Re: Initialization of static variables

["Carsten Flensburg" <flensburg@xxxxxxxxxx>]

Hello Joe,

> The problem is that I would like these variables reinitialized each time
> the program is called, but they are not (even if the program is bound
> with ACTGRP(*NEW)!).

- If the activation group that the program is running in is destroyed - and
that's the case when activation group *NEW is specified and the program
terminates - the storage that the static variables occupies is reclaimed.
And likewise, when the program is called again, a new, unique activation
group is created and the program's static storage is initialized.

This is why it is possible to call programs running in *NEW activation
groups recursively - all otherwise conflicting resourses are scoped to
individual, non-persistent activation groups that are destroyed when control
is returned to it's caller (regardless of LR setting).

- So your ACTGRP(*NEW) approach actually should work - that would also make
sense in regard to Barbara's statement about ILE-ish behaviour being
preferred over RPG-ish ditto when designing this aspect of the RPG/IV
runtime. How did you test the ACTGRP(*NEW) setting?

Best regards,
Carsten Flensburg



###########################################
This message has been scanned by F-Secure Anti-Virus for Internet Mail.
For more information, connect to http://www.F-Secure.com/