RE: Security Code

["Dan Bale" <dbale@xxxxxxxxxxxxx>]

> -----Original Message-----
> From: midrange-l-bounces@xxxxxxxxxxxx / James H H Lampert
>
> Actually, even if one has a system that requires one to obtain an
> authorization code even for a trial installation, and that has an absolute
> expiration date built into the code, there's always the possibility of
> defeating such a system by simply racking the system date back, unless
> you've provided code to detect such activity (and also
> deletion/reinstallation) through the use of what I refer to as the "muddy
> pawprint" technique.
>
> --
> JHHL

Hopefully, this doesn't occur often, if ever, on production boxes.  Maybe
some would go through the trouble on developer/test boxes.

Muddy pawprints:  If you've got someone "weasel-ey" enough to set back the
system date, you're probably dealing with someone who knows how to find that
hidden "pawprint" on the system.  As far as detecting such activity, I'd
speculate that one starts entering into the realm of security level issues
and MI techniques that aren't allowed at certain security levels, but I
acknowledge that this is based on hearsay and no actual experience.  If one
is trying to market software, sales can be lost if you force QSECOFR signon
to install.

By any chance does IBM absolutely log every RSTLICPGM execution that always
stays on the system?  If I could be assured that I could query such a log to
determine whether my product has *ever* been installed, short of doing a
scratch install of the OS.

db