----- Original Message ----- From: "Shane Kerr" <shane@time-travellers.org> To: <loganalysisat_private> Sent: Monday, October 29, 2001 1:50 AM Subject: Re: [logs] Logging standards? > On 2001-10-28 11:16:18 -0800, todd glassey wrote: > > And Shane, how exactly do you prove that no one violated that "LPAR > > Compartmentalization" you just heaved back at me. > > > > The whole point is that someone or something has to be empowered to > > verify the veracity of the acts taking place on a system otherwise > > there would be no need of this conversation group since there would be > > no logging because surely it violates that same privacy. > > > > I think the privacy you refer to is from "other users", but honestly > > how is it possible that you and the Systems Admins and Auditors would > > or could not know each other more intimately. > > > > So - I ask again - what real expectation of privacy is there in a Time > > Sharing System, other than by potential agreement between the users? > > This isn't the same question that you asked originally. You simply said > that users should have no expectation of privacy, because the scheduler > has to track what processes are doing. or some part of the system's auditing. > > To me, this is akin to saying you have no reason to expect > confidentiality from your physician because he has to know what ailments > you suffer from. How so - Your physician does not meet you in the middle of a crowded bus station to annouce the results of your tests for "the CLAP" or whatever STD you were referring to. Further while your records are private the reality of them is that some record that an event took place is not necesaarily saying that "you personally" got a STD so your model doesn't quite work. Still it ought to raise a few eyebrows based in its topic. But the net net is that I disagree. > Hogwash! How can I *prove* that my doctor isn't > sending reports of my various STD's to the press? You cant - > Well, I can't. That > doesn't mean I should expect my doctor to publish my private > information. The doctor is a bad model. Use something like the Doorman publishing documentation as to what time you came and went. Why is this an issue? > > If it was important, a system of external audits could be devised, using > the similiar forensic techniques to the ones that security folks do on > compromised systems. But I'm thinking this is only necessary for truly > critical systems, e.g. financial processing, medical data, Microsoft > Passport. Any system that contains data constrained by some legal framework needs an audit process to insure that this legal framework is adhered to. What more needs to be said here. That framework is the issue for you though it seems. That framework means that someone or something catalogued and corelated the events taking place under those constriants (don't get me wrong I actually agree with you that the human violator of my privacy is very different than the machine violator of my privacy) > Not that any of them do this ... probably because users think > they have no right to expect privacy. :( > > -- > Shane > Carpe Diem > > --------------------------------------------------------------------- > To unsubscribe, e-mail: loganalysis-unsubscribeat_private > For additional commands, e-mail: loganalysis-helpat_private > > --------------------------------------------------------------------- To unsubscribe, e-mail: loganalysis-unsubscribeat_private For additional commands, e-mail: loganalysis-helpat_private
This archive was generated by hypermail 2b30 : Mon Oct 29 2001 - 12:00:38 PST