root encryption vs just /home encryption?

["Jóhann B. Guðmundsson"]

Ralf Ertzinger wrote:
> Hi.
>
> On Tue, 25 Mar 2008 02:38:38 -0700, Andrew Farris wrote:
>
>   
>> You're probably right about that, but they should be relevant.  If a
>> machine containing my information is lost/stolen I do not care
>> whether the company thinks their encryption on it was *probably* good
>> enough, I should be notified the information is out of their control.
>>     
>
>   
First of all company's should never allow *employees* to leave with
security/corporate sensitive data from the premise's in the first place. 
Be it on encrypted
or not laptop's or any portable media format.  
( But then again they should not be mailing them either :) )

Second of all if the company is <sarcasm>*smart*<sarcasm> enough to allow
laptop or other portable media that contains security/corporate 
sensitive data leave
the premise's in the first place and then when that *data* gets 
*misplaced*, all parties involved
should be notified that the *information* is lost immediately.

Time is of the essence here..

In reality the scenario is more like this..

John Doe loses or *misplaces* the sensitive data, ( or is asked to mail it )
wastes couple of hours looking for
( or the people at the post office )
it and then finally reports the lost *data*, that is if he does not 
report it the following morning
or he realizes that he's ( probably ) gonna get fired
( yep him not the CEO/Goverment employee that allowed this to leave the 
premises in the first place )
and wastes more hours reflecting on his current situation.
( Depends on which sector your working in if you get trained to handle 
these situation )

The report gets in what happens now...
Damage control meetings yea!!! let's waste more time on that..
Then couple of days ( if lucky, more likely week or more )
Parties/Clients/Public is notified of the data loss and the person that 
lost the data got fired and they are assured
the data was "encrypted" and "unaccessible " by any means known to man, 
and if so *unlikely* the data is in the hands of
a criminal then that criminal is made out to be a common thief and or a 
drug user finding ways to finance his next fix
(something "low crime" people can commonly relate to instead of the 
actual real threat )..

This has given the attacker more than enough time to execute the second 
stage of his attack and or disappear..

> Even assuming the memory-pull-attack is technically feasible and workable
> under non-lab conditions 
It it's.
> I (as an attacker) would rather go against weak
> passwords or use trojans to get your secret data. 
> I don't think that
> throwing all our resources on this specific attack is a good use of
> our time
I  think there are others protecting their asset(s) that are working on 
finding a solution
to this problem and if/when they manage to come up with one i'm sure it 
will find it's
way to the open source community....  

All I was suggesting that where you "hash" encrypt in anaconda there 
would be a notification
telling the user(s) that thou he encrypted the drive it would be 
vulnerable to "cold boot" attack.
something along with line it's better to encrypt but it's not secure 
even thou governments and corporates have claimed it to be.

No need to be promoting false security..

Best regards
                   Johann B.

-------------- next part --------------
A non-text attachment was scrubbed...
Name: johannbg.vcf
Type: text/x-vcard
Size: 365 bytes
Desc: not available
Url : http://lists.fedoraproject.org/pipermail/devel/attachments/20080325/fd180a01/attachment-0002.vcf