aboutsummaryrefslogtreecommitdiffstats
path: root/protocols/jabber
Commit message (Collapse)AuthorAgeLines
* Fixed a long-standing bug where non-anonymous Jabber chatrooms could causeWilmer van der Gaast2010-05-11-1/+3
| | | | | | | | | | | full JIDs to be added to the local contact list, but not removed when leaving the room, eventually triggering errors like the one in #499. People kept sending me these dumps and I finally took time to make what turned out to be this one-line fix. This error should never appear again unless people are doing really crazy stuff.
* Added a mktime_utc() to misc.c using code that used to be in jabber_util.c.Wilmer van der Gaast2010-04-08-26/+3
| | | | | I want to use this in the Twitter module.
* Suppress empty "Headline:" messages for certain new XMPP broadcast messages.Wilmer van der Gaast2010-03-29-2/+2
|
* Merging in Jabber resource selection fix. This adds the activity_timeoutWilmer van der Gaast2010-03-18-47/+58
|\ | | | | | | | | | | | | | | setting, which makes BitlBee send a message to bare JIDs if there was no recent seen activity from any of the person's resources. This should fix most issues with messages going to the wrong resource (i.e. someone's mobile phone instead of something more sensible).
| * Fixed a bug that caused full JIDs to get lost sometimes with the new wayWilmer van der Gaast2010-03-18-6/+5
| | | | | | | | | | of storing full JIDs belongong to a contact.
| * Fixed jabber_util unittest. It passes, yet something still seems to beWilmer van der Gaast2010-03-12-1/+1
| | | | | | | | | | broken. :-(
| * Merging mainline.Wilmer van der Gaast2010-03-12-21/+27
| |\
| * | Restoring some logic that seemed broken but was important with handlingWilmer van der Gaast2009-11-19-1/+6
| | | | | | | | | | | | | | | | | | of chatrooms (and likely more things). The restored version is somewhat less confusing.
| * | Fixed a facepalm kind of NULL pointer dereference bug.Wilmer van der Gaast2009-11-19-4/+5
| | |
| * | resource_select now defaults to activity instead of priority. Also, addingWilmer van der Gaast2009-11-19-48/+54
| | | | | | | | | | | | | | | | | | | | | | | | | | | a activity_timeout setting. Now, messages to someone who hasn't spoken for a while will be sent to his/her bare JID, usually resulting in a broadcast. This should fix issues with messages sometimes arriving on someone's Crackberry/Android/etc instead of some place s/he's paying attention to. Last, the activity timer is only reset on incoming messages.
* | | Set resource_select to activity by default since priority has always been1.2.5Wilmer van der Gaast2010-03-17-1/+1
| | | | | | | | | | | | | | | a stupid default. More fixes coming up soon.
* | | Oops.. Today's Jabber fix could get stuck in a somewhat infinite loop if aWilmer van der Gaast2010-03-15-17/+3
| | | | | | | | | | | | | | | | | | | | | Jabber server returns an empty <iq type="result"/> response to the session establishment request (which is valid and actually done by the example, but my test Jabberd shows different behaviour). Fixed.
* | | Don't send bind and session requests at the same time when logging inWilmer van der Gaast2010-03-14-31/+34
| | | | | | | | | | | | | | | because some very picky jabberd's don't like it. (Fixes Bug #569)
* | | Don't send a <?xml> stream start after starttls since it upsets certainWilmer van der Gaast2010-03-14-3/+6
| |/ |/| | | | | | | | | jabberd's including Zimbra's. Thanks to jMCg and balzar in #bitlbee for helping with figuring this out.
* | Changed formatting of Jabber buddy info response.Wilmer van der Gaast2010-03-07-4/+5
| |
* | The Jabber part of the change. Also made representation of not-away a bitWilmer van der Gaast2010-03-06-17/+22
|/ | | | | more consistent. Except for free-for-chat, which is nuts anyway.
* Shuffling a few more #includes. Much more shuffling will be needed.Wilmer van der Gaast2009-10-23-1/+2
|
* If you can't find what you expect in an IQ packet, it doesn't mean it'sWilmer van der Gaast2009-10-17-2/+3
| | | | | | incorrect, so stop spamming. Ideally it should return a feature-not- implemented but I'll do that later.
* Fixing a crash bug that causes crashes only with buggy Jabber conferenceWilmer van der Gaast2009-10-04-6/+9
| | | | | | servers, sending a "you left this chat" without first acknowledging you entering it.
* Fixed a bug that caused crashes when joining certain kinds of Google TalkWilmer van der Gaast2009-06-03-2/+4
| | | | | | | | | chatrooms. Doesn't seem to be Google Talk specific, other than that this is the first time I see empty <x xmlns="http://jabber.org/protocol/muc#user"/> elements that cause this crash. A more proper/efficient fix may be to just move the break outside the inner if.
* Dropping warnings about unknown JIDs since they're spammy (mostly show upWilmer van der Gaast2008-12-14-4/+6
| | | | | | when the user logs in from mulitple locations) and the irc->debug setting shouldn't be read from inside the IM modules.
* Adding the password to the right part of the <presence> stanza so joiningWilmer van der Gaast2008-09-28-1/+1
| | | | | password-protected rooms *really* works.
* Restored support for password-protected chatrooms (for now only by acceptingWilmer van der Gaast2008-09-28-3/+3
| | | | | a password in the IRC JOIN command).
* Added SET_INVALID, which set evaluators should now return instead of NULLWilmer van der Gaast2008-08-24-3/+3
| | | | | | | when the given value is not accepted. This to allow certain variables actually be set to NULL (server, for example). This should fully close #444.
* From the department of over-engineering, now cached packet IDs are fullWilmer van der Gaast2008-06-22-19/+31
| | | | | | MD5 hashes instead of a known MD5 hash with a number. Just to make it harder to confuse BitlBee by sending it faked responses to packets.
* Now preserving case in JID resources, and handling them with case sensitivityWilmer van der Gaast2008-06-22-15/+18
| | | | | | | since apparently that's how the RFC wants it. (While the rest of the JID should be case IN-sensitive. Consistency is hard to find these days...) Also extended the unittests a little bit. Closes #422.
* Added jabber_util unittests (buddy_add/_by_jid only ATM).Wilmer van der Gaast2008-06-22-1/+3
|
* Using a port list instead of a single range for the allowed Jabber portWilmer van der Gaast2008-06-15-9/+27
| | | | | numbers, adding 80 and 443. Partially closes #265.
* Added parsing of Jabber chatroom invitations.Wilmer van der Gaast2008-06-15-0/+17
|
* query.h now defines a callback function type, not using void* for it anymore.Wilmer van der Gaast2008-04-05-2/+6
| | | | | | Got rid of the bogus window handler pointer as the first argument to the callback.
* Fixed stalling issue with OpenSSL and Jabber (#368).Wilmer van der Gaast2008-03-23-2/+7
|
* Fixed cached_id_prefix memory leak.Wilmer van der Gaast2008-03-16-0/+1
|
* Moved xmltree handlers initialization to xt_new().Wilmer van der Gaast2008-02-16-2/+1
|
* Moved xmltree to lib/ because I want to use it from more than just theWilmer van der Gaast2008-02-16-687/+1
| | | | | Jabber module.
* Added callback on Jabber add-to-roster requests so buddies get added to theWilmer van der Gaast2008-02-16-1/+26
| | | | | | | internal buddy list a little bit earlier. This should deal better with Jabber servers that send presence information of a new buddy before the roster push.
* Merge from Jelmer.Wilmer van der Gaast2008-02-16-1/+1
|\
| * Fix lcov dependencies.Jelmer Vernooij2008-02-16-1/+1
| |
* | Improved sasl_get_part() to deal with whitespace in challenge strings, asWilmer van der Gaast2008-02-16-7/+20
| | | | | | | | | | described in RFC 2831 secion 7.1 (the #rule description). Closes bug #362.
* | Added support for password-protected Jabber chatrooms.Wilmer van der Gaast2008-02-10-0/+2
| |
* | Messages from the user are also included in backlogs when joining a JabberWilmer van der Gaast2008-02-03-7/+16
| | | | | | | | | | | | chatroom. Until now they were ignored, which might make backlogs a little bit confusing.
* | Implemented XEP-0115. This adds some info to the <presence/> tags soWilmer van der Gaast2008-02-03-3/+35
| | | | | | | | | | | | clients interested in capabilities can cache discovery info, so they don't have to ask about it every time you/they log in.
* | Saner garbage collection of cached packets in the Jabber module. NowWilmer van der Gaast2008-02-03-10/+11
| | | | | | | | | | | | cached packets are removed after about ten minues instead of something between one and two minutes. Closes one issue in #354.
* | Keeping track of valid Jabber connections so _connected() events will beWilmer van der Gaast2008-01-24-1/+20
| | | | | | | | | | | | ignored if the connection's dead already. Necessary if using GLib for event handling for now. :-/
* | The Jabber module now uses imcb_chat_log() instead of imcb_log() whereWilmer van der Gaast2008-01-20-5/+41
| | | | | | | | | | possible.
* | GET_BUDDY_FIRST wasn't actually implemented, even though it was in useWilmer van der Gaast2008-01-20-7/+19
| | | | | | | | | | already. I don't want to know how long it took me to find out...
* | Fixed handing of failed groupchat joins.Wilmer van der Gaast2008-01-13-6/+5
| |
* | Fixed "Conditional jump or move depends on uninitialised value(s)" atWilmer van der Gaast2008-01-13-1/+1
| | | | | | | | | | jabber_pkt_presence (presence.c:174). Valgrind-Wilmer: 1-0.
* | Fixing bug #344, now away states should always be correct, even when peopleWilmer van der Gaast2008-01-11-19/+19
|/ | | | | | do complicated things with multiple resources. (There were two bugs and some duplicated code, so I moved things around a bit.)
* More consistency in error/warning errors. Until now "WARNING:" was usuallyWilmer van der Gaast2008-01-06-9/+9
| | | | | in upper case while "Error:" wasn't .... that doesn't really make sense.
* Added missing return in jabber_login().Wilmer van der Gaast2007-12-28-0/+2
|
generated by cgit v1.2.3 (git 2.25.1) at 2025-04-29 07:33:14 +0000