In <[log in to unmask]> [log in to unmask] writes:

> > For this kind of application, we can generate a encryption routine that will
> > be essentially unbreakable, so long as we don't tell anyone what the algorit
> thm
> > is. We can do this for the QC screen because we're running in a pure client-
> > server mode, albeit in an extremely thin-client mode (which is the new, mode
> ern
> > jargon for a host-terminal mode, which is where we started 30 years ago :-).
> .
> > We're running an HP3000-based program when we're in QC -- and in this
> > instance, we can clearly design both ends to match.
>
> For reliable, free (patent free/royalty-free/free source) encryptyion that you
> can actually *use* on the 3000, DES (or triple-DES) would be an easy to
> implement solution (again, it's free and public domain source code for DES
> compiles easily on the 3000). All patents have expired on it, it's VERY
> secure (with sufficient keysizes), and it's a proven/tested encryption
> technology that's available on all platforms.

[snip]

I guess I should also point out that as soon as you include any encryption
in your product, your product falls under the U.S. export restrictions on
munitions (which suck...to put it mildly). If you include a worthwhile
encrytpion technology (any DES with >56bit keys) you can't distribute your
code outside the U.S./Canada. If you use the legally exportable stuff (which
is mostly a waste of time as it's trivial to crack nowadays) you might as well
have not bothered. *sigh*

3k is part of one of the major lobbying groups (the ACP - Americans for
Computer Privacy) that's trying to get this crippling status improved, but
it's an uphill battle.

  -Chris Bartram