- From: Thomas Roessler <tlr@w3.org>
- Date: Wed, 26 Mar 2008 09:13:20 +0100
- To: Johnathan Nightingale <johnath@mozilla.com>
- Cc: Mary Ellen Zurko <Mary_Ellen_Zurko@notesdev.ibm.com>, "Yngve N. Pettersen" <yngve@opera.com>, public-wsc-wg@w3.org
On 2008-03-25 21:49:46 -0400, Johnathan Nightingale wrote:
>> <p>To derive a human-readable name from an AAC, user agents
>> MUST use the first of the following fields that is human-readable:</p>
>>
>> <olist>
>> <item>the Subject's Common Name (CN) attribute;</item>
> ...
>>
>> <item>the Subject's Organizational Unit (OU) attribute, in
>> combination with its Location (L) attribute;</item>
>> <item>the Subject's Organization (O) attribute.</item>
>> </olist>
>
> This seems like the inverse of intended order. To take Sierra Nevada
> Brewing's EV certificate as an example (https://www.sierranevada.com/):
ok, in that case I simply misread Yngve's mail -- I had understood
"in order of priority" to be "most important first"; happy to turn
that around.
> Accordingly, it seems to me that the order should be O, then CN,
> and honestly I'd leave OU out of it, but feel free to slot it in
> somewhere if you think it appropriate.
I don't actually care either way, but would like to hear from Yngve
before dropping that one.
>> <p>All Augmented Assurance Certificates MUST include
>> information that lets this algorithm terminate successfully,
>> i.e., return human-readable information.</p>
> I don't think we have it in our scope to offer normative language on how CAs
> issue certificates, (nor would we want to, I suspect)
In general, I'd agree. However, I do think it's appropriate to
spell out fundamental assumptions in terms of conformance clauses,
e.g. in this case.
> so this should probably read in terms of our own internal
> definition. E.g.
> <p>By definition, an Augmented Assurance Certificate will include
> information that lets this algorithm terminate successfully, i.e.
> return human-readable information.</p>
> (Possibly with appropriate tweaks to our definitions section, as well).
The trouble I have with this is that we have users or site
administrators possibly designating trust roots as being AA / EV.
So it is presumably worth spelling out a requirement on those who
generate certificates that they want to market to these
administrators as being EV certificates. Of course, we can also add
the implementation flip side, e.g.:
If the algorithm does not lead to human-readable information,
user agents MAY consider a certificate to be a usual validated
certificate, not an Augmented Assurance certificate, even if it
chains up to a <termref
def="id-augmented-qualified">AA-qualified</termref> trust root.
Cheers,
--
Thomas Roessler, W3C <tlr@w3.org>
Received on Wednesday, 26 March 2008 08:26:31 UTC