RE: Ignore *SYSTEM state of objects?? Search 400 article.

> I second that. Hacking programs to system state may be a common technique
> among SOME MI programmers, but it's one for which I've done my best to
> remain ignorant of the details, because (1) nearly everything I write is
> part of some commercial product that's expected to be
> Security-50-friendly, and (2) commercial products with hacks in them (the
> only reason I know of to patch to system state) tend to be less reliable
> (and less trusted) than those that stick to supported APIs.

And I can place myself in that camp too - nothing I have coded that has become 
part of a product that someone paid money for has ever needed to incorporate 
the patching of the objects state.

> While I still think it ridiculous to keep a system locked up with
> QFRCCVNRST set so high it won't accept non-observable programs compiled on
> an earlier release, this is a genuine security threat, and if some vendor
> is still putting out commercial products (or custom jobs, for that matter)
> that aren't Security-50-friendly, you're probably better off looking for a
> different software vendor.

I totally agree that there is no need to patch a program into system state for 
commercial products. If you re-read my emails on this patching subject, none of 
them advocate doing that.

--phil