| |
| Here is a nice list of things to do to improve tlsdate: |
| |
| |
| 1) hack the client handshake to not leak the clock to the server |
| set it to all zeros or something cute or something random |
| |
| 3) add HTTP GET request to avoid network fingerprinting |
| 6) skew the clock rather than slamming it |
| 11) verification of remote certificate for Tor nodes |
| 13) account for servers that do not send UTC (Microsoft sends local time) |
| 14) port to bssl, nss, gnutls, yassl, and other libraries |
| 15) starttls support (smtp, pop, imap, ftp, xmpp) |
| 16) ensure that 32bit time isn't near wrapping time on 32bit systems |
| 17) find others to audit it - we need more eyes! |
| 20) Add verification of remote servers by DANE/CAA DNSSEC protected records |
| 21) Integrate Chrome's CRL list into tlsdate |
| 22) Block revoked or bad certs such as MD5 inc. and others. |
| 24) Add OCSP check option |
| 25) Block weak signature algorithms |
| 26) Hard code block list of known horrible certs (extract from Chrome/FF) |
| 28) Check that extended key usage is empty, or includes TLS Server Auth |
| 29) extract the SubjectPublicKeyInfo from the certificates; match against |
| public keys |
| 31) Confirm HTTP and TLS date is within a sane range |
| 32) Integrate tack support https://github.com/tack/tackc |
| 33) Implement checking of RFC 2818 style wildcards: |
| http://wiki.cacert.org/WildcardCertificates |
| 35) seatbelt profile for Mac OS X |
| 36) SELinux policy for GNU/Linux platforms |
| 37) Port to some Windows operating system that anyone actually uses |
| 42) Unit-test everything |
| 47) Review cert validation and compare it with Chrome: |
| https://code.google.com/p/chromium/codesearch#chrome/src/net/base/x509_certificate.cc&l=500 |
| 48) Complain when server uses very weak DH group parameters |
| example weak server: https://demo.cmrg.net/ |
| 49) Add seccomp tlsdate and tlsdate-helper |
| 50) Add AppArmor fixes for Tails |
| 52) Port tlsdated to FreeBSD and other non GNU/Linux systems |
| |
| Patches welcome! |
| |