CP-18346 QuickSearch API does not highlight decimal field.

• CP-18378NullPointerException when paginating after sort by FK (and referenced record does not exist).

  On tabular view, when sorting by a FK column, if the FK record does not exist, the sort "does nothing".

  When moving to next page, we get a NullPointerException.

• CP-18402 Filter 'osd:label' is ignored on fields of type 'Locale'.

• CP-18903 An error occurs when applying quick search on some tables of under the 'Administration' area.

• CP-20051 When duplicating a dataset, the newly input label is not saved.

  When duplicating a dataset, the input label is not saved.

  Then, the duplicated dataset can display either the technical name or the label of the origin dataset.

• CP-20133NullPointerException when importing an archive file.

• CP-20398 Comparing records on the 'History' table causes a blank page.

  When users compare 2 records on the 'History' table in a dataset, a blank page is displayed.

• CP-20772 [Relation Matching] ClusterProducer returns no relation data causing wrong decision tree results.

• CP-20783 Quick search does not operate on the label of a 'Boolean' field.

  When users input a value for a 'Boolean' field on a table, then apply quick search, an incorrect result is returned.

• CP-20813 An error occurs when searching a long piece of text on the Quick search bar.

  A TooManyClauses error may occur, for an input with a large number of words (tokens) and/or a table with many columns.

• CP-20851 Modification of facet name should impact the persisted meta model

  For a mapped field, when the name of a facet is changed (for example, from xs:length to xs:maxLength), the description of the facet is not updated in the FCT database table. As a consequence, all following data model compilations detect a change and uselessly issue a DDL.

• CP-20882 Inter-database migration fails if the source is SQL Server

  If the source is SQL Server, the inter-database migration fails while migrating sequences.

• CP-20894 NullPointerException when importing an archive, if the module or schema is unavailable.

• CP-20923 Duplicate key exception when importing archive.

• CP-20930 NullPointerException when invoking TableViewFactory.findCustomViewFromPublicationName.

• CP-21058 In case of a cycle in field inheritance, the result page cannot be displayed.

  When users create an inheritance cycle on an inherited field of a record, this causes the whole view page to be in error (not allowing to correct the cycle).

  The record should be displayed, with an error message in place of the value of the inherited field.

• CP-21075 NullPointerException when importing archive in D3.

• CP-21106 The number of historized tables with a similar name in a repository is limited to 256.

• CP-21120 An exception is thrown when a foreign key defines a display pattern that includes a non-sortable node.

• CP-21187 LxtCompositeSearchStrategy is missing a method to retrieve all strategies assigned to a node (with duplicates).

  In the case a node has been assigned duplicate strategies in different search templates, TNode.getSearchStrategies().getSearchTemplateKeys() does not contain all the templates where the duplicate strategies have been set. 

• CP-21269 Fields under a list are not searchable.

• CP-21273 Index rebuild ends with InvalidProtocolBufferException.

  An InvalidProtocolBufferException is thrown during index rebuild (possibly also while performing a comparison) if the table includes a ValueFunction or any field access that requires a lookup on another table.

• CP-21279 Starting on v6 fails after an inter-database migration.

  Startup fails with a "This repository has last been run on an older version" error after an inter-database migration.

• CP-21292 Cannot migrate more than 2559745 records from a relational table to EBX®6.

• CP-21293 Quick search does not work on linked fields.

• CP-21375 VlnValueStates in dataset headers cannot be migrated.

  Migrating a dataset with a VlnValueState as the value for creator/last user causes a ClassCastException

• CP-21389 Inconsistent default values in workflow administration.

• CP-21421 Improve logging and exception management in serialization layers

• CP-21440 Performance issue with 'Mass delete' on a very large selection.

  The 'Mass delete' process takes very long if millions of records are selected.

• CP-21478 ClassCastException while replicating table with a foreign key to a table with composite primary key.

• CP-21482 Migration to v6 forces a full refresh of the replica tables.

• CP-21520 Avoid SORT/GROUP BY in query plan when clustering without relationship

• CP-21530 Wrong limit value when clustering

  When using RelBuilder#limit, Calcite automatically increases the limit by one. If Integer.MAX_VALUE is used as the limit value, Calcite computes Integer.MAX_VALUE+1 which overflows to Integer.MIN_VALUE.

• CP-21531 Improve management of large clusters

  Some datasets may contain records with a huge number of duplicates. This leads to a huge cluster of similar records, then to a lot of record pairs to be compared by the decision tree. Most of the time is then spent applying the decision tree on record pairs. 

  * In this case, the top-scored tuples in the cluster are the duplicates, and have the same score as the center. To fix this, deactivating the computation of all record pairs for these records would greatly reduce the number of record pairs sent to the decision tree, but still generate enough comparisons to generate a group with all these duplicates. 

  * The time to iterate the tuples in a cluster is longer than usual, it delays the exclusion of records. To fix this, the exclusion can be done before this iteration. 

  Note: also added a log to identify when a huge cluster is encountered. It still slows down the process a bit, and there is probably some action to handle those records directly in the data, outside of the matching process.

• CP-21576 Performance issue when checking whether a dataset has severe errors.

• CP-21596 Wrong serialization descriptor for non-stored field with transitive tableRef.

  If a field defines both:

  * a valueFunction/computationRule (i.e. is not a value-stored field)

  * a transitive tableRef

  a ClassCastException occurs during serialization.

• CP-21655 Increase the limit when fetching similar records in the matching process.

  When matching a single record, the number of candidates was limited to 100. This limit should be increased to 2000. 

Data modeling

• CP-1118 [35313] Schema compilation on 'Refresh all updated schemas' does not report errors in the UI.

  If a schema compilation ends in error on action 'Refresh all updated schemas', the error is not reported in the UI. Only an informative message is displayed, letting the user think that all went fine.

• CP-20958 A JavaScript error prevents from publishing the data model when an extension has messages.

• CP-20989 A cryptic error message is shown when publishing a data model which contains a search strategy with an unsupported language key.

  The displayed error contains an HTML element.

• CP-21119 The list of the impacted datasets is incorrect after changing the PK of one of the tables in the datamodel.

• CP-21148 An unfriendly error message in the 'Custom search strategies' creation form.

  On the 'Custom search strategies' creation form, when users do not define the value for the 'Name' field which is set as a mandatory field, and save record, an unfriendly error message gets displayed on 'Built-in strategy' field.

• CP-21255 SqlChecker error with data models if they have circular dependencies, including one data model in error

• CP-21316 Validating the 'Strategy' association table should throw an error if a strategy is associated with a linked field.

  Since a linked field is a virtual node and uses the strategy associated with the target node, it should not be possible to associate a search strategy with it.

• CP-21327 Display a wait page during the creation of the data model.

Data services

• CP-20629 The documentation for RESTEncodingHelper API is not clear.

• CP-20911 NPE occurs when a REST built-in service selects a history table with a foreign key field to a non-historized table.

  A NullPointerException occurs when the REST built-in service is used to select the content of a history table containing a foreign key field to a non-historized table.

• CP-21044 The query parameter 'ebx-indent' has no effect on REST built-in operations.

  When the 'ebx-indent' query parameter is set to false, the JSON content is always indented. All the built-in REST operations are concerned.

• CP-21103 REST URL in operation response must better manage 'disableRedirectionToLastBroadcast' parameter.

  When the request is applied on D3 dataspace.

  1. If 'disableRedirectionToLastBroadcast' parameter is used in request, the URL in JSON response must propagate it.

  2. If D3 redirection to a snapshot is activated, the URL in JSON response must keep the original dataspace parameter.

• CP-21324 Correct level and location of the built-in logs for REST data services.

  REST dataservices builtin logs are not in same level as 5.9.x WARN in INFO and INFO in DEBUG level. Those logs are present in restServices instead of dataServices category.

• CP-21510 Error while fetching history with built-in REST service with a tableRef to a different dataset/dataspace.

  The error "Specified context must be in the same data model as the field node holding this foreign key constraint" occurs while querying history, with a REST select operation for a table that contains a foreign key a to table belonging to a distinct dataset/dataspace..

Data validation

• CP-20752 Slow response for validation search, on a table with numerous errors.

• CP-20910 IllegalArgumentException when modifying a record after applying a validation search to the table.

• CP-21035 Parameter names not supported by a search strategy do not raise an error in the validation report.

  When using a parameter name not supported by a search strategy, the UI rejects the parameter; but when compiling the model, there is no error.

• CP-21067 The number of validation messages on the navigation pane is inconsistent.

  When users access tables in the dataset, the number of validation messages on the navigation pane changes continuously.

• CP-21068 The validation reports are not duplicated when creating a new child dataset.

  When users create a child dataset from its parent, the validation reports are not displayed on that child.

• CP-21227 IllegalStateException occurs after modifying the primary key of a table referenced by other tables.

• CP-21240 CalciteContextException: Column '...' not found in table '...' during a validation.

Data views

• CP-2663 [25608] Performance issue on hierarchy when users select all records in view, based on simple join/ ternary simple join model.

  On a join table with 50.000 records, it takes a long time to display a hierarchy view based on simple join/ ternary simple join model, if users select all records in the view.

• CP-18249 Errors occur when setting a filter on the 'Locale/URI' field on a custom tabular view.

  When users set a filter {matches/matches (case sensitive)/contains whole word/contains whole word (case sensitive)} on the 'Local/URI' field of a custom tabular view then apply it, an error occurs on the table.

  When users set other filter criteria on the 'Local/URI' field of a custom tabular view then apply it, no records are returned.

• CP-19358 An incorrect result is returned when searching with the 'matches' operator.

  When users add a criterion using the 'matches' operator, and pattern

  [a-z]*[\d]*

  or

  [a-z]*[\D]*

  the system returns an incorrect result.

• CP-20414 Differing results are returned when applying a filter on a hierarchy view and a tabular view.

  When users apply a quick search on a hierarchy view and on a tabular view (same data, same table) with the same criteria, the results are not the same.

• CP-20706 Exporting data from an association with custom view should not export all fields from association table

  Considering an association link (definition: By foreign key) with a custom view: while exporting data from the associated table, all table columns are getting exported with the default column position. Usually, the export contains only the columns of the current view.

• CP-20821 Performance issue when filtering on multi-valued field.

  When users apply an advanced search on a field under a group re-using complex type, the search takes a long time.

Data views - hierarchies

• CP-3851 [22500] The loading page notification persists when selecting all records in hierarchy view of a large table.

• CP-17239 The loading screen persists when applying a 'Search' filter, then applying 'Find in hierarchy'.

• CP-17452 A blank workspace is shown when selecting records on a hierarchy view that has large data.

  When users select records by using 'All' from 'Select and display' > 'Select records', the loading screen persists for quite a while, then a blank workspace displays.

• CP-18339 IllegalArgumentException when searching on ragged hierarchy.

• CP-18570 IllegalArgumentException while displaying hierarchy with composite links.

• CP-19145 A fatal error occurs when clicking 'Select all children' in the hierarchy view.

  When users add a new primary key to the model, then click  'Select all children' in the hierarchy view, a fatal error occurs.

• CP-19164 An incorrect result is displayed when applying a filtered hierarchy view on a target table which is also a join table.

• CP-19975 When deleting a record and applying the 'Select records' service, the records in the hierarchy view disappear.

  If, after deleting a record in the hierarchy view, users click on the 'All' or 'None' selection, the records on the table disappear.

  This bug also occurs when deleting a record and applying a value on the 'Find in hierarchy'.

• CP-20095 [All browsers] The filter icon is not displayed on the filtered node after applying an advanced search on the filtered hierarchy view.

  When users apply an advanced search on the filtered hierarchy view, the filter icon is not displayed in the filtered node, and a blank search pane is displayed.

• CP-20225 The number of selected records is not updated when a record is deleted on the hierarchy view.

• CP-20410 A blank workspace is displayed when applying the text search, then selecting all records on a hierarchy view that has large data.

• CP-20418 An error message is displayed when creating a hierarchy view that contains an HTML field on its label.

• CP-20677 An incorrect result is displayed when applying quick search on a 'Date' field in hierarchy view of a child dataset.

• CP-20678 Hierarchy search in error on a recursive link with a contextual date field link.

  When users create and apply a hierarchy view, then expand an unset node, an error message is displayed.

• CP-20679 A fatal error occurs when choosing a filtered record, then selecting all records in the hierarchy view of a child dataset.

  In the hierarchy view of a child dataset, when users apply quick search, tick on the filtered record, then select all records, a fatal error occurs.

• CP-20681 A fatal error occurs when selecting 'None' on hierarchy view of child dataset.

  In the hierarchy view of a child dataset, when users apply quick search, then choose all records, and finally choose 'None' selection, a fatal error occurs.

• CP-20693 A white page is displayed when selecting all records in the hierarchy view of a child dataset.

• CP-20700 Hierarchy: The elements returned by 'Find' are not filtered by the search.

• CP-20839 An error occurs when applying a hierarchy view on the "Activations" table.

  On the "Activations" table of the "Workflow launchers" under the "Administrations",  an error occurs when applying a hierarchy view with dimension "Tables/Activations". 

• CP-21022 An error occurs when applying search/ quick search on a hierarchy view based on a Boolean field.

• CP-21132 No children found if the next target level is recursive.

• CP-21220 An error message is displayed when creating a hierarchy view on a table whose label involves a field with certain search strategies.

  Error when users create a hierarchy view on a table, if it has a foreign key to a table whose label refers to a field with search strategy= Text/Excluded from search.

• CP-21233 Hierarchy Pagination on source fails on a binary-link dependency ordered by link.

  Hierarchy Pagination on source fails when searching on a binary-link dependency ordered by link.

   - 'next page' does nothing

   - 'previous page' (after a last page) fails.

• CP-21246 An exception occurs when creating a record under an orphan parent in the hierarchy view.

  An exception occurs when creating a record under an orphan parent in the hierarchy view

• CP-21290 'Find' fails on hierarchy with ternary recursive dependency.

  'Find' on hierarchy with a dimension such as 'active -> employee -> employee' fails with a ConversionException

• CP-21381 A member of a recursive hierarchy tree cannot also be the child of parent level.

  For a dimension as s -> t -> t, a member t of the recursive tree cannot be child of a s member, except if it is a root of the t tree.

• CP-21395 Hierarchy error when its dimension is based on a chain of composite foreign key dependencies.

• CP-21508 Hierarchy on mapped tables displays the wrong results.

  In dataset Administration -> History and Logs, define a hierarchy on the table  'Repository connection log' with the dimension Mode -> Repository. 

  If we open any Mode member, all the records are displayed.

Data workflows

• CP-2239 [31108] The data context variable standing for output parameter of validation service is not displayed.

  After defining a data context variable for 'Output parameters' of the service '[built-in] Validate a dataspace, a snapshot or a dataset': this variable is not displayed.

• CP-17206 The PDF export pop-up of 'Workflow diagram' is poorly displayed.

  In 'Workflow diagram', the Export settings (PDF) form moves up when users open the 'Format' drop-down list, and the pane that contains the diagram comes without a right border.

• CP-19440 When opening a sub-workflow, its parent workflow item is not displayed in breadcrumb.

• CP-19752 When using the 'Duplicate' service in a workflow user task, primary key fields that are also foreign keys are not editable.

• CP-19882 The workflow model does not comply with the modeling default values.

  When users set the default values of the 'Activate quick launch' and 'Automatically open the first step' as 'Yes', then create a workflow model: the default values of these fields in this model are 'No'.

• CP-20221 Improve speed performance of workflow write actions.

• CP-20747 'Find in hierarchy' bar should not be displayed on the workflow model.

  When users access the hierarchy view on the workflow model, then apply the 'Find in hierarchy', no matching record is highlighted and the 'Find in hierarchy' pane is closed.

• CP-20767 Unwanted redirection to the inbox, when closing a snapshot through the 'Publication history' service.

• CP-21143 A wrong message is displayed when running the 'Clean interaction table' service on the 'Interaction' table of the 'Interaction' dataset under Administration.

• CP-21218 The priority is not updated on 'Data workflow' when users set the default priority on 'Administration'.

• CP-21338 Default value for workflow activation should be 'true '.

  The default value should be 'true', since workflow is most commonly used.

• CP-21390 Spelling error in a message for workflows

  The following French message for workflows contains some spelling errors :

  Impossible d'ouvrir le bon de travail à cause d'une erreur inatendue. Veuillez contater votre administrateur.

  It should be instead :

  Impossible d'ouvrir le bon de travail à cause d'une erreur inattendue. Veuillez contacter votre administrateur.

Dataspaces & History

• CP-19944NullPointerException when importing an archive to a data space.

• CP-19987 The first page of the dataspace pagination is displayed after clicking on the breadcrumb of a dataspace.

  When users click on the breadcrumb of a dataspace, they are re-directed to the first page of the pagination. 

• CP-20788 The "Workflow launcher" service behaves abnormally on the History table.

  When users click on the "Workflow launcher" service under the "Actions" menu on the History table,  the History table is closed and the data table is displayed.

• CP-20796 The "Sort by relevancy" service should not be offered on the History table.

• CP-21017 Error when several threads concurrently access the history of a table.

Distributed Data Delivery (D3)

• CP-21433 Resetting all schemas can deadlock with D3 event notifications.

Infrastructure

• CP-21380 Clarify the installation notes for LZ4.

  The application servers' installation note are not clear enough for customers wishing to install the LZ4 library natively.

Queries

• CP-21005 ClassCastException when querying OResource fields: String cannot be cast to OResource.

• CP-21169 SQL query with AVG function fails with CannotPlanException.

• CP-21178 Inconsistent entries for default value of Query Cache entryExpirationInMinutes.

• CP-21277 Calcite metadata rowCount returns incorrect value for LuceneToEnumerableCount.

• CP-21337 No results returned for quick search on a 'Resource' field.

• CP-21340 Sorting by a 'Resource' field causes exception when paginating.

• CP-21367 Quick search does not operate on the label of a 'Locale' field.

  Quick search does not operate on the label of a 'Locale' field.

• CP-21529 LEFT/RIGHT JOIN with OFFSET returns wrong results.

  The problem can be reproduced with e.g.

  SELECT n.n_nationkey, n.n_name, r.r_name

  FROM nation n LEFT JOIN region r

  ON n.n_regionkey.r_regionkey = r.r_regionkey

  ORDER BY n.n_name OFFSET 1

  This query returns an incorrect result, since the OFFSET is applied more than once, probably due to a bug in Calcite's SortJoinTransposeRule (see CALCITE-4617)

• CP-21676 Wrong translation for osd:label filter on locale field.

• CP-21708 Repeated access to Lucene indexes slows down the diff between the semantic table and the replica.

  If a full diff with the replica has to be computed (for example: when activating replication), the repeated access to the persistent Lucene indexes takes very long, for tables with millions of records.

REST Toolkit

• CP-19778 Exception when requesting REST with a custom servlet path.

  If a custom servlet path configuration is set: when sending a REST request generated from the UIResourceLocator#getURLForRest API, a ServletException occurs on the server side.

  On the browser, the response is 200 OK, but it contains the HTML error content, not the expected JSON content.

• CP-21034 The UI should not depend on the activation of the built-in REST service, defined from the global permissions dataset.

• CP-21188 RequestUriFilter does not properly handle base URI definition according to built-in data services lineage and modules public path prefix property.

  When the _ebx.module.publicPath.prefix_ property and the built-in data services lineage are defined, their URI path value are concatenated resulting in wrong URI builds.

  If no built-in data services lineage and no modules public path prefix are defined, then default values must be applied.

  * server URI path: /

  * built-in data services URI path: /ebx-dataservices

  * base URI path: the one from the HTTP request

  * request URI path: the one from the HTTP request

  If no  built-in data services lineage is defined but there is a modules public path prefix, then it must be considered:

  * server URI path: modules public path prefix

  * built-in data services URI path: modules public path prefix + /ebx-dataservices

  * base URI path: modules public path prefix + partial base URI path which begin with the servlet name segment

  * request URI path:  modules public path prefix + partial request URI path which begin with the servlet name segment

  If there is a built-in data services lineage, then it must be considered:

  * server URI path: built-in data services lineage URI path minus the servlet name segment 

  * built-in data services URI path: built-in data services lineage URI path

  * base URI path: server URI path + partial base URI path which begin with the servlet name segment

  * request URI path: server URI path + partial request URI path which begin with the servlet name segment

User interface

• CP-825 [37252] When setting the 'Node category' field to 'Hidden' for a table, under the 'Action' menu of an associated data set, 'Categories' is not shown.

• CP-18601 The horizontal scroll bar is not displayed, in a form opened from a work item.

• CP-19041 The 'collapsed' state in custom perspective does not remain.

  After clicking on the 'Collapse' button of a custom perspective, then opening the 'Select perspective' menu, the navigation pane of custom perspective is changed to 'Expand' state.

• CP-19134 An incorrect behavior occurs when a non-admin user changes the owner on the information page of a dataspace.

  When a non-admin user changes the owner on the information page of a dataspace and saves it, no error message is displayed. When clicking on the 'Close' button, the error message is displayed with duplicated content.

• CP-19789 [All browsers] The service icons overlap the label view when accessing as a non-admin user.

  A view created by an admin user is shared with all profiles. When accessing this view as a non-admin user, all service icons overlap the label view.

• CP-19829 The default value of the 'Progress strategy' field on the creation form is not synchronized after updating the configuration in the 'Modeling default values' table.

  When '[not defined]'/'Display the work items table'/'Automatically open the next step' is set to be the default value for the following fields:

  * Progress strategy for a user task

  * Progress strategy for a subworkflow invocation

  * Progress strategy for a script task

  * Progress strategy for a condition

  Then, on the creation form of a user task/sub-workflow invocation/script task/condition, the value of the 'Progress strategy' field is [not defined].

• CP-19868 The focus on custom perspective item does not take the full width.

• CP-19870 In custom perspective, the selected item loses its selected style after a few milliseconds.

• CP-19968 The dataspace tree does not expand to the search result.

  When users search on a dataspace tree, if the search result is a child dataset of level 2, the dataspace tree does not expand to that result.

• CP-20007 [All browsers] A JavaScript error occurs after accessing a snapshot.

• CP-20232 Exception thrown on advanced search for a 'Resource' field.

  If users create a criterion for the resource field on the advanced search pane, then apply it, an error occurs.

• CP-20264 [Edge44] Some paging buttons are not displayed and highlighted when accessing a table.

  On the Edge browser, when users access a table that has over 10 records, some paging buttons: 'next>|', 'last>|' are not displayed and highlighted. When users hover over it, they are displayed and highlighted.

• CP-20316 An error message is shown when applying the quick search on the 'Active workflows' table.

  When users apply the advanced search on the 'Priority' field, an error message is shown.

• CP-20394 [All browsers] The View menu is not fully displayed when creating a view with a long name in the 'Administration' area.

  When creating a view with a long name in the 'Administration' area, the view menu is not fully displayed.

  The issue also happens on the 'Data workflows' area.

• CP-20427 An item is missing on the breadcrumb of a dataspace.

  When opening a permission record or a history record in a data space, the breadcrumb is not displayed fully.

• CP-20516 [Chrome/Edge]The label of the tables under the 'Extensions' is blurred when zooming out the page 80%.

  When zooming out the page 80% and accessing any table under 'Extensions', the first letter of the label is blurred.

• CP-20646 Sub tabbed pane should fill parent width.

  A tabbed pane under another tabbed pane has a special display. It should have the same width than the parent tab.

• CP-20785 Some matching records are not highlighted when repeatedly clicking on the 'Next' button in the 'Find in hierarchy' pane.

• CP-20811 The matching records are not highlighted when executing the 'Find in hierarchy' on the integer field.

• CP-20859 In the 'Record permission script' extension, an alert is displayed when user clicks the 'Revert' button when script has been modified.

  A warning displays that "modifications may be lost".

• CP-20868 When navigating to a new dataspace, the previously selected dataset node should remain selected.

  When user changes the dataspace, the UI tries to keep the same dataset selected, if it exists.

  The same process should apply to the node selected in the dataset.

• CP-20899 Custom strategy label overwrites reserved name.

  A user imports an XML file in the strategy table of the search extension. If this file contains a record for which the name of the strategy is a reserved name, the record will be displayed in error in the table.

  But in the extension search, the name of the associated built-in strategy will be overwritten by the label of the custom strategy in error.

• CP-20904 The authorization header is not correctly escaped

  The authorization header contains unexpected characters, which may prevent the user from logging.

• CP-20931 Extensions are displayed in RW policy when they should be displayed in RO.

  When setting the value access policy of a model in "Read-Only", the extensions under the 'Extension' tab are still displayed as though they could be overwritten. It is not the case - the user cannot save the modifications brought to the extension- but the display should not allow any modifications.

  When accessing extensions through the navigation pane, the same issue occurs. If we are not in the case of a table, we have still the option to modify and try to save the extension. 

• CP-20948 The documentation pane of the default strategy is truncated in the case of a foreign key.

  In the case of a foreign key, the default strategy documentation pane is not fully displayed. 

  The only solution is to increase the height of the extension frame to fit the whole description.

• CP-20951 [All browsers] A blank extension content is displayed when clicking on the 'Save custom layout' button in the 'Set custom form' area under the 'Extensions' tab of a table.

• CP-20952 The record is still highlighted when modifying/removing a search on the hierarchy view.

  If users apply a search combined with finding in hierarchy: the record remains highlighted when users modify or remove that search.

• CP-21031 The 'fullscreen' state of the Data Model Extension should be persisted.

• CP-21033 When duplicating a dataset, and inputting an existing name for it, the error message is not user-friendly.

• CP-21045 ConcurrentModificationException thrown when adding a criterion.

  A ConcurrentModificationException may be thrown when adding a criterion in the advanced filters pane.

• CP-21046 When closing a user service form, the user is not redirected to the right view.

  When closing a user service form with the default 'Close' button, the user is not redirected to the right view.

  This occurs when the record was opened from a work item that selects a custom hierarchical table view.

  In that case, the user is redirected to the default view instead of the custom one.

• CP-21064 In custom forms, the 'Save', 'Save and close' and 'Revert' button should be hidden.

  When displaying a record with a custom layout: if this record is inherited or in read-only mode, then these buttons should not be displayed.

• CP-21070 Error on a hierarchy view based on osd:select dependency.

  When users create and apply a hierarchy view, then apply quick search/advance search on the "Publication" table on "Data Modeling" dataset under "Administrations", an error occurs.

• CP-21087 When overwriting a record, the user is redirected to the model-driven form.

  When overwriting or inheriting a record from a custom form, the user is redirected to the model-driven form instead of the custom form.

• CP-21105 Encoding mismatch when enabling file upload for the form with UserServiceDisplayConfigurator.

• CP-21122 The 'Clear field' buttons are not displayed on Date/Time/DateTime fields in the advanced search pane.

• CP-21123 A non-admin user cannot access a child dataset.

  A non-admin user cannot access a child dataset when that user does not have permission in the parent dataset.

• CP-21144 A blank workspace is displayed when combining a quick search with an advanced search, and overwriting a record on a child dataset.

  On a table of a child dataset, when users combine a quick search with an advanced search then overwrite a record, a blank workspace is shown.

  When users apply a quick search with an advanced search and go to next page/ last page then delete a record, an error message is displayed.

• CP-21197 A fatal error occurs when creating a multi-valued field with data type: 'Password'.

  When users create a multi-valued field with the data type as 'Password', a fatal error occurs. A JavaScript error also appears in the console.

• CP-21225NullPointerException when loading a filter containing a criterion on validation.

  When loading a filter having a criterion on "Message" or "Severity", an NullPointerException is thrown.

• CP-21242 The quick search feature causes BootstrapMethodError.

• CP-21313 In custom forms, groups are not expanded by default.

  In custom forms, when adding a 'var' block referencing a group inside a 'Form', the group is always collapsed.

• CP-21341 Error messages are not visible when using a workflow launcher.

  When using a workflow launcher, the messages indicating that the workflow could not be launched should be errors, and should pop up automatically.

• CP-21431 A white page is displayed when closing a record after applying advanced search on enumeration Locale/URI field.

  When users apply an advanced search on an enumeration Locale/URI field, then open the filtered record and close it, a white page is displayed.

• CP-21507 Avatar in header should be wider.

  Avatar size is too small; user initials are too close from  avatar border.

• CP-21542 Some JS buffer errors occur in the log files.

  Some JS buffer errors occur in the log files :

  2021-05-28 14:53:21,738 CEST ERROR log.kernel 1149 [http-bio-8080-exec-2] 127.0.0.1 7D481DD26E68CCAF7ACE6E4DBFADE533:0 ebx-manager P .sessionIn.user_service wbp:_Manager.html 0xaruccs61ek6zyc6u22d81n3b9jiiv5 on_select_datas UIBuilder's JSBuffer is not empty. Some JavaScript code has been lost.

  No noted functional error

• CP-21652 PerspectiveHandler#getCurrentPerspective(boolean) may return null in some unintended cases.

  The documentation of PerspectiveHandler#getCurrentPerspective(boolean) states:

  {quote}Returns the perspective applied in the current session, or recursively in parent if lookupInParentSession is true.

  - If the current session is the root session, always returns a perspective.

  - If the current session is a sub session and lookupInParentSession is true, always returns a perspective.

  - If the current session is a sub-session and lookupInParentSession is false, may return null.{quote}

  But in some circumstances:

  * it can return null for a root session,

  * it can return null for a sub-session even if lookupInParentSession is set to true.

• CP-21658 Wrong prototype for UIHttpManagerComponent#initFromHttpRequest().

  A second parameter has been added to UIHttpManagerComponent#initFromHttpRequest(). It should be removed.

Users & Permissions

• CP-21243 Access rules on occurrence returning 'Hidden' with a message are not filtering the queries accordingly.

  When an access rule on record returns 'Hidden' access permission with an additional UserMessage, the record is yet returned from RequestResult.

6.0.0 Closed issues

Core engine

• CP-742 [36071] @36071 Exceptions can occur while requesting data while customer modules are not fully compiled or have been unregistered.

  Exceptions can occur while requesting data while customer modules are not fully compiled or have been unregistered.

• CP-1961 [31455] The default log file for a given category should not be created when the logs are redirected elsewhere.

• CP-16455 [37260] Unclear error message when an OperationException is thrown during the "apply last modifications" action.

• CP-16534 The API for RepositoryAnonymizer.RecordContext is incorrect.

  The method RepositoryAnonymizer.RecordContext.setValue(SchemaNode aNode, Path aPath)

  has the wrong parameter types, and should be changed to: setValue(Object aValue, SchemaNode aNode).

• CP-17052 An error occurs when building an XPath filter that uses a validation predicate.

  An error occurs with a NullPointerException when building an XPath filter that uses a validation predicate.

• CP-17156 An error occurs when publishing a data model declaring a replication with a name longer than 32 characters.

  An error occurs when publishing a data model with a replication that has a key length greater than 32 characters. The max key length should be 64 characters.

• CP-17527 A deadlock may occur on startup on a JBoss EAP server.

  When using a JBoss EAP server, a deadlock may occur on startup.

• CP-18296 An exception occurs while renaming a mapped column under a group through the API.

• CP-18609 If the JavaMail jar is missing, an error message should explicitly state it.

  When deployed on Tomcat, if the JavaMail jar is not installed, startup fails without relevant information.

• CP-18677 Upgrade script to 5.7 assumes that the table name starts with the repository prefix.

  If a repository has run on EBX® 5.0 to 5.2, and used history or relational tables, these have been created without a repository prefix. However, installation script TSA.6 - issued for upgrading a repository to 5.7 - assumes that the name of any relational table starts with the repository prefix. The upgrade to 5.7 fails with this message:[SqlChecker] Table EBX_MY_TABLE does not exist.

• CP-18782 Mapped mode fails with the H2 1.4.200 database when a field name is a reserved word.

  Mapped mode fails with the H2 1.4.200 database when a field name is a reserved word.

• CP-18946 An error occurs when requesting the history of a table in all dataspaces.

• CP-20392 ObjectAlreadyExistsException when registering a scheduler cron task, if it already exists.

  An exception is thrown when a module registers several times a scheduler cron task with  the same identification. A warn log should only be added.

• CP-20546 An error can be silently dropped when executing an SQL request.

• CP-20828 A blank workspace is displayed when the 'Seniority (days)' field contains special characters.

  In the 'Administration' area, when users create the 'Seniority (days)' field with special characters then re-access the 'Task scheduler' dataset, a blank page is displayed and an exception is thrown in the log file.

Data modeling

• CP-2425 [28754] The method SchemaDocumentationContext.getSession can return null erroneously.

• CP-16910 The list of data model versions is empty if it has not been previously published.

  The section that displays the versions of a data model is empty if the data model has not been previously published.

• CP-17491 Error at the data model compilation when inheriting a field from another data model.

  An error occurs at the data model compilation when a field of the data model inherits from a field in another data model through 2 foreign keys.

• CP-17583ClassCastException referencing field from an included schema in a computation rule.

  ClassCastException referencing field from DaqaMetadata in a computation rule.

• CP-17595 Constraint on the "source location" field in the DMA should return a warning instead of an error.

  Constraint on the "source location" field in the DMA should return a warning instead of an error. 

• CP-17835 The "Trim white spaces" property is not inherited from a simple data type.

  When the "Trim white spaces" property is set to false on a simple data type, and a field reuses this type, this property is not inherited, even though it looks like it is in the DMA. Also, the property is still not considered when overloading the value.

• CP-18527 Records can still be duplicated when this action is forbidden on the data model.

  Users set 'Duplicate record' on the table of a data model to 'Forbidden' but in associated datasets, records can still be duplicated.

• CP-20274 The XSD generated by the Data Model Assistant is not formatted.

  The datamodel XSD file no longer has any tabs/spaces to make the XML human-readable.

Data services

• CP-552 [36564] Error when using built-in REST service with filter by record label

  A NullPointerException occurs when using the built-in REST service to get table data with filter by predicates on a record label (osd:contains-record-label).

• CP-1120 [36858] REST 'get history' returns an unexpected error when the table has locale 'zh-cn'

  When a record has the locale zh-cn, the REST call to get the history of a table returns an unexpected error. 

• CP-16393 Lineage base URL makes EBX® inaccessible.

  Lineage base URL makes EBX® inaccessible. 

• CP-16482 Comparing with filter isResolvedMode returns a different result than from the Web.

  In unresolved mode, the comparison result of two records by a workflow using a built-in service is different from the one using the public API DifferenceHelper.compareOccurrences(Adaptation, Adaptation, boolean).

• CP-16638 When setting a toolbar with isForTable=false in a UserServiceDisplayConfigurator, table-level services are offered instead of record-level services.

  When setting a toolbar with isForTable=false in a UserServiceDisplayConfigurator, table-level services are offered instead of record-level services. 

• CP-17197 REST Toolkit: UnsupportedOperationException error thrown when combining standard authentication, POST verb and application/x-www-form-urlencoded content-type

  An UnsupportedOperationException error is thrown during the standard authentication process.

  The exception is caused by a replacement of the provided password by an obfuscated one and it must be done for security reason.

• CP-17227 HTTP Request Authentication scheme is performed even if a built-in authentication is tried

  During the authentication activation / deactivation feature development, a condition has been removed and must be added anew. If a built-in authentication scheme can be applied to the given request, then the HTTP Request Authentication scheme and the following anonymous authentication schemes must not be tried.

• CP-17747 Incomprehensible error message while inserting data using REST data services.

  Such a message is returned as soon as there is a node under the table path in the base URI, which is related neither to the record primary key, nor to the technical primary key.

• CP-17881 Provide more detail when authentication is denied because of multiple methods in a REST request.

  When a REST request is received with multiple authentication methods configured, only the message "Authentication failed" is returned. No information regarding the cause is provided, even in the log files at debug level.

  A reason for the failed authentication should be given.

• CP-18562 Property 'ebx.http.authorization.header.policy' should not impact basic authentication

  Currently, setting ebx.http.authorization.header.policy from file ebx.properties impacts all authentication modes.

  Standard "Basic" authentication mode then becomes incompatible with most REST clients.

  This setting should only concern the 'Token Authentication Scheme'.

• CP-18589 The 'REST forward authentication' should not be limited to use the built-in operations.

  The introduction chapter of the REST dataservices mention a limitation regarding REST forward authentication. The limitation restricted to the built-in services is incorrect, it should be removed.

• CP-18590 The 'ebx.dataservices.rest.auth.tryBasicAuthentication' property does not deactivate the Basic authentication method.

  When the 'ebx.dataservices.rest.auth.tryBasicAuthentication' property is set to 'false', the 'Basic' method is not deactivated; the documentation of this property should be improved.

• CP-18785 REST built-in select operation with 'includeSortCriteria=true' parameter aborts.

  The REST select operation aborts if the following conditions are all true:

  'includeSortCriteria=true,' the 'sort' parameter is defined, and the pagination mecanism is activated.

• CP-19491 Incorrect permission evaluation with REST built-in service, when updating a record from 'overwrite' to 'inherit'.

  When changing the inheritance mode of an 'overwriting' record back to 'inherit', the permission to evaluate should be 'Delete a record' instead of 'Overwrite inherited record'.

• CP-19503 Label and details URL for foreign key field are unavailable on REST history content.

  When a user requests the history content through the REST built-in service, the label and URL of details for foreign key fields are unavailable. However, this information is displayed in the user interface.

• CP-19690 Error when REST request contains an XPath predicate in the URI

  When users request a GET method containing an XPath predicate, an error occurs.

Data validation

• CP-1195 [36513] Duplicated mandatory error message on a foreign key field.

  Duplicated mandatory error message on a foreign key field when a specific constraint is declared on this field.

• CP-18637 Mandatory message constraint not displayed for list types.

  Multi-valued fields never display their mandatory message constraint.

• CP-19257 A foreign key with a blocking constraint prevents from merging dataspace.

  A foreign key that references another dataset of the same dataspace prevents from merging, if the 'blocksCommit' policy of the constraint is set to 'onInsertUpdateOrDelete'.

• CP-19407 OutOfMemoryError while getting the validation report.

  An OutOfMemoryError occurs when a dataset has too many messages, due to printing the validation report into the logs.

• CP-19434 Blocking foreign key constraint during merge.

  If the blocksCommit property onInsertUpdateOrDelete is set on a foreign key field, then the merge operations may be blocked, even if there are no errors on that foreign key.

• CP-17968 Association definition inconsistency when creating a child dataspace.

  An error occurs when accessing a child dataset using that use an association with a foreign key on a explicitly defined dataspace, due to the association definition inconsistency.

• CP-19561 Validation error appears in the record form, even for hidden fields

  Even when a field is hidden programmatically, the validation error displays in the record view

Data views

• CP-610 [36975] Impossible to compare 2 records if the table contains a data complex type with the "Access properties" as "Not a dataset node".

  When comparing 2 records, if the table contains a data complex type which sets the "Access properties" to "Not a dataset node", then it does not work.

  The compare service should work normally. The data complex type that does not have the permission should be hidden from the compare screen.

• CP-945 [37039] The 'Find in Hierarchy' service does not work when the table has more than one primary key field.

  When searching with the 'Find in Hierarchy' mode activated, the matched record is not highlighted if the table has more than one primary key field.

• CP-1569 [33794] Running the 'Move up' service from a leaf node of a child data set raises an error, and the leaf node is not moved.

  The two following error messages are shown when users run the 'Move up' service: "The requested operation has failed. Reason: Cannot create the new record since its primary key is already used." and "Move up".

• CP-1710 [32335] When a user who has no rights to share views duplicates a view, it is shared with the same profiles as the original view.

  When a user who has no rights to share views duplicates a view, the view is shared with the same profiles as the original view. These users should not be able to share any view, even by duplicating a shared view.

• CP-1950 [29725] An error occurs when calling the REST data service with the osd:contains-record-label function.

  An error occurs when calling the REST data service select operation with the osd:contains-record-label function.

• CP-3653 [22652] On the hierarchy view, [unset] node only displays the number of records without displaying any record after searching.

  On the hierarchy view, after searching with large volumes of data (~500000 records), the [unset] node displays the page numbers but does not display any record.

• CP-3857 [22123] Inconsistent behavior between creating and attaching a record under an unset node of a hierarchy.

  In a hierarchy view based on recursive ternary simple join tables, when the user creates a new node from an unset node, an error displays. When the user attaches a node to an unset node, no error displays.

• CP-4244 [21609] Some leaf nodes are not selected when running the service "All" in a simple join hierarchy view.

  In a simple join hierarchy view, when running the service 'All' under the 'Select and display' menu, some leaf nodes in the last page are not selected.

• CP-4359 [23080] Hierarchy nodes which are not orphan nodes are displayed under an unset node.

  In a recursive ternary hierarchy, a hierarchy node which is not an orphan node is displayed under an unset node.

• CP-4943 [17398] Incorrect behavior displaying the page when creating new records after making a global selection.

  This issue occurs when the user stays on the last page to create a record on a hierarchy view. If the user makes a global selection and then creates a new record, the page containing newly created records is not displayed.

• CP-5483 [17372] Unexpected error is displayed when checking the constraint of a field under a complex list.

  This issue occurs on a field under a complex list which defines a constraint. After the user creates a record with the specified value for this field, an unexpected error displays.

• CP-16514 A java.lang.StackOverflowError occurs when creating a pruned hierarchy over a link table.

  A java.lang.StackOverflowError occurs when creating a pruned hierarchy with a filter on the dependency link table.

• CP-17537 The message box is shown inappropriately when opening a deleted record.

  When users open a deleted record, an error is shown. When opening the message box, an inappropriate message is displayed.

• CP-17541 The 'Display the message box' button is not shown upon the first time accessing a fatal dataset after publishing.

  When users access a fatal dataset for the first time after publishing, the 'Display the message box' button is not displayed.

• CP-17555 [Chrome] The checkbox of any search on the 'Filter' pane does not work properly.

  When users click on the checkbox of any search on the 'Filter' pane, the 'Validation search' is expanded/ collapsed instead of  being applied.

• CP-19687 Filter on link table may disappear when creating a custom view.

  When creating a custom view that specifies a filter on link table as a dimension criteria, the filter disappears after creating the view, saving then reopening. 

Data views - hierarchies

• CP-16757 Performance issues while building a hierarchy tree.

  New optimizations have been implemented in the hierarchy tree construction algorithm.

• CP-17806 The arrow button is available for an unset node after running the "Attach to another parent" service in the hierarchy view.

  When using the "Attach to another parent" service of an unset node in the hierarchy view, arrows are available for unset nodes.

Data workflows

• CP-1000 [35068] A user is able to launch a workflow from a URL without the required permission.

  A user is able to launch a workflow from a manually built URL without the required permission.

• CP-16739 The filter is kept when switching between the custom perspective action and the advanced perspective.

  The filter is kept when going from an inbox as a custom perspective action to the inbox in the advanced perspective.

• CP-17015 The session interaction is not valued in the AccessRule execution.

• CP-17216 According to the documentation, the data context wizard should be available by default, but it is missing.

  The context wizard is missing for scripts and user tasks.

• CP-17223 The space at the bottom of the work item card is too large.

  In both densities, the space at the bottom of the work item is not compliant with the space on the left, right and top.

• CP-17440 The link between 'Initial transition' and 'Final step' is lost when deleting steps by pressing 'Delete' on the keyboard.

  On the workflow diagram of a workflow model, users select all steps between the 'Initial transition' and the 'Final step', then press the 'Delete' button on the keyboard, the link between 'Initial transition' and 'Final step' is deleted as well.

• CP-17629 The icon path for the workflow diagram should be relative.

  The path for the icon is absolute and not relative. Thus, when testing in the cloud for example, the icon does not display correctly.

• CP-19487 Error when using service extensions in perspectives.

  When accessing service extensions in perspectives, we get the below error: Perspective action is badly configured: action 203 in perspective validation-overview

Dataspaces & History

• CP-16773 An error occurs when anonymizing list fields.

  If you invoke RepositoryAnonymizer#anonymizeHistory with a consumer that modifies an aggregated list element, then the following exception occurs when constructing the DDL statements:

  java.lang.ArrayIndexOutOfBoundsException: 37

Distributed Data Delivery (D3)

• CP-19973 A communication error in D3 federation broadcast is not detailed in console and log.

  When a broadcast in federation mode is configured, if the communication with a replica node does not work, the stack information is not added to the D3 primary console and D3 log.

Documentation

• CP-17099 REST built-in authentication methods are missing in the documentation.

  The REST built-in authentication methods for 'Anonymous' and 'HTTP Request' are missing in the documentation.

• CP-17639 The custom module log threshold documentation is incorrect.

  The property ebx.log4j.category.log.module.xxxxxx defined in the main configuration file ebx.properties should be ebx.log4j.category.log.wbp.xxxxxx.

• CP-19646 Bronken links in pdf documentation if the anchor contains a point character.

  When an anchor contains the point ('.') character, the pdf link is incorrect. The product documentation contains 2 broken links.

Import / Export

• CP-20109 XML Export is not formatted

  The exported  XML misses tabs for formatting.

Java EE integration

• CP-17392 When deploying EBX® over Weblogic, the Jersey library should be prevented from scanning REST Toolkit services.

  When deploying EBX® over Weblogic, the Jersey library should be prevented from scanning REST Toolkit services.

  Consequently, the scanning operation has to be disabled.

• CP-18044 The shipped Apache Xerces library may lead to a StackOverflow error on custom code.

  The shipped Apache Xerces library may lead to a StackOverflow error on custom code that validates an XML file. This issue must be fixed by upgrading the shipped Apache Xerces and Xalan libraries (plus any related libraries).

• CP-18909 Weblogic cannot start up since CXF has been upgraded to 3.2.12

  CXF has been upgraded to 3.2.12 for security reason. However, this version prevents Weblogic server from starting up. 

  Some CXF MessageBodyWriter cannot be registered into Weblogic's HK2 component.

• CP-20339 Fix the JMS section and improve JDBC section in the JBoss EAP 7.1.x installation note example.

  In the JBoss EAP 7.1.x installation note example.

  The JMS section is incorrect since the  EBX_JMSConnectionFactory is defined as a queue and should be specified in a <connection-factory> element.

  The JDBC section has been improved to specify sample as for PostgreSQL.

Messages & i18n

• CP-17533 The UserMessage JavaDoc is not fully compliant with the MessageFormat pattern.

  When using a UserMessage the pattern syntax is not fully compliant with MessageFormat as described in the documentation. For example, it is currently not possible to declare a format type and a style pattern to a format element.

Miscellaneous

• CP-18686 Event broker aborts on firing dataspace creation event, if the creation has aborted.

  When a dataspace aborts on creation and if a subscriber has subscribed to the dataspace creation event, an exception is thrown.

REST Toolkit

• CP-17057 When a JSON body request is required but is not available, or the format is incorrect, a 500 HTTP status code is returned.

  When no body request is sent by the client and a JSON body request is required, a 500 error is returned instead of a 400 bad request.

  When a JSON body request with bad format is sent by client, a 500 error is returned instead of a 400 bad request.

• CP-18370 REST Toolkit returns HTTP error status 500 with a null body request.

  The JSON parser aborts and returns a HTTP error status 500, when a request is sent, using POST method with a Content-Type set to application/json and a request body set to null.

• CP-19259 Asymmetrical representation of URI in Rest Toolkit JSON serialization.

  When serializing a URI which is placed within an Object, a default Johnzon Adapter converts it into a String. However, the internal URISerializer takes the priority when the URI is placed within a list (at whichever level) or directly returned by the REST Java method. A URI type is then used.

  *Workaround:* Use the String representation of the URI, instead of the URI itself. If the _absolute_ and _serverRelative_ attributes are needed, use a DTO holding the String representation of this data.

• CP-20186 Too many SocketExceptions due to browser connections management with Tomcat.

  When the browser manages connections, it can close some of them. The server detects SocketException with "broken pipe", "socket write error" message and logs in restService category of the EBX® server. For Tomcat server, it catches and rethrows the exception with org.apache.catalina.connector.ClientAbortException.

User interface

• CP-533 [35889] The sub-page's name ("Information" and "Permission") of a dataset disappears from the navigation pane when opening it twice.

  When users open a sub-page (Information/ Permission) twice on a dataset, the name of that page disappears from the navigation pane.

• CP-925 [36282] The number of messages in the message box accumulates when an error occurs.

  The number of messages in the message box accumulates when users receive an error message in the Web component link generator.

• CP-998 [36221] [LAYOUT][IE] The second header of a modification workflow model step displays incorrectly in 'Compact' mode.

  Users open an existing workflow model step when the density is set to 'Compact' mode, the second header displays incorrectly.

• CP-1238 [36029] [Chrome] The workspace is corrupted when saving a record with cross-script labels.

  When saving a record with cross-script labels, if this label field is also a unique field, the workspace is corrupted.

• CP-1265 [36313] Accessing a dataset via a permalink in the node scope results in an erroneous empty page.

  If there is a field at the root of the dataset, accessing it via a permalink in the node scope leads to an erroneous empty page.

• CP-1467 [34590] [LAYOUT] The design for the 'Duplicate a dataset'/'Create a data set' page is wrong.

  On the 'Duplicate a dataset'/'Create a data set' page, the design for fields is not the same as the what is stated in the guideline.

• CP-3727 [22232] The favicon defined for a child perspective is ignored.

• CP-5429 [17579] If a hierarchy view is applied in a filtered foreign key selector, records that do not match the filter can be selected.

  If a foreign key has a filter, when the user opens its selector and applies a recursive hierarchy view, then it is possible to select all the nodes of the hierarchy, including those that do not match the filter.

• CP-16339 EBX® does not work if we specify the following parameter in ebx.properties in order to deploy the static resources on another server:port than the server:port of servlet.

  EBX® does not work if we specify the following parameter in ebx.properties in order to deploy the static resources on another server:port than the server:port of the servlet:

  ebx.externalResources.http.host=otherhost.

• CP-16342 EBX® does not work with the HTTP Servlet path parameter set in admin

  When the HTTP Servlet path parameter is set in admin configuration, EBX® doesn't work.

• CP-16351 An error occurs when performing actions on the grid edit for an association by an XPath predicate in a user service.

  An error occurs with a NullPointerException when performing actions on grid edit for an association by an XPath predicate opened in a user service.

  The error occurs when the user clicks on the "Close" or "Save and Close" buttons in the modal opened for grid edit.

• CP-16379 Backed-up user page does not display when implemented in a custom directory.

  When a custom directory is implemented and a user has one or more backed-up users, the backed-up user selector page does not display after a successful login.

• CP-16390 Record navigation arrows are only displayed for the records on the first page.

  Record navigation arrows are only displayed for the records on the first page. They should also be displayed for records on the other pages.

• CP-16401 The dataset selector does not open automatically.

  When a dataspace is selected and more than one dataset can be selected, the dataset selector should open automatically.

• CP-16447 The trackingInfo value, when defined on a perspective action, is not available in the session.

  The trackingInfo value, when defined on a perspective action, is not available in the session. The getTrackingInfo() method returns null. 

• CP-16471 State of selected records on custom perspectives is kept when switching to the advanced perspective.

  Access built-in services to select more than one record on a custom perspective, then switch to the advanced perspective, the state of selected records is kept.

• CP-16524 Blank space appears when using a web component inside a tabbed pane.

  A custom form that uses a tabbed pane with a tab displaying a web component will have its layout broken.

  It appears as though its display shifted to top, leaving a blank space at the bottom.

• CP-16529 A user service 'Close action' does not work in a workflow user task.

  The 'Close action' of a user service does not work in a workflow user task: the result screen contains only the workflow 'Comment' link and the 'Reject' and 'Accept' buttons.

• CP-16581 A concurrency conflict happens when a record is updated through a trigger in the parent session.

  A concurrency conflict happens when a record is updated through a trigger in the parent session.

  Such an issue happens if the update is done in a child dataspace and if the record in the parent session is sharing its revision with its parent dataspace.

• CP-16587 Too many calls to authentication methods since EBX® 5.9.0.

  Too many calls to authentication methods since EBX® 5.9.0. Especially when authenticating through a custom directory using the method authenticateFromHttpRequest.

• CP-16656 When using the link workflow.workItem.link.allocateAndStart the user is not redirected to the work item when he does not have access to the advanced perspective.

  When using the link workflow.workItem.link.allocateAndStart the user is not redirected to the work item when he does not have access to the advanced perspective. He is instead redirected to the start page of a custom perspective.

• CP-16663 The user is logged-out after completing a work item opened in a Web Component from a filtered inbox view.

  The user is logged-out after completing a work item opened in a Web Component from a filtered inbox view. 

• CP-16681 Impossible to compute the size in a tabular view in some cases.

  When the size of a tabular view on a large table is already computed, if the size becomes unknown - e.g. because a search criterion has changed - then the "Compute size" menu item is not displayed.

• CP-16694 When using UserServiceNext in a UserServiceExtended<DatasetEntitySelection>, the user is not redirected.

  When using UserServiceNext in a UserServiceExtended<DatasetEntitySelection>, the user is not redirected.

• CP-16725 Icons missing after upgrading

  After upgrading to the new version, some icons do not display due to a caching issue.

  Although it can be fixed by cleaning the browser cache, it is impractical.

• CP-16733 Faulty display when trying to display an association widget several times in a UserService.

• CP-16855 User Service, JS error occurs when there is a tabbed pane embedding other tabbed panes.

  If a user service tries to specify a tabbed pane which embeds another tabbed pane, a JS error will occur on display.

• CP-16921 [All browsers] The 'Close' button overlaps the 'Help' button in a pop-up when in Compact mode.

  In a random pop-up screen when publishing an erroneous model, the 'Close' button overlaps the 'Help' button.

• CP-16925 Unchecking a checkbox on a Boolean field sets the field value to null instead of false.

  Unchecking a checkbox on a Boolean field sets the field value to null instead of false.

• CP-16927 [All browsers] The 'Close' button overlaps with the 'Display the message box' button in a pop-up when in Compact mode.

  In a pop-up screen, when adding an invalid formula in the computation rule, the 'Close' button overlaps with the 'Display the message box' button.

• CP-16932 When duplicating a record through an association, its primary foreign keys that target other tables are in read-only.

  When duplicating a record through an association, its primary foreign keys that target other tables are in read-only. They should be in read-write instead.

• CP-16992 Built-in service @hierarchyEditLinkRecord does not open the right record.

  Built-in service @hierarchyEditLinkRecord always redirects the user to the record that has been opened first for a similar association element (last dimension of the association in the hierarchy tree). 

• CP-17000 An error occurs when accessing history from an association.

• CP-17033 [IE][Edge] Unable to access a table content when the initial request contains a parameter whose value is very long.

  The content of a table cannot be displayed if using the Microsoft Internet Explorer or Edge browsers when the initial request contains a HTTP parameter whose value is very long.

  Nothing is displayed in the workspace pane.

• CP-17059 Under certain conditions a NPE can occur when selecting the dataspace tree.

  When using the REST UI service on the dataspace version 1 for a tree operation, and under certain conditions, a NPE can occur when selecting the dataspace tree.

• CP-17114 [IE10] The background color of the 'Confirm password' label displays incorrectly when creating a new user in the Users table under the Administration area.

  On IE10, when creating a new user in the Users table of the Directory, under the Administration area, the background color of the 'Confirm password' label is displayed incorrectly.

• CP-17115 Theme color for the selected item in the perspective navigation tree is not applied.

  The theme color for the selected item in the perspective navigation tree is not applied.

• CP-17256 An error occurs when extracting the XPath expression of a perspective action.

  An error occurs with a NullPointerException when extracting the XPath expression of a perspective action. It happens when the XPath parameter of the action is not specified.

• CP-17270 Workspaces in sub-sessions with a scope greater than the node replace the root workspace.

  If a web component with a scope data, dataset or dataspace is displayed in a workspace, then only the workspace content of this web component is displayed.

• CP-17313 [IE10] Cannot edit the user password, the button does not react to clicks.

  On IE10, when the password is not empty, the user cannot click on the "Change password" button, nothing happens - as if the button was disabled.

• CP-17469 EBX® becomes inaccessible when a redirection restriction is specified in the exit policy.

  The application becomes inaccessible when a redirection restriction is specified in the exit policy.

  It occurs when the property ebx.servlet.useLocalUrl parameter of ebx.properties is set to 'true'.

• CP-17536 An incorrect message is shown after running "Refresh data model" under 'Development tools' in the 'Administration' area.

  When users run 'Refresh all data models'/'Refresh updated data models', an incorrect message is shown.

• CP-17540 The message under 'Display the message box' is misplaced when an error occurs on the 'Web Component link generator' preview.

  When users open an erroneous preview in the 'Web component link generator' screen, the message under 'Display message box' is misplaced.

• CP-17545 Cross origin Javascript error.

  When EBX® GUI is displayed in an IFrame, when the Iframe container is not in the same domain than EBX®, a cross origin Javascript error can occur.

• CP-17561 An error appears when saving a record on a user service launched from a record that belongs to another table.

  An error appears when saving a record on a user service launched from a record that belongs to another table.

  The 'An error occurred while displaying field. Please contact your administrator.' message is displayed near a foreign key field when the user service page is closed.

• CP-17622 Embedded Web Components modify the parent EBX® breadcrumb.

  When a Web Component is embedded in another EBX®, it modifies its breadcrumb once loaded.

• CP-17631 An error occurs when the authorization header contains special characters.

  An error that prevents the user from logging in occurs when the authorization header contains special characters.

• CP-17655 HTTP parameters are lost while accessing a permalink through a custom login page.

• CP-17656 Permalinks generated through the GUI can be broken.

  When a custom login page is declared, the permalink to access a record is broken.

  [|http://localhost:8080/ebx/?instance=toto&xpath=%2Froot%2Ftable&branch=toto]

• CP-17671 Erroneous date in a form due to the application server and the browser being on different time zones.

• CP-17765 An error occurs in a user service whose form contains a submit button named 'submit'.

  An error is displayed when executing a user service which contains a form with a submit button called 'submit'.

• CP-17788 Table header icons should have light color on a dark background.

  The background color and font color of the table header can be customized in the administration.

  However, when the font color is "light", the icons of the table header do not change accordingly.

  They remain grey, whereas they should be white.

• CP-17813 In a programmatic record label, 'UILabelRendererContext.isHTMLForbidden()' never returns true.

  In a programmatic record label, UILabelRendererContext.isHTMLForbidden()never returns true.

• CP-17853 The task name is erroneous after running service "Hide in the workflow progress view".

  When users run the service "Hide in the workflow progress view", on the "Workflow model steps" screen, the task name is erroneous.

• CP-17976 Within a user service, cannot select or close the foreign key selector in combo box.

  The foreign key selector of a combo box, when used in a user service, seems to be broken.

• CP-18032 When the page becomes obsolete, clean the list of widgets kept in the session's page history

• CP-18166 A technical error occurs if a relative "login URL" parameter is defined

  A technical error occurs when accessing the default URL if the "Login URL" parameter is set to a relative path.

• CP-18282 It is not possible to enter a carriage return on a multi-line text field in grid-edit mode.

• CP-18289 In grid edit, it is impossible to select edited cell text with keyboard.

  In a standard form, user can select a text of a focused textarea by using SHIFT+LEFT/UP/RIGHT/BOTTOM. This cannot be achieved in grid edit.

• CP-18336 Unwanted consecutive slash characters in EBX® URLs.

  Unwanted consecutives slash characters appear in EBX® URLs, leading to URLs being blocked by security tools.

• CP-18397 Display subtabs on several lines in a custom form pane.

  Display subtabs on several lines in a custom form pane when there are too many tabs to fit on a single line. It was the behaviour prior to EBX® versions 5.9.

• CP-18416 UILabelRendererForHierarchyContext.setDefaultLinkEnabled does not work.

  The method UILabelRendererForHierarchyContext.setDefaultLinkEnabled(false) does not disable the default preview link for hierarchy node.

• CP-18446 Fix the focus style of the message box button in DMA, WMA, administration.

  The legacy message box button is not round like in new UI.

  The cause seems to be an optional counter which is not used in new ui, and should be deleted.

• CP-18448 Fix collapse/expand button localization.

  Collapse button is "collapse".

  Messages should start with a capital.

• CP-18449 Error when creating/duplicating a dataset with a very long name.

  Creating or duplicating a dataset with a name longer than 64 characters causes the display of an undetailed error.

  See screenshot enclosed.

• CP-18465 The space around error page is not in accordance with the selected density.

• CP-18519 An error occurs when changing the locale in a workitem opened from a UserServiceExtended.

  An error occurs when changing the locale in a workitem opened from a UserServiceExtended using UserServiceNext.nextWorkItem.

  The error only occurs if the service was launched in the context of a custom perspective.

• CP-18521 Application locking does not work when connecting to EBX® with credentials as HTTP parameters.

• CP-18643 Password field always causes a message prompt when leaving creation form

  This is the case even when the form is empty.

• CP-18659 Widgets on boolean not seen as updated when updated via a TableTrigger.

  The radio button group widget is not seen as updated in the GUI, when updated via a TableTrigger.

• CP-18766 When creating a record from a foreign key, the predicate filter is ignored

  When creating a record from a foreign key, the predicate filter is ignored.

  The value of the nodes referenced by the predicate should be initialized

• CP-18794 Performance issue with combo boxes, due to Ajax prevalidation.

  When Ajax prevalidation updates are executing on a form, it prevents every combo box of the form from displaying; this leads to performance issues.

• CP-18826 Fatal error on perspective navigation.

  When moving a service from a dataset to another in the same dataspace via perspective actions, a fatal errors occurs.

  [^StackTrace.txt]

• CP-18913 On completion of a workitem displayed in a root session web component, the user is redirected to the exit policy URL instead of the component redirection URL.

  On completion of a workitem displayed in a root session web component, the user is redirected to the exit policy URL instead of the component redirection URL.

  When no exit policy URL is defined, the user is correctly redirected to the component redirection URL.

• CP-18935 Improve the CSRF token mechanism for better protection against CSRF attacks

• CP-19006 A technical error occurs for a relative 'login URL' parameter, if 'useLocalUrl' is true

  A technical error occurs when accessing the default URL if the 'Login URL' parameter is set to a relative path, along with ebx.servlet.useLocalUrl=true in ebx.properties.

• CP-19089 [IE11] 'Access denied' returned to a JavaScript function in a portal integration

  'Access denied' returned to a JavaScript function when using IE 11 in a portal integration.

• CP-19287 [Firefox] An alert box can be erroneously displayed when leaving the page.

  An alert box can be displayed when leaving the page, whereas no modification was made to the current form.

  It occurs on Firefox and from a record opened from a foreign key preview button.

• CP-19310 A white screen is displayed when accessing a foreign key selector view.

  A white screen is displayed, and a NullPointerException is thrown, when accessing a foreign key selector view.

• CP-19332 Error when using relative path in addJS_* methods in WidgetWriter.

  Using relative path in the class WidgetWriter is not possible in this following methods:

  * addJS_getNodeValue

  * addJS_setNodeValue

  * addJS_setNodeMandatoryIndicator

  These methods should be able to detect if the input path has already been normalized. If not, the path must be normalized.

• CP-19354 An error page is shown when application locking is enabled, and using SSO non-admin login.

  An error occurs when application locking is enabled and a custom directory is used for SSO authentication by using RemoteUser or UserPrincipal in the method: com.orchestranetworks.service.directory.Directory#authenticateUserFromHttpRequest.

• CP-19376 Cannot set the 'mandatory' indicator for field, when using a custom widget on a record form.

  On a simple record form (without custom UIForm), if a custom widget is set to be displayed on a given field: in the widget, the method to set the 'mandatory' indicator does not work.

• CP-19460 Browser does not spellcheck the content of HTML inputs

  HTML inputs are used for long texts, and the user might want to let the browser spellcheck such input.

  An option should be added in administration, to enable spellcheck on HTML inputs.

• CP-19562 In the dataspace administration tree, the font color of the selected item does not depend on the background.

  In Administration > Dataspaces > Dataspaces, it is possible to open a dataspace to see its information.

  The selected dataspace is then displayed in the navigation menu as selected, on a dark background with the default theme.

  In this situation, the font color should be white, but is black there.

• CP-19764 [All browsers] A Javascript occurs when dragging the scroll bar on the 'Dataspace key' field under the 'Data services' area.

  When users open the drop-down list of the 'Dataspace key' field under the 'Data services' area, then drag its scroll bar, a Javascript error occurs.

• CP-19914 Work items in state "Started" and "Allocated" are displayed when users only choose the "Allocated" service on the Work items table.

• CP-19995 NPE when creating a record through a table reference declared by a BeanDefinition

• CP-20020 Documentation pane defined in User Service bean element is not displayed in form rows list.

  When form rows are not enclosed in  startTableFormRow/endTableFormRow methods, there are displayed in div.

  Compared to table form rows, a regression on documentation pane appears on field defined by a bean element where it displays "Object not found".

• CP-20120 The owner of a branch should see descendant hidden dataspaces.

  In order to be able to close a dataspace, a branch owner should be able to access any of its children, even if permission has been set to

  'hidden' in one of them.

• CP-20168 Textarea fields should not be limited to 1 million characters.

• CP-20271 Buttons on tabular view header do not show light icons when highlighted.

  On the tabular view, the selection/sort button and "?" button show dark icons when they are highlighted (clicked). They should show light icons.

• CP-20460 When a custom login page is implemented, the application locking page redirects to the login page.

  When custom login page is implemented, application locking page is redirecting to standard login page.

  The expected behavior is that the application locking page should redirect to the custom login page.

• CP-20559 Dataspace Selector pane border color should be the same as the input.

Users & Permissions

• CP-16472 Job title updated for capitalization in the default directory.

  There is no need to update the job title for capitalization in the directory. It leads to unwanted behaviors, ex: VP, Vp.

• CP-16719 Performance issue while computing the list of services accessible in a toolbar when the add-ons are activated.

• CP-16818 Performance issue while computing the list of services accessible in a toolbar when the add-ons are activated.

• CP-16850 Tracking information is not valued in the Access Rule execution.

  The tracking information declared in a workflow user task definition is null while building the navigation pane.

• CP-17339 The ActivationContext.setDisplayForLocations() behavior needs to be clarified.

  When calling the ActivationContext.setDisplayForLocations() with the ActionDisplaySpec.HIDDEN_IN_DEFAULT_MENU parameter without specifying any toolbar location, the user service is still displayed in the default menu.

  As it could be expected that such a call could hide the service in that default menu, the current behavior must be clarified in order to avoid any misunderstanding.

• CP-18400 The services menu is displayed incorrectly for a user with hidden permission in a dataset.

  When admin users set hidden permission for a non-admin user in a dataset, this non-admin user cannot view the dataset's information: the root node is not shown and the 'Delete' and 'Duplicate' services are missing.

• CP-19240 The documentation of UserServicePermission.getDisabled() could lead to confusion.

  The documentation of UserServicePermission.getDisabled() states that the "user service is neither displayed to the user nor can be executed by the user in the current session context".

  This may lead to confusion as ServicePermissionRule.getpermission() offers the ability to decide if the user service is disabled for display or for execution.

• CP-19329 Permissions for table view user services are not evaluated when displaying history.

  Programmatic permissions for table view user services are not evaluated when displaying history.