Re: Geekspeak I don't understand
Peter Fairbrother <zenadsl6186-1HOZaDBbGgxaa/[email protected]> Thu, 14 May 2015 22:01:03 +0100
| Newsgroups | gmane.law.cryptography.uk |
|---|---|
| Message-ID | <[email protected]> |
On 14/05/15 18:23, Alan Braggins wrote: > On 13 May 2015 at 23:10, Jon Ribbens <jon+ukcrypto-JgzTmhv+UHffC7kMvaharFpr/1R2p/[email protected]> wrote: >> Although, now I think about it, I seem to recall that >> this kind of certificate pinning tends to have a deliberate exception >> in it for locally-added certificates. > > Certainly Chrome has - I haven't checked Firefox > https://code.google.com/p/chromium/codesearch#chromium/src/net/http/transport_security_state.cc&l=135 Yes, Firefox too - in Firefox the default value of 1 for the preference security.cert_pinning.enforcement_level allows locally-added certificates. https://wiki.mozilla.org/SecurityEngineering/Public_Key_Pinning There is a list of pinnings which are on by default - they come with the browser. They typically include TOR, Twitter, Google, Microsoft, some Mozilla and Firefox sites- but not Facebook. So far this seems to work well enough. Pinnings can include more than one CA, which is good. The problems start when pinnings for new sites are dynamically added to the list. Quote from https://wiki.mozilla.org/SecurityEngineering/Public_Key_Pinning/Implementation_Details : "Something is Broken, and we Think it's Pinning" -- Peter F