What are some scenarios where Capability Based Security might be thought of as unnecessary (but isn't/wasn't)?
What are some scenarios where Capability Based Security might be thought of as unnecessary (but isn't/wasn't)?
We made due with band-aids for a long time, but they've reached their limits. Virtualization, Containerization, were all approaches that offered extremely course grained capabilities, in that they allowed you to specify a virtual disk for the machine, and no other access.
We tried enumerating all the bad programs - virus scanners.
We tried blaming the users - "You clicked on a bad link", Why did you open that attachment, etc.
We tried blaming the programmers - You need to write "secure code"
We tried blaming the compilers - Rust, etc.
We blamed everything except the Operating Systems that were not fit for purpose.
That needs to change, or we'll never be able to secure or data.