|
s2smanager: Don't tostring() the data before sending, sends2s already does this.
|
Matthew Wild |
Sat, 21 Nov 2009 13:47:45 +0000 |
|
s2smanager: Don't re-destroy destroyed sessions
|
Matthew Wild |
Sun, 14 Mar 2010 02:57:22 +0000 |
|
s2smanager: Add open_stream and close methods to resting sessions
|
Matthew Wild |
Sun, 14 Mar 2010 02:56:57 +0000 |
|
sessionmanager, s2smanager: Close session on </stream:stream>
|
Matthew Wild |
Thu, 11 Mar 2010 01:04:19 +0000 |
|
sessionmanager, s2smanager: Destroyed sessions are now simply resting (not dead) until they are collected - prevents a whole class of tracebacks
|
Matthew Wild |
Mon, 08 Mar 2010 02:13:41 +0000 |
|
Merge with 0.6
|
Matthew Wild |
Thu, 25 Feb 2010 18:03:15 +0000 |
|
s2smanager: Log and fail gracefully when unable to create socket for outgoing s2s connection (thanks foucault)
|
Matthew Wild |
Thu, 25 Feb 2010 17:53:27 +0000 |
|
s2smanager: Fix for the logic SRV record priority comparison (thanks darkrain)
|
Matthew Wild |
Thu, 25 Feb 2010 17:52:33 +0000 |
|
s2smanager: Use s2smanager log() if session doesn't have a logger (thanks Flo)
|
Matthew Wild |
Fri, 19 Feb 2010 03:30:27 +0000 |
|
sessionmanager, s2smanager: Fix for syntax of null_data_handler() (thanks Nolan)
|
Matthew Wild |
Fri, 12 Feb 2010 17:27:53 +0000 |
|
sessionmanager, s2smanager: Give sessions dummy data handlers that log when data is received by a destroyed session
|
Matthew Wild |
Fri, 12 Feb 2010 17:14:54 +0000 |
|
s2smanager, mod_compression, mod_tls: Changed event.session to event.origin for s2s-stream-features event for consistency.
|
Waqas Hussain |
Fri, 12 Feb 2010 04:30:17 +0500 |
|
s2smanager: Update for new server SSL API (don't pass a sslctx)
|
Matthew Wild |
Sun, 31 Jan 2010 15:41:33 +0000 |
|
s2smanager: Fix syntax error)
|
Matthew Wild |
Sat, 30 Jan 2010 18:45:46 +0000 |
|
s2smanager: require_s2s_encryption -> s2s_require_encryption
|
Matthew Wild |
Sat, 30 Jan 2010 17:00:47 +0000 |
|
s2smanager: Make require_s2s_encryption do what it says on the tin
|
Matthew Wild |
Sat, 30 Jan 2010 16:42:27 +0000 |
|
s2smanager: Whitespace
|
Matthew Wild |
Thu, 28 Jan 2010 13:39:23 +0000 |
|
s2smanager: Log error and fail gracefully if we can't open a socket for an outgoing s2s connection (thanks foucault)
|
Matthew Wild |
Tue, 19 Jan 2010 19:58:02 +0000 |
|
s2smanager: No-one noticed that Prosody stopped sending to/from on stream headers recently? :)
|
Matthew Wild |
Tue, 19 Jan 2010 04:35:10 +0000 |
|
s2smanager: Split sending of stream header into a :open_stream() method
|
Matthew Wild |
Sat, 09 Jan 2010 06:53:23 +0000 |
|
s2smanager: Allow new_outgoing() to create sessions without automatically connecting them (set 3rd parameters to false)
|
Matthew Wild |
Mon, 04 Jan 2010 17:32:26 +0000 |
|
s2smanager: Stamp a 'to' on the reply stream:stream. This makes prosody happy.
|
Paul Aurich |
Sun, 03 Jan 2010 11:26:36 -0800 |
|
s2smanager, hostmanager: Make dialback secrets per-host
|
Matthew Wild |
Sun, 03 Jan 2010 18:55:42 +0000 |
|
s2smanager: Trim trailing whitespace
|
Matthew Wild |
Sun, 03 Jan 2010 18:37:03 +0000 |
|
s2smanager: Give reasons to destroy_session in more cases, including when DNS lookup fails
|
Matthew Wild |
Thu, 17 Dec 2009 21:50:02 +0000 |
|
Merge with backout branch
|
Matthew Wild |
Thu, 17 Dec 2009 21:38:16 +0000 |
|
s2smanager: Send reply stream header only after checking we serve this host (sorry darkrain... :) )
|
Matthew Wild |
Thu, 17 Dec 2009 21:36:31 +0000 |
|
Backed out changeset 78c5cb163ea9, to fix another way. Haven't done this for a while :)
|
Matthew Wild |
Thu, 17 Dec 2009 21:33:35 +0000 |
|
s2smanager: More whitespace
|
Matthew Wild |
Thu, 17 Dec 2009 21:18:02 +0000 |
|
s2smanager: Mark stream as opened sooner, this prevents a duplicated stream header on host-unknown errors (thanks darkrain and his mail server)
|
Matthew Wild |
Thu, 17 Dec 2009 21:17:47 +0000 |
|
s2smanager: Whitespace
|
Matthew Wild |
Thu, 17 Dec 2009 21:14:44 +0000 |
|
s2smanager: Have both destroy_session and bounce_sendq accept a reason string to include in bounced stanzas
|
Matthew Wild |
Thu, 17 Dec 2009 18:59:29 +0000 |
|
s2smanager: Allow configuration of the dialback_secret in the config
|
Matthew Wild |
Thu, 26 Nov 2009 17:17:54 +0000 |
|
Merge with 0.6
|
Matthew Wild |
Wed, 25 Nov 2009 03:30:41 +0000 |
|
s2smanager: Log warning when trying to send a stanza from a host we don't serve, instead of a traceback (thanks stpeter)
|
Matthew Wild |
Wed, 25 Nov 2009 03:30:00 +0000 |
|
s2smanager: Fix syntax error introduced in merge
|
Matthew Wild |
Sun, 22 Nov 2009 04:53:02 +0000 |
|
Merge with trunk
|
Matthew Wild |
Sun, 22 Nov 2009 04:50:42 +0000 |
|
s2smanager: Update for new net.server API (s2s still doesn't work with libevent for other reasons)
|
Matthew Wild |
Sun, 22 Nov 2009 04:42:03 +0000 |
|
s2smanager: Don't log full stanza when sending outwards
|
Matthew Wild |
Sat, 21 Nov 2009 13:49:48 +0000 |
|
s2smanager: Don't tostring() the data before sending, sends2s already does this.
|
Matthew Wild |
Sat, 21 Nov 2009 13:47:45 +0000 |
|
s2smanager: Fail outgoing s2s connection if hostname does not pass idna_to_ascii(), thanks Flo + waqas
|
Matthew Wild |
Fri, 20 Nov 2009 15:07:39 +0000 |
|
s2smanager: Queue db:verify unless we already sent a db:result (if we had then it can could a dialback deadlock). Also remove some redundant code which could cause a db:result to be sent while still negotiating features (e.g. TLS) and break things. Collectively these fix a 'random' s2s failure (usually with ejabberd for some reason) - resulting in an 'unbound prefix' XML error, or 'ssl handshake failure'. Was this commit message long enough? I think so.
|
Matthew Wild |
Fri, 20 Nov 2009 04:39:54 +0000 |
|
s2smanager: Use session-specific logger to log outgoing data on outgoing s2s connections
|
Matthew Wild |
Sun, 18 Oct 2009 16:16:27 +0100 |
|
s2smanager: Add incoming_s2s to the prosody object
|
Matthew Wild |
Sat, 17 Oct 2009 15:25:30 +0100 |
|
s2smanager: Fix access of 'config' global without requiring configmanager
|
Matthew Wild |
Fri, 16 Oct 2009 22:19:20 +0100 |
|
s2smanager: Make s2s-stream-features a per-host event
|
Matthew Wild |
Thu, 08 Oct 2009 23:40:54 +0100 |
|
s2smanager: Don't send version=1.0 in reply to incoming stream tag if they didn't advertise 1.0 (makes happy some old ejabberd versions)
|
Matthew Wild |
Tue, 06 Oct 2009 15:02:16 +0100 |
|
s2smanager: Improve log message on missing to/from in stream header
|
Matthew Wild |
Tue, 06 Oct 2009 11:02:33 +0100 |
|
s2smanager: Mark session as secure if it was negotiating TLS and opens a new stream
|
Matthew Wild |
Mon, 05 Oct 2009 14:50:27 +0100 |
|
Specify XMPP version on response stream to incoming s2s connection.
|
Paul Aurich |
Sat, 03 Oct 2009 23:23:00 -0700 |
|
Minor changes; outgoing TLS works.
|
Paul Aurich |
Sat, 03 Oct 2009 19:20:20 -0700 |
|
s2smanager: Small comment
|
Matthew Wild |
Sun, 04 Oct 2009 16:40:24 +0100 |
|
s2smanager: Don't wait for stream:features from non-XMPP-1.0 hosts
|
Matthew Wild |
Sun, 04 Oct 2009 16:39:57 +0100 |
|
s2smanager: Send stream:features for incoming s2s connections which support XMPP 1.0
|
Matthew Wild |
Sun, 04 Oct 2009 16:38:33 +0100 |
|
s2smanager: Give outgoing sessions a host field (the local one)
|
Matthew Wild |
Sun, 04 Oct 2009 16:34:17 +0100 |
|
Merge with 0.5
|
Matthew Wild |
Sun, 27 Sep 2009 12:26:51 +0100 |
|
core.s2smanager: Always use last record in the DNS cache
|
Matthew Wild |
Sun, 27 Sep 2009 11:59:11 +0100 |
|
Merge with 0.5
|
Matthew Wild |
Mon, 21 Sep 2009 14:42:29 +0100 |
|
core.s2smanager: Don't recurse CNAMEs infinitely :)
|
Matthew Wild |
Mon, 21 Sep 2009 14:40:14 +0100 |
|
Merge with 0.5
|
Matthew Wild |
Sun, 20 Sep 2009 15:17:37 +0100 |
|
s2smanager: Timeout unauthed s2s connections
|
Matthew Wild |
Sun, 20 Sep 2009 15:16:25 +0100 |
|
Merge with 0.5
|
Matthew Wild |
Sat, 19 Sep 2009 17:48:15 +0100 |
|
s2smanager: Compatibility with domains which use CNAMEs
|
Matthew Wild |
Sat, 19 Sep 2009 17:47:26 +0100 |
|
Merge with 0.5
|
Matthew Wild |
Sat, 19 Sep 2009 17:42:16 +0100 |
|
s2smanager: Pass A-record lookups through adns module
|
Matthew Wild |
Sat, 19 Sep 2009 17:40:19 +0100 |
|
core.s2smanager, xmppserver_listener: Move the responsibility of setting session.log from the latter to the former for incoming sessions (outgoing are already like this)
|
Matthew Wild |
Sun, 16 Aug 2009 21:41:53 +0100 |
|
Remove version number from copyright headers
|
Matthew Wild |
Fri, 10 Jul 2009 03:11:45 +0100 |
|
s2smanager: Fix to correctly bounce stanzas if first connection attempt fails instantly
|
Matthew Wild |
Wed, 08 Jul 2009 03:14:12 +0100 |
|
s2smanager: Log the hostname and address when s2s connection fails instantly
|
Matthew Wild |
Fri, 03 Jul 2009 14:58:11 +0100 |
|
Send xml:lang in stream headers, fixes #78
|
Matthew Wild |
Wed, 01 Jul 2009 17:53:18 +0100 |
|
s2smanager: Remove srv_hosts from session when connected, this fixes attempting to reconnect s2s sessions when they are closed during shutdown
|
Matthew Wild |
Fri, 19 Jun 2009 17:27:03 +0100 |
|
s2smanager: Use a UUID as the dialback secret
|
Waqas Hussain |
Thu, 04 Jun 2009 21:12:36 +0500 |
|
s2smanager: Timeout DNS requests after 60 seconds (or dns_timeout in config)
|
Matthew Wild |
Tue, 26 May 2009 21:59:28 +0100 |
|
s2smanager: Only send reply </stream:stream> when connection is still open
|
Matthew Wild |
Tue, 26 May 2009 01:07:52 +0100 |
|
s2smanager: Miscellaneous logging improvements, changing levels, improving messages and using session loggers where possible
|
Matthew Wild |
Thu, 30 Apr 2009 02:43:12 +0100 |
|
core.s2smanager: Buffer data sent before connection
|
Matthew Wild |
Mon, 20 Apr 2009 22:14:31 +0100 |
|
core.s2smanager: Logging improvements
|
Matthew Wild |
Mon, 20 Apr 2009 22:13:09 +0100 |
|
core.s2smanager: Don't initiate dialback if we don't have a connection yet
|
Matthew Wild |
Sun, 12 Apr 2009 13:14:45 +0100 |
|
Remove more debugging code which was accidentally committed (don't trust hg)
|
Matthew Wild |
Fri, 10 Apr 2009 10:38:04 +0100 |
|
core.s2smanager: Remove some debugging
|
Matthew Wild |
Fri, 10 Apr 2009 10:36:01 +0100 |
|
Merge
|
Matthew Wild |
Fri, 10 Apr 2009 10:31:38 +0100 |
|
core.s2smanager: Look away... wait for it... now! Hopefully s2s should be back to normal.
|
Matthew Wild |
Wed, 08 Apr 2009 20:01:53 +0100 |
|
core.s2smanager: Check for valid host and port before attempting connection
|
Matthew Wild |
Mon, 06 Apr 2009 23:43:48 +0100 |
|
sessionmanager, s2smanager, mod_component: Send reply </stream:stream> when client closes stream
|
Matthew Wild |
Sun, 05 Apr 2009 20:06:39 +0100 |
|
Fixed: Some nil access bugs
|
Waqas Hussain |
Mon, 30 Mar 2009 05:26:10 +0500 |
|
Fixed: s2smanager: Apply nameprep on hostnames passed in stream tag (part of issue #57)
|
Waqas Hussain |
Mon, 30 Mar 2009 03:06:01 +0500 |
|
0.3->0.4
|
Matthew Wild |
Fri, 20 Mar 2009 20:16:25 +0000 |
|
core.s2smanager: Remove some old commented code, break a long line in 2
|
Matthew Wild |
Mon, 02 Mar 2009 13:52:08 +0000 |
|
Update copyright notices for 2009
|
Matthew Wild |
Fri, 30 Jan 2009 17:59:26 +0000 |
|
0.2->0.3
|
Matthew Wild |
Fri, 30 Jan 2009 17:40:25 +0000 |
|
GPL->MIT!
|
Matthew Wild |
Fri, 30 Jan 2009 17:22:56 +0000 |
|
Port to new server.lua, quite some changes, but I believe everything to be working
|
Matthew Wild |
Sat, 17 Jan 2009 04:45:08 +0000 |
|
Bounce stanza errors on failed s2s
|
Waqas Hussain |
Fri, 19 Dec 2008 01:57:13 +0500 |
|
Add support for dialback piggy-backing. Fixes #37. Thanks to CShadowRun for helping me test :)
|
Matthew Wild |
Sat, 13 Dec 2008 17:43:52 +0000 |
|
0.1 -> 0.2
|
Matthew Wild |
Wed, 10 Dec 2008 15:44:03 +0000 |
|
Slightly more secure dialback secret generation
|
Matthew Wild |
Sun, 07 Dec 2008 03:12:22 +0000 |
|
Remove print()s from sessionmanager and s2smanager
|
Matthew Wild |
Sat, 06 Dec 2008 23:14:39 +0000 |
|
Remove an incorrect line which I didn't add, and fix the proper way. Corrects the sending of stanzas over unauthed s2sout's. Also fixes mod_dialback to send stanzas and not strings.
|
Matthew Wild |
Fri, 05 Dec 2008 05:23:42 +0000 |
|
Fix for s2s with jabberd2 (we weren't routing db:verify's over s2sout_unauthed)
|
Matthew Wild |
Fri, 05 Dec 2008 04:44:19 +0000 |
|
Don't send stream:features to incoming s2s connections
|
Matthew Wild |
Thu, 04 Dec 2008 17:10:02 +0000 |
|
Fix logger ids for c2s and s2sout
|
Matthew Wild |
Thu, 04 Dec 2008 16:41:32 +0000 |
|
Insert copyright/license headers
|
Matthew Wild |
Wed, 03 Dec 2008 14:39:07 +0000 |
|
No more reading 1 byte at a time from sockets
|
Matthew Wild |
Fri, 28 Nov 2008 00:08:23 +0000 |
|
Switched from md5 to sha256 for dialback key generation
|
Waqas Hussain |
Fri, 28 Nov 2008 01:13:34 +0500 |
|
Log reason for connection failure
|
Matthew Wild |
Wed, 26 Nov 2008 23:47:23 +0000 |
|
(Basic) IDNA and SRV fallback support
|
Matthew Wild |
Wed, 26 Nov 2008 23:44:30 +0000 |
|
Remove or comment useless prints, or change them to log()
|
Matthew Wild |
Thu, 20 Nov 2008 03:00:43 +0000 |
|
Remove version=1.0 on s2s stream headers, again.
|
Matthew Wild |
Thu, 20 Nov 2008 01:14:31 +0000 |
|
Remove an old FIXME comment
|
Matthew Wild |
Wed, 19 Nov 2008 23:25:04 +0000 |
|
Comments and logging fixes
|
Matthew Wild |
Wed, 19 Nov 2008 22:50:37 +0000 |
|
Send version=1.0 again in s2s stream header
|
Matthew Wild |
Wed, 19 Nov 2008 22:05:44 +0000 |
|
Fix for setting the correct host on the socket, seems to fix s2s with XCP
|
Matthew Wild |
Wed, 19 Nov 2008 22:04:39 +0000 |
|
Don't send version 1.0. Who cares about Gmail?
|
Matthew Wild |
Wed, 19 Nov 2008 05:34:00 +0000 |
|
Send version=1.0 in s2s stream header. Send s2s stream:features when in 1.0 mode.
|
Matthew Wild |
Wed, 19 Nov 2008 05:13:07 +0000 |
|
Don't attempt to auth connection unless stanzas are being sent across it
|
Matthew Wild |
Wed, 19 Nov 2008 05:12:21 +0000 |
|
Log sent s2s stanzas
|
Matthew Wild |
Wed, 19 Nov 2008 05:11:37 +0000 |
|
We have SRV resolving \o/
|
Matthew Wild |
Tue, 18 Nov 2008 22:41:04 +0000 |
|
session:disconnect() -> session:close() for consistency with other Lua APIs
|
Matthew Wild |
Tue, 18 Nov 2008 19:44:54 +0000 |
|
Quite some changes, to:
|
Matthew Wild |
Tue, 18 Nov 2008 17:52:33 +0000 |
|
Fixed s2s manager's send queue to not store recieved stanza objects (as these can chnage after the send call)
|
Waqas Hussain |
Mon, 17 Nov 2008 18:53:13 +0500 |
