aboutsummaryrefslogtreecommitdiffstats
path: root/core
Commit message (Collapse)AuthorAgeFilesLines
* s2smanager: Log error and fail gracefully if we can't open a socket for an ↵Matthew Wild2010-01-191-0/+5
| | | | outgoing s2s connection (thanks foucault)
* s2smanager: No-one noticed that Prosody stopped sending to/from on stream ↵Matthew Wild2010-01-191-1/+1
| | | | headers recently? :)
* xmlhandlers: Move stream-error detection to somewhere more appropriate (it ↵Matthew Wild2010-01-191-4/+6
| | | | was never reached) - s2s stream errors now reported properly
* xmlhandlers: More refactoring, split up stream_ns and stream_tag, add ↵Matthew Wild2010-01-191-1/+7
| | | | stream_error_tag so that callers don't need to be so worried about the separator we use
* xmlhandlers: Define ns_separator and ns_pattern to save repeating it in ↵Matthew Wild2010-01-181-3/+5
| | | | literal form throughout the file
* stanza_router: Log the name of unhandled stanzas (thanks bear)Matthew Wild2010-01-181-1/+1
|
* sessionmanager: Added resource prepping, and invalid resource checking to ↵Waqas Hussain2010-01-111-1/+3
| | | | the bind_resource function.
* componentmanager: Add dialback_secret to components to allow s2s to work ↵Matthew Wild2010-01-111-3/+9
| | | | (thanks Flo)
* configmanager: Add parsers() method to return an array of supported config ↵Matthew Wild2010-01-091-2/+11
| | | | formats
* s2smanager: Split sending of stream header into a :open_stream() methodMatthew Wild2010-01-091-2/+11
|
* s2smanager: Allow new_outgoing() to create sessions without automatically ↵Matthew Wild2010-01-041-3/+5
| | | | connecting them (set 3rd parameters to false)
* s2smanager: Stamp a 'to' on the reply stream:stream. This makes prosody happy.Paul Aurich2010-01-031-1/+1
|
* s2smanager, hostmanager: Make dialback secrets per-hostMatthew Wild2010-01-032-6/+8
|
* s2smanager: Trim trailing whitespaceMatthew Wild2010-01-031-7/+7
|
* s2smanager: Give reasons to destroy_session in more cases, including when ↵Matthew Wild2009-12-171-2/+2
| | | | DNS lookup fails
* Merge with backout branchMatthew Wild2009-12-171-1/+0
|\
| * s2smanager: More whitespaceMatthew Wild2009-12-171-3/+1
| |
* | s2smanager: Send reply stream header only after checking we serve this host ↵Matthew Wild2009-12-171-4/+4
| | | | | | | | (sorry darkrain... :) )
* | Backed out changeset 78c5cb163ea9, to fix another way. Haven't done this for ↵Matthew Wild2009-12-171-1/+2
|/ | | | a while :)
* s2smanager: Mark stream as opened sooner, this prevents a duplicated stream ↵Matthew Wild2009-12-171-2/+1
| | | | header on host-unknown errors (thanks darkrain and his mail server)
* s2smanager: WhitespaceMatthew Wild2009-12-171-1/+0
|
* s2smanager: Have both destroy_session and bounce_sendq accept a reason ↵Matthew Wild2009-12-171-3/+6
| | | | string to include in bounced stanzas
* modulemanager: Fire item-removed events on module unload.Waqas Hussain2009-12-101-0/+9
|
* modulemanager: Fixed a global access.Waqas Hussain2009-12-101-2/+2
|
* componentmanager: Set ssl_ctx[_in] for components correctly, and use global ↵Matthew Wild2009-12-041-5/+12
| | | | SSL context if available
* Disable SSLv2 by default, it's known to be insecure.Paul Aurich2009-12-041-2/+2
|
* modulemanager, net.dns: Remove trailing whitespaceMatthew Wild2009-12-021-3/+3
|
* modulemanager: Fire events module-loaded and module-unloaded, both receive a ↵Matthew Wild2009-12-011-0/+2
| | | | table with 'module' and 'host' fields, thanks bjc
* Merge with 0.6.Waqas Hussain2009-11-291-3/+3
|\
| * core.xmlhandlers: Fixed processing of empty namespaces (which caused an ↵Waqas Hussain2009-11-281-3/+3
| | | | | | | | issue with jwchat).
| * Merge with 0.6 on prosody.imvault/0.6.00.6.0Matthew Wild2009-11-252-12/+28
| |\
* | | s2smanager: Allow configuration of the dialback_secret in the configMatthew Wild2009-11-261-2/+1
| | |
* | | Merge with 0.6Matthew Wild2009-11-251-0/+4
|\| |
| * | s2smanager: Log warning when trying to send a stanza from a host we don't ↵Matthew Wild2009-11-251-0/+4
| | | | | | | | | | | | serve, instead of a traceback (thanks stpeter)
* | | core.sessionmanager, net.*_listener: Remove the evil collectgarbage() callsMatthew Wild2009-11-241-1/+0
| | |
* | | modulemanager: New module API methods for getting config options with type ↵Matthew Wild2009-11-231-1/+82
| | | | | | | | | | | | conversion, get_option_string, get_option_number, get_option_boolean, get_option_array, get_option_set
* | | Merged with 0.6.Waqas Hussain2009-11-222-12/+28
|\ \ \ | | |/ | |/|
| * | sessionmanager: Added function send_to_interested_resources().Waqas Hussain2009-11-221-0/+15
| | |
| * | sessionmanager: Fixed and cleaned function send_to_available_resources(). ↵Waqas Hussain2009-11-221-12/+7
| | | | | | | | | | | | The 'to' attribute for presence subscription stanzas is now preserved.
| * | loggingmanager: Explicitly flush log messages if the __FLUSH_LOG environment ↵Waqas Hussain2009-11-221-0/+6
| |/ | | | | | | variable is defined (workaround for MSVCRT buffering piped output).
* | s2smanager: Fix syntax error introduced in mergeMatthew Wild2009-11-221-1/+1
| |
* | Merge with trunkMatthew Wild2009-11-223-63/+63
|\ \
| * | stanza_router: Don't log full stanzas destined for s2sMatthew Wild2009-11-221-1/+1
| | |
| * | objectmanager: Convert to unix line-endingsMatthew Wild2009-11-211-60/+60
| | |
| * | s2smanager: Don't log full stanza when sending outwardsMatthew Wild2009-11-211-2/+2
| | |
| * | s2smanager: Don't tostring() the data before sending, sends2s already does this.Matthew Wild2009-11-211-1/+1
| |/
* | s2smanager: Update for new net.server API (s2s still doesn't work with ↵Matthew Wild2009-11-221-3/+3
| | | | | | | | libevent for other reasons)
* | sessionmanager: Use : syntax for calling connection methodsMatthew Wild2009-11-211-2/+2
|/
* s2smanager: Fail outgoing s2s connection if hostname does not pass ↵Matthew Wild2009-11-201-1/+6
| | | | idna_to_ascii(), thanks Flo + waqas
* s2smanager: Queue db:verify unless we already sent a db:result (if we had ↵Matthew Wild2009-11-201-5/+1
| | | | 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.