Messages in this thread | | | Subject | Re: Flame Linus to a crisp! | From | Valdis.Kletnieks@vt ... | Date | Thu, 24 Apr 2003 01:56:40 -0400 |
| |
On Thu, 24 Apr 2003 06:39:50 BST, viro@parcelfarce.linux.theplanet.co.uk said:
> Excuse me, but I don't get the last part. You know that > F had been built in environment of unspecified degree of security > from source that had been kept in <---> > written by programmers you don't know > who had been hired in conformace with criteria <---> > and released after passing QA of unknown quality (but you can bet > that they had missed some security holes in the past) > under a license that almost certainly disclaims any responsibility. > > Care to explain how does one get from the trust in above to "trusted to run"?
On top of which, if a buffer overflow is found, the exploit will run in the context of the signed program. What it *does* mean is that once the ankle-biting script kiddie breaks in, the kernel will hopefully refuse to run their unsigned exploits. [unhandled content-type:application/pgp-signature] | |