I think getting them revoked would be the first step.  If you make the data
available about which CAs still have 1024 bit certs or lower, we could email
the CAs and find out what is going on.

Jeremy

-----Original Message-----
From: dev-security-policy
[mailto:dev-security-policy-bounces+jeremy.rowley=digicert.com@lists.mozilla
.org] On Behalf Of Kurt Roeckx
Sent: Saturday, June 21, 2014 10:15 AM
To: mozilla-dev-security-pol...@lists.mozilla.org
Subject: Only accepting 2048 bit or better certificates

Hi,

The CA/B baseline requirement say that all RSA keys that are used since
since 1 january 2014 should have been at least 2048 bit.
All shorter than 2048 should have either expired or been revoked by that
date.  But it's still not the case.  We're currently around 0.24% of the
certificates that are being seen on the internet that still are too short.

I've made a graph of the progress of this at which you can see at:
http://www.roeckx.be/certificates/rsa_small_zoom.png

If I do a linear interpolation of the last 3 months it looks like we might
end up with 0% around January 2015, only 1 year after it was supposed to be
the case.  I hope the current trend stays that way.

But I would like to start enforcing the 2048 bit as soon as possible.  Do we
have some criteria for at which point we're willing to break compatibility?

There are still a few new certificates generated with 1024 bits.
I've been filing bugs about those and there were only a few so far this
month.  Maybe we can set a date from which we won't be accepting
certificates with a smaller than 2048 bit key generated after that date?

Should I put an effort into trying to get those certificates that are still
seen revoked?


Kurt

_______________________________________________
dev-security-policy mailing list
dev-security-policy@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-security-policy

_______________________________________________
dev-security-policy mailing list
dev-security-policy@lists.mozilla.org
https://lists.mozilla.org/listinfo/dev-security-policy

Reply via email to