diff options
| author | Dominyk Tiller | 2014-10-18 00:48:57 +0100 |
|---|---|---|
| committer | Mike McQuaid | 2014-10-19 12:43:16 +0100 |
| commit | 73ca769de13b72682d0dcfe7d54f31f96fd5e5ca (patch) | |
| tree | c707f5b972d931d8a90dde63eaafeca62233ce60 /Library/Formula/boost-python.rb | |
| parent | 625fe7889b9dc8b95cd1ec0334d9d95ae2a9e77c (diff) | |
| download | homebrew-73ca769de13b72682d0dcfe7d54f31f96fd5e5ca.tar.bz2 | |
polarssl: killing sslv3
Proposing we take a sharp axe to sslv3 in this PR. Reasons:
1) We want to do this with OpenSSL as soon as practical. PolarSSL works in a roughly similar way but is much less widely used, only having one Homebrew dependency (and even that is optional), which makes it a nice platform for testing.
2) PolarSSL has no mechanism to mitigate the SSLv3 POODLE mess in the manner OpenSSL has. That mechanism isn’t available to PolarSSL, and consequently upstream’s recommendation [here](https://polarssl.org/tech-updates/security-advisories/polarssl-security-advisory-2014-03-poodle-attack-on-ssl-v3) is just “feel free to kill sslv3 like this”.
3) If anything goes horribly wrong with this change, and I don’t think it will FWIW, it’s a 2 minute job to revert and rebuild to fix. That’s the kind of risk bracket I feel relatively comfortable jumping into for once.
Diffstat (limited to 'Library/Formula/boost-python.rb')
0 files changed, 0 insertions, 0 deletions