Re: mutt 1.5.24 released

2015-09-01 Thread Phil Pennock
On 2015-09-01 at 12:08 -0700, Brendan Cully wrote: > [1] http://www.mutt.org/kevin.key > GPG fingerprint: 8975 A9B3 3AA3 7910 385C 5308 ADEF 7684 8031 6BDA % curl -I http://www.mutt.org/kevin.key HTTP/1.1 404 Not Found I used `gpg --recv-key 0xADEF768480316BDA` so I'm good, I'm not after a w

[PATCH] TLS SNI support

2009-06-14 Thread Phil Pennock
e + + -- Tommi Komulainen tommi.komulai...@iki.fi -Updated by Jeremy Katz -ka...@linuxpower.org +Updated by: + Jeremy Katz + Phil Pennock diff --git a/mutt_ssl.c b/mutt_ssl.c --- a/mutt_ssl.c +++ b/mutt_ssl.c @@ -323,6 +323,20 @@ static int ssl_negotiate (CONNECTION *co SSL_set_mode (ssld

[patch] Send TLS SNI when setting up TLS

2011-08-01 Thread Phil Pennock
t as a client for the SMTP/TLS side of that. Regards, -Phil # HG changeset patch # User Phil Pennock # Date 1312190773 14400 # Branch HEAD # Node ID a1e4667211b4c5f2dcca06ad8f352cb86dac94c3 # Parent b01d63af6feab5e0b791ad12b04a1068d6a41cf6 OpenSSL SNI support diff --git a/README.SSL b/REA

Gmail custom search support

2011-08-01 Thread Phil Pennock
at the above URL documenting a capability string of "X-GM-EXT1" but real connections seeing a capability string of "X-GM-EXT-1". I added logic to handle capability aliases so that mutt can support both. Ah, the joys of site-local extensions. Regards,

Re: [patch] Send TLS SNI when setting up TLS

2011-08-01 Thread Phil Pennock
On 2011-08-01 at 10:53 -0700, Brendan Cully wrote: > On Monday, 01 August 2011 at 07:18, Phil Pennock wrote: > > +#if OPENSSL_VERSION_NUMBER >= 0x0090806fL && !defined(OPENSSL_NO_TLSEXT) > I'd prefer an autoconf test for SSL_set_tlsext_host_name instead of > the ve

[patch] fix TLS initialisation for OpenSSL1.0.1

2012-05-23 Thread Phil Pennock
Folks, OpenSSL 1.0.1 introduced support for TLSv1.1 and TLSv1.2. These are not used by mutt. This patch fixes that. Counter-intuitively, the OpenSSL folks have TLSv1_client_method() negotiate *only* TLSv1.0, and SSLv23_client_method() remains the only method which can negotiate different versio

Re: Design choices

2012-08-03 Thread Phil Pennock
I'm mostly avoiding this conversation, as I avoid most, because my mutt contributions are at the level of "several patches, only a few accepted into mutt itself". One point I'll comment on, with my MTA developer/maintainer hat on: On 2012-08-02 at 16:38 +0200, Vincent Lefevre wrote: >

Re: [patch] fix TLS initialisation for OpenSSL1.0.1

2012-12-17 Thread Phil Pennock
On 2012-05-23 at 15:57 -0400, Phil Pennock wrote: > OpenSSL 1.0.1 introduced support for TLSv1.1 and TLSv1.2. These are not > used by mutt. This patch fixes that. > > Counter-intuitively, the OpenSSL folks have TLSv1_client_method() > negotiate *only* TLSv1.0, and SSLv23_client_m

Re: toward a 1.5.22 release

2012-12-18 Thread Phil Pennock
On 2012-12-18 at 14:06 -0800, Michael Elkins wrote: > My current plan is to ensure that all the bugfixes that are > applied to the mutt packages in various OS distributions have been > rolled into the hg repo, then push out a 1.5.22 release. > > I've found the patches from some distros here: >

FreeBSD mutt segfault, keymap, empty map->keys

2014-10-11 Thread Phil Pennock
Looking for pointers to where to investigate more closely as I chase this down. Started getting repeatable mutt crashes, if I leave mutt running for a while and come back to it; session is IMAP using Kerberos auth. Running on FreeBSD, mutt built using Ports (poudriere). The crash is in keymap.c.