Re: Authentication, cancels, etc

New Message Reply About this list Date view Thread view Subject view Author view

From: Charles Lindsey (chl@clw.cs.man.ac.uk)
Date: Tue Oct 02 2001 - 05:45:52 CDT


In <8A2IRGkZcDB@3247.org> list-ietf-wg-apps-usefor@faerber.muc.de (Claus Färber) writes:

>Brad Templeton <brad@templetons.com> schrieb/wrote:
>> On Sun, Sep 30, 2001 at 11:47:00AM +0200, Claus Färber wrote:
>>> Brad Templeton <brad@templetons.com> schrieb/wrote:
>>>> Quite possibly, but I really don't see more than the tiniest
>>>> fraction of sysadmins wanting to do this sort of detailed
>>>> configuration.
>>>
>>> So they trust someone (X) which does that for them and tell their
>>> server to accept such messages from everyone certified by X.

>> That was what I was saying. No need for a complex UI and format to
>> describe the reasons for cancels.

>No, there's a slight difference:

>Your solution would only allow differntiating on a per-canceller
>basis.

>I proposed a solutions where certificates would be valid on a per-
>canceller-reason basis.

No, you are confusing two issues. The purpose of the signature is to
establish that the cancel/nocem/whatever indeed came from Alice, and not
from Hipcrime. It has nothing to do with authority.

The separate question is which attributes/reasons given by Alice (and you
now know it was indeed from Alice) you should decide to act upon. That is
just a matter of keeping on file a list of known issuers, and the
attributes/reasons you will accedpt from each (so, for Alice, you might
accept her "spam" attribute, but not her "patent" attribute).

There is indeed a further issue of how you decide which attributes to
accept from which issuers, and there is a place there for "advisors" (such
as Rosslyn Hengeveld) who will publicly announce that Alice is good for
spam cancels, but not for patent cancels except within the USA). Doubtless
the advisor will sign her recommendations, but they should be communicated
in a separate channel from the actual cancels/nocems.

It is, in essence, no different from deciding whose newgroup control
messages you will accept. In practice, most people will believe a signed
control message if the issuer/key is on Tale's list.

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133   Web: http://www.cs.man.ac.uk/~chl
Email: chl@clw.cs.man.ac.uk      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5


New Message Reply About this list Date view Thread view Subject view Author view


This archive was generated by hypermail 2b29.