[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Sensitivity Labels
Gentlefolk,
I propose that we officially remove the recommendation for Sensitivity
Labels from RFC-1825, for several reasons:
A) Although there are many (> 6) interoperable implementations of
RFC-1828 and RFC-1829, none of them implement Sensitivity Labels.
B) Since RFC-1828 and RFC-1829 are more than ready to go to Draft
Standard, but interoperability of Sensitivity Labels has not been
demonstrated, by RFC-1602 we MUST remove Sensitivity Labels from our
official WG documents.
C) Sensitivity Labels are ill-defined.
D) Commercial vendors have not found a demand for Sensitivity Labels.
> From: Ran Atkinson <rja@cisco.com>
> There are at least 2 independent implementations of RFC-1825 that include
> support for sensitivity labels.
>
Please indicate which implementations?
And if you cite NRL, please detail commands for manual configuration of
the security association that implement the feature.
> There is no plan to move RFC-1825 through RFC-1827 forward prior to
> or during LA in any event.
>
Then, you have not followed the Standards Process in RFC-1602. The time
for updating them is upon us.
Last Spring, we (the WG) allowed Ran Atkinson to publish RFC-1825 to
RFC-1827 without removing some of the nits, in order to make forward
progress. Since then, he has been browbeating Photuris with an obscure
"requirement" (clearly "RECOMMENDED" not "REQUIRED" in RFC-1825) for
Sensitivity Labels.
There have been two respondents that have called for Sensitivity Labels
(in Photuris) in the past 3 weeks:
1) Ran Atkinson
2) Theodore Ts'o
I have spoken to the latter, and he assures me that his interest is in
having a "level playing field" for the protocol descriptions, not in
actually implementing Sensitivity Labels; nor has MIT any requirement or
current plans for deployment.
I will note here that someone has sent private messages "quoting" me as
stating: "i don't care for sensitivity labels". I have examined both
the IPSec and my personal archives, and cannot find this string at any
location. Dissemination of such false and out of context quote
information is excrable.
My previous _privately_ expressed position was "Personally, I don't
care. It is a pretty simple thing to add syntactically, but takes 10-20
pages to describe all the security considerations."
I will note that I have previously cooperated with regard to Sensitivity
Labels. Last September, at the request of Ran and NSA, I added the
"Modify" capability to Photuris. The sole reason was to modify
Sensitivity Labels on the fly. Last October, the WG, including
Sommerfeld, Orman, Bellovin and Housley (citing earlier IEEE arguments),
called for the removal of the Modify capability.
As to WG consensus, no public support (other than Atkinson) was
expressed for inclusion of Sensitivity Labels in Photuris. Therefore,
it was removed to the "for future work" Extensions draft.
Bill.Simpson@um.cc.umich.edu
Key fingerprint = 2E 07 23 03 C5 62 70 D3 59 B1 4F 5E 1D C2 C1 A2