Releases prior to 7.7.06 pre-load various samples into the IBISAMP application. As of 7.7.06, on a new installation, the IBISAMP application is created, but is not pre-loaded. The server Web Console has a new feature on the ribbon and on the application tree (under new), Tutorials (the Create Tutorial Framework page), which has a pull-down for various samples. The DMC also has this feature on the application tree.

There are currently about 10 different tutorial/sample selections available on the pull-down select list to match various customer needs. The bulk of the prior IBISAMP sample objects can be generated by selecting the Create Legacy Sample Tables and Files tutorial. Other prior IBISAMP DataMigrator sample objects (usually starting with the characters dm*) are now loaded by choosing their respective DataMigrator tutorials. Under the new method, the tutorials/samples may be loaded to any application, not just IBISAMP.

If you are doing just a software refresh, the prior IBISAMP objects will be unchanged (because a refresh does not touch app directories).

The Java Listener on OpenVMS Alpha has pre-set values for Initial Java Heap Size (JVM_INIT_HEAP) and Maximum Java Heap Size (JVM_MAX_HEAP) because the internal Java defaults for these values are insufficient (and cause features to fail). Plus, the OpenVMS Alpha Java documentation explicitly states that the internal defaults should not be depended on.

The pre-set values do not reflect any tuning, but are simply known working values. Specific tuning should be done based on your application need. Java’s own internal default for this heap size is insufficient and in-turn causes server features to fail. The pre-set value does not reflect any specific tuning, but is simply a known working value. Specific tuning should be done based on application need.

Since OpenVMS Integrity (IA64) Java does not have issues with these internal defaults, they are not pre-set. However, applications may benefit from specific tuning.

As of 7.7.02, OpenVMS 8.x builds are natively IEEE builds and the SET VMSFLOATCONV = IEEE/G-FLOAT command is used to SET the float read method for single (F)/double (D) float data. This requires separate configurations for each float type and a hub-sub arrangement to join data By default, an IEEE build needs no explicit SET command to access IEEE float data, although it is allowed. To access G_FLOAT data, the following command must be added to the server edasprof.prf:

SET VMSFLOATCONV = G-FLOAT

This may be done from the Web Console using the server profile edit feature.

The SET command is ignored on non-applicable platforms, so application sources may be shared. The setting also does not apply to the OpenVMS 7.x builds, which are strictly G_FLOAT builds.