> On 1 Nov 2024, at 7:28 AM, Roman Fischer <[email protected]> wrote:
> 
> Key-generation isn't that cheap and the larger the keys get, the more 
> expensive it gets.
> Also, the CA's would probably need to feed and query one central database of 
> "used" keys to prevent the re-use.

Private Key Compromise Transparency (PKCT), as mentioned in 
https://mailarchive.ietf.org/arch/msg/trans/tB8YhAapz_6RN9MJVMKlRCR9HK0/, might 
help to make this information available to all CAs.

>  
> -Roman
>  
> From: Mike Shaver <[email protected] <mailto:[email protected]>>
> Sent: Freitag, 1. November 2024 13:16
> To: Roman Fischer <[email protected] 
> <mailto:[email protected]>>
> Cc: [email protected] <mailto:[email protected]>
> Subject: Re: Assuming keyCompromise for unspecified-reason revocations
>  
> I guess I’m curious about the opposite question: why is it important that a 
> key be permitted to be reused ever? I understand that there are some 
> operational shortcuts that can be taken to avoid deploying a new key along 
> with a new certificate, but I’m not sure if there are other reasons to not 
> just require a new private key for every certificate. They are cheap to 
> generate!
>  
> I feel a nagging at the back of my brain as I ask this question, so I 
> anticipate that someone will provide a tremendously obvious answer.
>  
> Mike
>  
> On Fri, Nov 1, 2024 at 6:00 AM Roman Fischer <[email protected] 
> <mailto:[email protected]>> wrote:
> Dear Jamie,
>  
> I don't see what's the final goal here. Is it to prohibit anybody from 
> re-using any key that was ever involved in a certificate that was revoked 
> (any reason could be "false" or "mistaken")? Or also keys of certificates 
> that expired (not everybody who has a key compromise reports it or revokes 
> the certificate, they might just create a new key and let the cert expire)?
I was only suggesting to prohibit reuse of keys for certificates revoked 
without reason (“unspecified” reasonCode) as there is a chance they are 
compromised. 

Though, preventing overall reuse of any key seems like the safer approach but I 
don’t really know what type of negative implications it might bring.

-- 
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 visit 
https://groups.google.com/a/mozilla.org/d/msgid/dev-security-policy/F61FC7BF-E7BA-40DA-877C-3C6AE6605BFF%40gmail.com.

Reply via email to