Firewall Wizards mailing list archives
Re: Trusted Unices Aren't?
From: Rick Smith <rick_smith () securecomputing com>
Date: Fri, 30 Oct 1998 17:09:15 -0600
At 10:59 AM 10/29/98 -0500, Gordon Greene wrote:
...... I think Multics was the first OS designed with MLS in mind. I don't have direct experience with it, so I don't know how far it got, but I distinctly recall Multics being the origin of security labels in an OS.
The first mention I remember of security labels was in an obscure system called the ADEPT-50, which showed up before Multics actually ran (mid 60s instead of late 60s). The AIM labeling work on Multics took place after the basic OS was designed, and of course the B2 evaluation didn't happen till the mid '80s. Multics did some serious innovation, but it can't fairly claim MLS. I think it's arguable that only a handful of systems were ever designed from the ground up for MLS, and those were A1 or A1 candidate systems: SCOMP, LOCK, GEMINI, and perhaps some custom things like Boeing SNS and BLACKER. All others tried to graft MLS onto a polyglot of existing architectural constraints, usually Unix based. Rick. smith () securecomputing com
Current thread:
- Sidebar on a historical security model, (continued)
- Sidebar on a historical security model David Collier-Brown (Nov 02)
- Re: Trusted Unices Aren't? ark (Nov 02)
- Re: Trusted Unices Aren't? Gordon Greene (Nov 02)
- Re: Trusted Unices Aren't? Gordon Greene (Nov 02)
- Re: Trusted Unices Aren't? dreamwvr (Nov 02)
- Re: Trusted Unices Aren't? dreamwvr (Nov 02)
- Re: Trusted Unices Aren't? Paul D. Robertson (Nov 02)
- Re: Trusted Unices Aren't? dreamwvr (Nov 02)
- Re: Trusted Unices Aren't? Rick Smith (Nov 02)
- Re: Trusted Unices Aren't? Paul D. Robertson (Nov 02)
- Re: Trusted Unices Aren't? Rick Smith (Nov 02)
- Re: Trusted Unices Aren't? Rick Smith (Nov 02)
- Re: Trusted Unices Aren't? Paul McNabb (Nov 02)
- Re: Trusted Unices Aren't? Rick Smith (Nov 02)