Susanne ...
----- Original Message -----
From: "Susanne Okunick" <susanne.okunick@xxxxxxxxxxxxxxxxx>
To: "todd glassey" <tglassey@xxxxxxxxxxxxx>
Cc: <ietf-ltans@xxxxxxx>; <thomas.kunz@xxxxxxxxxxxxxxxxx>
Sent: Monday, September 10, 2007 8:31 AM
Subject: Re: draft-ietf-ltans-dssc-00 comments
Susanne
that doesn't make sense from either EU Laws or those of the US as well.
Sorry if this is a like a bucket of very cold water in the face, but its
time to wake up and smell the coffee... Any long term document storage
and management protocol MUST take into account any and all laws which
would constrain the operations and formulation of the proofing models it
(said protocol) produces.
You are right.
Thanks!
And within that global requirement, all evidence-standards pretty much
require a 'full chain of custody' to be proven and without that, their
content will not be admissible in the US Courts and any Court's the US
has Joint Judicial MRA's with(Mutual Recognition Agreements). As such
the entire chain of custody and each signing and resigning needs to be
resolved and proven or there is an imperfect history for the LTANS
protected file.
While this may seem like its not something that is important the
'setting aside of this requirement' will in the US Make LTANS unusable
since the Court's wont accept it as a reliable method of storing
information. That means the ETSI TSA recommendation is PROBABLY also in
jeopardy since it clearly violates the intent and scope of those same
laws as well...
The requirement 'full chain of custody' does not contradict LTANS/our
draft.
But its not the focus or only form that the LTANS system can be used in -
creating wiggle room for introducing problems to the trust model and its
portability... The idea is to create ' trusted records which are
ultimately portable"...
The LTANS system is supposed to provide the "Trust Anchor" for those
records...
E. g. ERS does not prohibit to keep policies, protocols or whatever
needed for the complete evidence as required in the respective country.
No but it doesnt differentiate between hierarchical trust models and
flattened-out ones. I.e. ones compressed so only the most recent
trust-event is visible externally. .. and why this is important is that if
the document is routinely 'flattened' out - that is 'made such that there
is only one operable layer of policy and history' visible, then that chain
of custody is eliminated even if there is some history provided internally
to the file. Each 'stage' and each certification(IMHO) MUST be
reproduceable over the storage lifetime of the LTANS-stored Object.
Actually ERS allows the integration of such data.
yes butit gets converted to payload as opposed to keeping it as operable
policy.
The same is true for our draft: All policies additionally may kept by an
archive service and presented to court.
As to multiple proofs, the key issue is whether they are presented as an
integrated part of the document/signature proofing blob or as an external
'assurance' tool. My concern is that the Court's