What's New in OpenEdge 12.4

ABL SAX-READER object

ABL developers can now use ABL object-oriented class methods as a callback target for the OpenEdge SAX-READER object. Before this release, only procedures could be used as callback target for the SAX-READER object. Developers can now use an object-oriented solution for handling callback events within their application and avoid mixing procedural and object-oriented code. Overloaded methods can also be specified as callback handlers.

For more information, see the the following topics:

• SAX callback reference
• ABL SAX callbacks
• EVENT-HANDLER-OBJECT attribute

SOAP web services

ABL developers can now use class methods as SOAP header callback targets for SOAP web service operations. Before this release, only procedures could be used as callback target to handle SOAP service callbacks. ABL developers can now register class methods as event handlers to use built-in ABL functionality to run SOAP web service operations asynchronously. This feature allows overloaded methods in ABL classes to be specified as callback handlers.

For more information, see the the following topics:

• SET-CALLBACK( ) method
• Statements to consume SOAP web services
• Attributes, methods, and events to consume web services
• Read an XML file using methods as callback

Support for expressions in VAR statement

Expressions are now supported during variable initialization using the VAR statement. Expressions can be used to initialize a scalar variable as well as parameters to a constructor when initializing a class variable. The ability to use expressions simplifies your code, allowing you to combine operations into a single statement.

The following example shows defining and initializing scalar variables using expressions:

VAR INT a = 50, b = 20, x = a + b , y = a - b, z = x - y.

The following example shows defining and initializing variables with a built-in function, a user-defined function, and a method as the expression:

VAR DATETIME dtm = DATETIME(TODAY,MTIME).
VAR INTEGER x = myFunction().
VAR INTEGER y = classMethod(output v).

The following example shows defining determinate and indeterminate array variables and initializing them using expressions:

VAR INT[2] w = [123, a + b].
VAR INT[3] x = [funct( ), 123, a + b].

VAR INT[ ] y = [123, a + b].
VAR INT[ ] z = [funct( ), 123, a + b].

The following example shows expressions used as parameters to a constructor:

VAR INTEGER x = 10, y = 20.
VAR point p1 = NEW point (x - 5, y + 5).

VAR point p2 = NEW point (classMethod() - offset, someobject:getylocation()).

For more information, see VAR statement or Learn about the VAR statement.

Support for instantiation and initialization of a class using the VAR statement

In this release of OpenEdge, you can now instantiate a class when you declare a class variable with the VAR statement by using the NEW function. This feature simplifies your code and allows you to do both operations in a single statement.

VAR myclass myobj = NEW myclass("Progress", 2021, ?).

For more information, see VAR statement or Learn about the VAR statement.

Support for array object variable instantiation and initialization using the VAR statement

In this release, you can now create an array of objects using the VAR statement. This feature simplifies your code, allowing you to combine multiple operations into a single statement.

The following example shows the instantiation and initialization of a determinate array object variable:

VAR StateClass[2] objArrayA = [NEW StateClass(“MA”), NEW StateClass(“NH”)].

The following example shows the instantiation and initialization of an indeterminate array object variable:

VAR StateClass[] objArrayB = [NEW StateClass(“MA”), NEW StateClass(“NH”)].

For more information, see VAR statement or Learn about the VAR statement.

This release introduces a new ABL AGGREGATE statement for computing aggregate values on the database server-side. Previously, aggregate calculations were performed client-side using both the ACCUMULATE statement and the ACCUM function. Now, you can simply use a single AGGREGATE statement for the COUNT, TOTAL, and AVERAGE aggregate operations. The AGGREGATE statement simplifies your code and improves performance as the computation is done efficiently on the database server-side.

The following example uses the AGGREGATE statement to calculate the average balance for users in Chicago:

VAR DECIMAL avgBalance.

AGGREGATE avgBalance = AVERAGE(Balance) FOR Customer
  WHERE Country EQ 'USA' AND City EQ 'Chicago'.

MESSAGE "Average balance: " avgBalance
  VIEW-AS ALERT-BOX.

For more information, see the AGGREGATE statement.

In this release, the AVM no longer ignores errors that occur when evaluating a query expression. Previously, if an error occurred, the value of the expression was set to the Unknown (?) value, and the query would continue to execute. This behavior affects all statements that involve query WHERE clauses, such as FOR EACH statements, FIND statements, and OPEN-QUERY statements. For a dynamic query, the error may not be detected until the first GET-FIRST/NEXT/PREV/LAST method after the query is opened. It is possible that mistakes in existing ABL will be uncovered by this change. Now, any error that occurs is raised and handled by the normal error handling mechanisms.

Here is an unrealistic, but illustrative, example:

FOR EACH customer WHERE custNum > INTEGER("A"):
...
END.

This code will now raise an error.

Previously, the code resolved to the following:

FOR EACH customer WHERE custNum > ?:
...
END.

To catch the error raised, use a CATCH block in the block containing the FOR EACH statement (not inside the FOR EACH block as that block has not actually started yet).

If using this feature causes any issues, you can use the -zNoQryExpErrors startup parameter to revert to the old behavior. Please note that this is meant to be used as a temporary work-around. It is possible that by using the startup parameter, you may occasionally return incorrect data from the query. Progress Software strongly recommends that in the long-term, you fix your code to avoid any new errors rather than use this undocumented startup parameter.

Support for .NET Core has been added to OpenEdge Open Client Toolkit. Open Client for .NET supports .NET framework 4.8, .NET Core 3.1, and .NET Core 5.0. Open Client applications can be run on .NET platforms that support .NET standard 2.0 targeted libraries. Progress Software also supports running Open Client application on Linux .NET core 3.1, Windows .NET core 3.1, and .NET framework 4.8 on Windows.

For more information, see .NET Open Client Support FAQ.

An ABL service is a set of business logic that can be accessed through a URL path.

For more information, see ABL Service Deployment.

Starting in OpenEdge release 12.0, Progress Software stopped shipping the 32-bit OE Studio product, which included AppBuilder. AppBuilder is a multi-purpose application development environment. It can be used as a visual programming environment to create applications. This release includes AppBuilder in both the 32-bit and 64-bit 4GL Development System products.

For more information, see Learn About AppBuilder.

Certain administrative tasks being performed on an active database are complicated by data changing while the task is underway, due to continuing inflight transactions and new logins. Data administration activities like mass load, purge, synchronizing replication source to target or making a large schema change can be easier to perform with data change paused. The solution is to temporarily restrict user activity and perform maintenance tasks without taking the database down.

This release introduces a restricted online mode that enables you to execute maintenance tasks without user activity and with all the buffer pool and helper processes still available, which improves the performance for these tasks. When this mode is enabled only database utilities can connect to the database.

For more information, see Restrict database access.

OpenEdge provides a mechanism for clients to automatically connect to an OpenEdge Replication target database, if the connection to the replication source database fails. In this release, PAS for OpenEdge agents can now reconnect to the new replication source when -autoReconnect is used.

For more information, see Automatic database reconnection with OpenEdge Replication.

Now you can quickly and efficiently truncate all the data associated with a particular table residing in Type II areas. The data storage space that was occupied by the table is returned to the storage area for reuse by other objects.

In this release, the PROUTIL utility lets you truncate, and optionally move, a table while the database is running, without having to recreate the schema.

For more information, see Truncate tables.

Space management for the before-image (BI) files was introduced in OpenEdge 12.3 with a mechanism that scans the BI file clusters for committed transactions and recyclable clusters. In this release, space management has its own before-image manager process, probim. To run BI space management in this release, you must set -biscantime to a value greater than zero and run the before-image manager (BIM).

probim <db-name>

For more information, see PROBIM command.

You can now remove a named field from the schema definition while still being connected to the database by means of the online Drop Field transaction. The data of the field in the database table is also effectively removed in the process without requiring any exclusive lock.

For more information, see Drop a field online.

You can now use the JDBC Cancel() API to terminate a SQL query in a JDBC client. When the JDBC client executes the API, the OpenEdge SQL server cancels the query and discards all running operations for that query without impacting other queries or applications. The JDBC Cancel() API can be used to cancel SQL SELECT queries that are taking longer than expected.

The Data Administration tool connects the two databases, compares the schemas, and generates a .df file that contains the changes required in order for the first database schema to match the second database schema. This is referred to as the delta.df.

Four sections are added to the new delta.df file. The sections explain the operations that are performed before and after the deployment of an application:

For more information, see Dump incremental delta.df supporting online schema changes.

By default, ABL and SQL create new data objects in the schema area, and the Data Dictionary creates new database objects in the next storage area after the schema area. To take advantage of the performance capabilities of Type II areas, you can specify a default area for the creation of tables, LOBs, and indexes, so that they are created in a Type II area instead of a Type I area like the schema area. You may also change the default object creation area at any time, whether or not the database is running. Default area assignments eliminate downtime when moving tables, LOBs, and indexes from a Type I to a Type II storage area.

For more information, see Set default storage areas for creating database objects.

You can now use the PROUTIL utility to configure or display default areas.

For more information, see PROUTIL SETOBJECTDEFAULTAREA qualifier.

To create objects in default areas, you can use the Index Editor option, Modify Table option or Field Editor Add option in the Data Dictionary.

For more information, see Schema Menu Reference.

The Data Dictionary displays the default area settings for database objects in the Database Properties screen and the Utilities menu’s Information option.

For more information, see Database Properties.

OpenEdge Replication now supports Transport Layer Security (TLS) for network communication between the source and target databases. This provides OpenSSL encryption for data on the move: data replication between the Replication Server and Replication agent, the Replication Transition process, and coordination of the replication processes.

This is independent of and complementary to OpenEdge Transparent Data Encryption (TDE) which is for data at rest.

For more information, see Transport Layer Security.

OpenEdge now supports client-side TLS authentication from an ABL client socket, in addition to the server-side authentication that was available in earlier OpenEdge versions. The OpenEdge system or security administrator can configure TLS client authentication from an ABL client socket, for a TLS-enabled connection, to any HTTPS server that requires TLS client authentication. Here HTTPS server includes PAS for OpenEdge instances and non-OpenEdge servers such as Apache HTTP Server and Microsoft IIS. This feature enables you to meet the enterprise security policy requirements.

The CONNECT() method of the Socket object connects a socket to the specified TCP/IP port on the specified host.

For more information, see CONNECT( ) method (Socket object).

OpenEdge databases that use Transparent Data Encryption (TDE) have an encryption policy for the database master key (DMK) known as the encryption DB policy, for short. You can perform TDE DB policy management tasks on your TDE-enabled databases while the database is running.

If you need to change the encryption DB policy for one of the preceding reasons, you can also change the cipher for certain database objects at the same time.

For more information, see Online TDE DB policy management.

An emergency shutdown of a multi-session agent is initiated when a severe error occurs. The shutdown process causes the multi-session agent to redirect all new requests to a new multi-session agent and records a message in the syslog (UNIX) or eventlogs (Windows) and generates a Protrace file. The shutdown process also creates a new file, pasoe-agent-crash-trace-process-id, in the $DLC/WRK/InstanceName/work directory.

The default wait time for the shutdown process is 10 minutes but it can be customized as required.

For more information, see Multi-Session agent initiated graceful shutdown.

In this release, the PROSHUT -replSync option offers a better way to enforce a guaranteed synchronization of all databases without any pending transactions after a broker shutdown.

For more information, see Enforce source and target database synchronization during shutdown.