Package Manager Denies Permission to Install

[Kam Leo]

On Wed, Jan 21, 2009 at 2:25 PM, Craig White <craigwhite at azapple.com> wrote:
> On Wed, 2009-01-21 at 13:42 -0800, Kam Leo wrote:
>> On Wed, Jan 21, 2009 at 12:47 PM, Patrick O'Callaghan
>> <pocallaghan at gmail.com> wrote:
>> > On Thu, Jan 22, 2009 at 2:12 PM, Kevin Kofler <kevin.kofler at chello.at> wrote:
>> >> Richard Hughes wrote:
>> >>> Sure, but my point if that GTK code is untrusted, and just not designed
>> >>> to be run with elevated privileges. A buffer-overflow is an easy exploit
>> >>> if the code is running as uid 0, whether running as setuid or as root.
>> >>
>> >> Why would you overflow a buffer on your own machine where you're already
>> >> root? It makes sense to attack a setuid binary on a machine you're not root
>> >> on, but it doesn't make sense to attack your own machine.
>> >
>> > Really?  In that case I invite you to visit my website evil.com and
>> > click on a few links. Better still, log into my friendly server and
>> > run a few of my apps. They're running on my machine, not yours. Of
>> > course the GUI runs on your machine via X11 ...
>> >
>> > poc
>>
>> A GUI is not required to compromise a machine. No need to go after the
>> root user either. Take a gander here:
>> http://www.linuxsecurity.com/content/blogcategory/89/102/7/0/
> ----
> well now...there's a cogent argument.
>
> Suggesting that even though few of the applications that run on X are
> audited for security when run as superuser, that it becomes acceptable
> to do so because other exploits exist that don't require X to propagate.
>
> Interesting logical expansion
>
> Craig

The fallacy is believing you automatically obtain security by auditing
applications running on X for execution by the superuser or preventing
root from logging into X.