I think all the fellow monks and readers here will be more than delighted to read your explanation about how to do this when you code a web app, for instance, that runs in a server managed by a third party and which needs to connect to yet another database server.

How do you propose such problem be tacked?

That said, I think the post makes it very clear that the proposed piece of code is not secure. /I/ happen to think that a well understood level of obscurity is better than no obscurity at all. Leaving your passwords inside the scripts is bad, bad, bad. But obscuring them as shown, at least has the benefit of forcing the sysadmin to do something deliberate to read them, which is /very/ useful to prove intentionality should the need arise.

Finally, I can only interpret your answers (perceived tone and lack of a real identity) as rudeness, which I believe does not have a place in the monastery. If you do not agree to a point, it is usually a well respected practice to elaborate your answers and provide reasonable alternatives. This also allows the comunity to identify your views with an identity, giving you a chance to see for yourself how your opinions fare among the rest of us.

BTW, I just realized that I wrote AC instead of AM in the first update. Sorry about that.

Best regards

-lem, but some call me fokat