String variables may be assigned any text such as a line in a work file. They can be accessed globally across work files. Every string variable has a content at all times since it is assigned the character ''BLANK'' (X'40') on initialization.
In compatibility mode, string variables do not have a character set. Their contents are always interpreted in the current character set.
Even after its content has been deleted or it has been assigned an empty string, a string variable is always initialized with the character ''BLANK'' (X'40').
Note
Workarounds (save data in string variables, delete all work files, insert from the string variables) can be used in compatibility mode to change the character set between saving and inserting a string. In this case, the data to be copied is of course transferred unchanged. The result is not usually meaningful and will certainly be different from in Unicode mode. For example, after the sequence
@CODENAME EDF047
@CREATE 1 ' πλάτων '
@CREATE #s0:1
@DELETE
@CODENAME EDF041
@CREATE 1 #S0
line 1 of a work file in compatibility mode has the content Ðëáôùí.
In Unicode mode, EDT attempts to convert the string ' πλάτων ' into EDF041. This is either rejected or the substitute character is used.