ti.sql presently depends on init.sql, because it needs the initial value of &_ti_sequence. Would it make sense then to explicitly call init.sql within ti.sql, as the script hangs if we forget to do that?
Alternatively, I think I could come up with a solution that doesn't need any initialization. The idea is to generate a new value for tracefile_identifier based on the parsed current value, as opposed to maintaining a counter in a sqlplus variable.
ti.sql presently depends on init.sql, because it needs the initial value of &_ti_sequence. Would it make sense then to explicitly call init.sql within ti.sql, as the script hangs if we forget to do that?
Alternatively, I think I could come up with a solution that doesn't need any initialization. The idea is to generate a new value for tracefile_identifier based on the parsed current value, as opposed to maintaining a counter in a sqlplus variable.
What do you think?