I second that concern. I know at least one organization that is not
going to consider implementing key pining precisely for that reason.
They may consider a report-only mode, but as they were badly bitten by
browser bugs in the report-only mode implementation of
Content-Security-Policy that effectively changed report-only into
enforcing, they would wait with that until the implementation matures.

I think it would be easier to convince them if they could implement
key pinning themselves in incremental steps. For example, a browser
can provide JS API to access certificate chain and a possibility to
"pin" a particular JS source so it would be always executed from the
local cache.

On 25 May 2014 22:45, Yoav Nir <[email protected]> wrote:
> Hi
>
> We’ve had a few threads ([1]) on this list about administrators needing to be 
> careful when activating HSTS and/or HPKP, because a common mistake like 
> forgetting to replace the certificate in time can brick your site. The usual 
> conclusion was that companies and organizations without the proper technical 
> capacity to maintain a valid certificate should not use these features ([2])
>
> So today we got a real live demo how this can happen even to big, well-funded 
> companies ([3]). HSTS and HPKP were not involved - this was not a browser, 
> but a client that was programmed to hard-fail when receiving an expired 
> certificate. It is the equivalent of a site with HSTS, and if it can happen 
> to them, it could happen to anybody.
>
> I’m not saying we should revisit our previous decisions. Obviously Apple can 
> fix this now in minutes (probably already has), and all the consequences are 
> some bad press. I just think it is illuminating that even at large companies 
> there is no guarantee that an action that needs to be performed once a year 
> will be done correctly.
>
> Yoav
>
> [1] http://www.ietf.org/mail-archive/web/websec/current/msg01119.html (just 
> an example - there were many threads)
> [2] http://www.ietf.org/mail-archive/web/websec/current/msg01213.html
> [3] http://www.macrumors.com/2014/05/25/apple-software-update-invalid/
> _______________________________________________
> websec mailing list
> [email protected]
> https://www.ietf.org/mailman/listinfo/websec

_______________________________________________
websec mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/websec

Reply via email to