On Thu, 11 Jan 2007 09:38:29 +0900 Georgi Georgiev <[EMAIL PROTECTED]>
wrote:
| Quoting Ciaran McCreesh <[EMAIL PROTECTED]>:
| > On Thu, 11 Jan 2007 09:07:54 +0900 Georgi Georgiev <[EMAIL PROTECTED]>
| > wrote:
| > | Further, by adopting ACCEPT_RESTRICT, it would be possible to be
| > | able to say: ACCEPT_RESTRICT=-sandbox: Do not let any ebuild touch
| > | anything outside the sandbox.
| > | ACCEPT_RESTRICT=-userpriv: Do not let any ebuild run with elevated
| > | privileges.
| >
| > Which gains what, exactly? These are not things about which the end
| > user should be concerned.
| 
| A user shouldn't be concerned if an ebuild wants to leave the
| sandbox when not supposed to?

Correct. *Developers* should be concerned about whether their package
installs and uninstalls correctly. If an ebuild is leaving the sandbox,
it's doing so because it's necessary (at least at present -- this
proposal will make it more like "because the developer couldn't be
bothered to fix something").

Remember that packages can still do bad stuff to the filesystem even
when sandbox is turned on. The point of sandbox is to be a safety
feature, not a security measure.

-- 
Ciaran McCreesh
Mail                                : ciaranm at ciaranm.org
Web                                 : http://ciaranm.org/
Paludis, the secure package manager : http://paludis.pioto.org/

Attachment: signature.asc
Description: PGP signature

Reply via email to