To support what Jeremy and Rick said, I'll point out that it's entirely
possible to spin up multiple logs for different uses of X.509 certificates.
I understand that different X.509 certificates are used by different client
software, each kind of client software could use a different set of logs
(and these sets would be monitored differently depending on the
application).
Nothing prevents the logs to be based on the same software and
infrastructure.

Specifically:
* There's already a reference implementation for a CT log server as
open-source code (and we're working on a more robust version).
* Given Ben's interest in binary transparency and general transparency, I
believe he won't outright reject running extra logs for other cert types
(on Google's infrastructure).
* Code for general transparency service would allow running a transparency
log for arbitrary binary blobs. Ben may be able to share some details on
such an effort.




On Mon, Jul 21, 2014 at 9:44 PM, Jeremy Rowley <[email protected]>
wrote:

>  Because of the differences in root program requirements, operational
> requirements, and verification needs, from my experience, most client and
> code signing systems already fork from the SSL cert systems.  Client certs
> are especially different in their issuance process. For most CAs, I imagine
> it’s more work to add CT support for the three base certificate types than
> simply enable it for SSL.
>
>
>
> Jeremy
>
>
>
> *From:* Trans [mailto:[email protected]] *On Behalf Of *Kyle Hamilton
> *Sent:* Monday, July 21, 2014 2:00 PM
> *To:* Rick Andrews; [email protected]; [email protected]; [email protected]
>
> *Subject:* Re: [Trans] List of Roots Accepted by Log Servers
>
>
>
> Until there's adequate support in SCT for all types of certs, there is
>
> no drive to add SCT to other protocols and clients.  Therefore, there
>
> will never be a requirement for it.  Ultimately, the unwillingness to
>
> support it now shall only guarantee that it will never be required.
>
>
>
> But unless and until Google wants to step up and actively promote it, I
>
> guess Google gets a lack of it.  Therefore, according to your argument,
>
> Google now has de facto administrative control over Internet security
>
> infrastructure administration and buildout.
>
>
>
> Value might be perceived because of the lack of cost to modify ****all**
>
> issuance systems to be on the same code base, as opposed to forking them
>
> all.  However, as a reminder, forks increase maintenance cost and
>
> decrease capacity, capability, and interoperability across multiple
>
> systems -- particularly on previously-enabled systems that haven't yet
>
> been fully brought into production.
>
>
>
> -Kyle H
>
>  On 7/21/2014 8:08 AM, Rick Andrews wrote:
>
> True. The point I'm trying to make is that we're a long way from requiring 
> SCTs for all types of certs, SSL and non-SSL. Until all applications require 
> SCTs for all types of certs, I see no reason to add the complexity of logging 
> all certs. At this point in time, Google wants SCTs for EV SSL certs, and 
> that's what I'm shooting for. Chrome knows which roots are enabled for EV, 
> and so those are the only roots that log servers MUST accept.
>
>
>
> Other CAs might want to log all their certs. Relying parties might want to 
> push every cert they see into log servers. But I see value in limiting which 
> of my roots are accepted by log servers, and I'd like the ability to control 
> that.
>
>
>
> -Rick
>
>
>
> -----Original Message-----
>
> From: Gervase Markham [mailto:[email protected] <[email protected]>]
>
> Sent: Monday, July 21, 2014 2:33 AM
>
> To: Rick Andrews; Mehner, Carl; [email protected]
>
> Subject: Re: [Trans] List of Roots Accepted by Log Servers
>
>
>
> On 19/07/14 02:02, Rick Andrews wrote:
>
>  Unless I've lost control over my private keys, I don't need to monitor
>
> CT logs to see if anyone else has issued an SSL cert from one of my
>
> inappropriate roots. I can just look in my database.
>
>
>
> Not implying anything about your security, Rick, but one of the benefits of 
> CT is that people who don't realise that they've lost control of their 
> private keys can find out! :-)
>
>
>
> Gerv
>
> _______________________________________________
>
> Trans mailing list
>
> [email protected]
>
> https://www.ietf.org/mailman/listinfo/trans
>
>
>
> _______________________________________________
> Trans mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/trans
>
>
_______________________________________________
Trans mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/trans

Reply via email to