Purpose
The inactive nametab table consists of two tables, DDXTT and DDXTF.
Apparently for performance reasons, both were created as transparenttables in the database, but each with a RAW field for the crucialNametab headers and the field data. To read the inactive Nametab table,structures X030L and X031L were used as a template. The data was thenread using field symbols.
To Unicode-enable all programs that read or write to the inactivenametab table, the field symbol-based access had to be modified.
Since this meant editing all programs using DDXTT/DDXTF and since therehave long since been requests to be able to read (or change) theinactive nametab data directly in the database or analyze the datausing transaction SE16, we decided to replace the RAW fields in bothtables with explicit fields.
This requires that the inactive nametab table is not used except forupgrades, that is, that the table is empty at all times or does onlycontain editable entries which can be reproduced.
Sending the table in the 'normal' way is not possible, sinceDDXTT/DDXTF are used for the upgrade itself. During the upgrade, aconversion would then be requested for both tables, and both would beaccessed. In this case, the system would terminate.
This XPRA uses the inactive activation of DDXTT/DDXTF during an upgrade(both are saved as LIMU TABD with a modified structure). The XPRAregenerates the active nametab table from the active sources andperforms a DROP/CREATE on the database. The contents of DDXTT/DDXTF arewritten to a log in advance.

Features