Ryan wrote:
> However, an additional consideration is that keyCompromise revocations are 
> likely more valuable than other forms of revocations, both in terms of 
> efficient and timely delivery and in user risk.

I would still like to know what use cases (if any) the consumers of revocation 
information have for differentiating between the "other forms of revocation".  
If consumers of revocation information are, at most, going to create two 
buckets - one containing keyCompromise revocations, the other containing 
everything else - then ISTM that there is no value in mandating the use of 
affiliationChanged, superseded, cessationOfOperation, and privilegeWithdrawn; 
and if so, then I think CAs should be permitted to omit reasonCodes for 
non-key-compromise revocations.

________________________________
From: [email protected] <[email protected]> on 
behalf of Ryan Sleevi <[email protected]>
Sent: 02 February 2022 22:35
To: Doug Beattie <[email protected]>
Cc: Kathleen Wilson <[email protected]>; [email protected] 
<[email protected]>; Ryan Sleevi <[email protected]>
Subject: Re: Revocation Reason Codes for TLS End-Entity Certificates


CAUTION: This email originated from outside of the organization. Do not click 
links or open attachments unless you recognize the sender and know the content 
is safe.



On Wed, Feb 2, 2022 at 7:25 AM Doug Beattie 
<[email protected]<mailto:[email protected]>> wrote:

Will the CA block further issuance when the request for revocation does not 
include PoP which could DoS them for renewal using the same key pair?   To me, 
if the subscriber can’t provide PoP of the private key the unspecified reason 
code would be more accurate.  What’s the value to the subscriber, CA and 
ecosystem to treat that case as key compromise vs. unspecified?



I’m probably just not understanding the background and value for the second 
rule around processing requests for revocation with key compromise without PoP.

As hopefully my reply to Aaron captured a little, it's about where the burden 
rests.

Today, for most TLS issuance, no POP is required. That's because TLS itself 
doesn't need a POP, because it's an online protocol - the POP is delivered 
in-band, and it's not an identity-attestation system based on a directory (e.g. 
compared to S/MIME, where a sender needs to look up your public key ahead of 
time to encrypt something to you)

So the functional change of requiring the POP is that very few people, today, 
could request keyCompromise without doing more work. That's not ideal.
Further, however, is that for situations that are not uncommon, such as 
malicious deletion or ransomware, there is zero guarantee the victim would be 
able to prove keyCompromise at that time.

This is very similar to the discussions in the past of how many hoops a CA can 
place to request revocation (i.e. "You can only request revocation on the fifth 
Tuesday of every February under the full moon"). For Subscribers, and users, 
this matters.

However, an additional consideration is that keyCompromise revocations are 
likely more valuable than other forms of revocations, both in terms of 
efficient and timely delivery and in user risk. A policy that restricts when 
and how Subscribers can request this revocation is thus one that limits the 
value of that, by making it harder, which harms end users more. The more 
barriers placed for Subscribers, the harder it is to get this information in a 
timely fashion.

So, to end users, it's ideal where Subscribers can request any revocation 
reason that they want (... within reason), and for imposing obligations on CAs, 
to use particular revocation reasons when they're made aware, either internally 
or by externally reports. That protects users the most.

However, because there's no POP, that does offer _some_ abuse scenarios from 
malicious entities wishing to abuse the policies, and so some safeguards are 
needed. The question posed to this group is, seemingly, do we want to throw the 
baby out with the bathwater? Namely, should Subscribers have flexibility to (as 
easily as possible) request the method of their choice, or is the risk of abuse 
too great to trust them?

I'd like to find a solution where we can empower Subscribers as much as 
possible, because that can help protect Users the greatest. I think you're 
right that we want to figure out how to narrowly scope the abuse scenarios, so 
definitely, thanks for raising 6.1.1.3. We should try to find a way to best 
balance things, don't you agree?

--
You received this message because you are subscribed to the Google Groups 
"[email protected]" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to 
[email protected]<mailto:[email protected]>.
To view this discussion on the web visit 
https://groups.google.com/a/mozilla.org/d/msgid/dev-security-policy/CAErg%3DHEaptK1R7PMBFNYvp6_hHcMD4FRL7C3b4OP1yK2YaJFzQ%40mail.gmail.com<https://nam04.safelinks.protection.outlook.com/?url=https%3A%2F%2Fgroups.google.com%2Fa%2Fmozilla.org%2Fd%2Fmsgid%2Fdev-security-policy%2FCAErg%253DHEaptK1R7PMBFNYvp6_hHcMD4FRL7C3b4OP1yK2YaJFzQ%2540mail.gmail.com%3Futm_medium%3Demail%26utm_source%3Dfooter&data=04%7C01%7Crob%40sectigo.com%7C725715f3c42d4427d58e08d9e69c4c10%7C0e9c48946caa465d96604b6968b49fb7%7C0%7C0%7C637794382886534220%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C3000&sdata=9ci%2BzraTiaeG3cQ7vKvXSDF6DKE%2FLzdsjPIHiEplPa4%3D&reserved=0>.

-- 
You received this message because you are subscribed to the Google Groups 
"[email protected]" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/a/mozilla.org/d/msgid/dev-security-policy/MW4PR17MB47293D9D086787BD3B7EB74EAA299%40MW4PR17MB4729.namprd17.prod.outlook.com.

Reply via email to