aboutsummaryrefslogtreecommitdiff
path: root/src (follow)
Commit message (Collapse)AuthorAgeFilesLines
...
* mod_multicast changes:Alexey Shchepin2015-08-051-41/+81
| | | | | | | | | | - catch exceptions - do ets:give_away for multicastp table on init - don't send multicasts to itself - don't check user@server for multicast support - handle empty disco items - ignore cdata in <addresses/> - properly check for subdomains
* Use websocket pings also on old style connectionsPaweł Chmielowski2015-08-051-4/+3
|
* Update FSF addressChristophe Romain2015-08-055-20/+15
|
* mod_client_state: Throttle stanzas by defaultHolger Weiss2015-08-051-2/+2
| | | | | | Enable the mod_client_state options "drop_chat_states" and "queue_presence" by default. These features are now well-tested in practice.
* Rename muc_filter_packet to muc_filter_messageHolger Weiss2015-08-041-1/+1
| | | | | The hook is just triggered on groupchat messages, not on other types of stanzas.
* New hook: muc_filter_presenceHolger Weiss2015-08-041-110/+118
| | | | | Add a hook that allows for modifying or dropping presence stanzas sent to MUC rooms.
* Add split_jid/1Mickael Remond2015-08-041-1/+8
| | | | | This is a convenience reverse of make_jid/1. It allows extracting the jid parts without relying on using the jid record structure, to abstract details.
* Do not crash SIP authentication on SCRAMed passwordsEvgeniy Khramtsov2015-08-021-2/+6
|
* Don't copy messages of type "normal" without bodyHolger Weiss2015-07-281-4/+7
| | | | | | | Don't carbon-copy messages of type "normal" that don't have a body element as an immediate subtag. Those messages are usually generated by clients or servers (as opposed to messages written by humans). This includes MAM messages, for example.
* Also check for <no-storage/> hintHolger Weiss2015-07-282-18/+18
| | | | | | Check for the <no-storage/> and <no-permanent-storage/> hints in addition to <no-store/> and <no-permanent-store/>. XEP-0334 (0.1) mentions both variants, and unfortunately, both of them are in use.
* XEP-0198: Don't add <delay/> tags to IQ stanzasHolger Weiss2015-07-281-0/+2
|
* Reuse some strings in WebAdmin's Listening PortsBadlop2015-07-271-3/+3
|
* Fix problem with merging values from multiple config filesPaweł Chmielowski2015-07-241-5/+6
|
* Ensure config snippet is loaded at module installation (EJAB-1741)Christophe Romain2015-07-241-1/+3
|
* Allow contribution to include .yml or .yaml config fileChristophe Romain2015-07-241-1/+1
|
* Allow migration of old pubsub items with xmlelement in body (#479)Christophe Romain2015-07-241-1/+23
|
* Merge pull request #642 from weiss/pep-on-caps-updateChristophe Romain2015-07-242-25/+38
|\ | | | | Also send PEP notifications when local contact updates CAPS
| * Trigger PEP notifications on CAPS updatesHolger Weiss2015-07-072-29/+33
| | | | | | | | | | | | Let mod_pubsub send last items whenever a contact updates the entity capabilities. This was already done for remote contacts and is now also done for local contacts.
| * mod_pubsub: Explain caps_update usage in a commentHolger Weiss2015-07-061-0/+9
| | | | | | | | | | The reason to use the caps_update hook for sending last items to remote contacts is probably not obvious.
* | Provide command for ODBC's convert_to_scramBadlop2015-07-231-0/+4
| |
* | Don't crash web admin when displaying info about websocket using usersPaweł Chmielowski2015-07-222-1/+6
| |
* | Let nodetree call default flat plugin nowChristophe Romain2015-07-221-8/+8
| | | | | | | | this fix mistake from previous commit 311fedaa
* | Allow include of simple dependencies (EJAB-1737)(#391)Christophe Romain2015-07-221-3/+74
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Either contributed module include dependencies this way deps/ dep1/ src/ include/ dep1/ src/ include/ Or includes rebar.config or rebar.config.script: In this case, only git is supported (if git command available) and ext_mod checkout code in deps directory. In both case, only basic built procedure is supported. ext_mod does not do more than bare compilation like this: erlc -I include src/*erl
* | Let flat be default plugin (#609)Christophe Romain2015-07-2212-2119/+1941
| |
* | Do not init nodes from mod_pubsub (#609)Christophe Romain2015-07-221-9/+0
| | | | | | | | | | Let each plugin do its own initialisation in init_plugins init_nodes was an old function to be removed.
* | Merge pull request #608 from weiss/bare-jid-subscriptionChristophe Romain2015-07-221-3/+6
|\ \ | | | | | | PubSub: Check for node subscription of bare JID
| * | PubSub: Check for node subscription of bare JIDHolger Weiss2015-06-191-3/+6
| | | | | | | | | | | | | | | | | | Don't just check whether the full JID is subscribed when a node subscription is required to list or publish items. If the bare JID is subscribed, these requests are now also accepted.
* | | Merge pull request #632 from weiss/increase-max-ack-queueEvgeny Khramtsov2015-07-211-2/+2
|\ \ \ | | | | | | | | XEP-0198: Increase default value of "max_ack_queue" option
| * | | XEP-0198: Increase timeout during resumptionHolger Weiss2015-07-031-1/+1
| | | | | | | | | | | | | | | | | | | | Give gen_fsm:sync_send_all_state_event/3 a little more time to transfer the (possibly somewhat large) c2s #state during session resumption.
| * | | XEP-0198: Increase default "max_ack_queue" valueHolger Weiss2015-07-031-1/+1
| | | | | | | | | | | | | | | | | | | | | | | | During login, clients might receive a relatively large number of stanzas in one go. For some users, the default value of the "max_ack_queue" option turned out to be too small in that situation.
* | | | Merge pull request #618 from weiss/muc-message-hookEvgeny Khramtsov2015-07-211-14/+26
|\ \ \ \ | | | | | | | | | | New hook: muc_filter_packet
| * | | | New hook: muc_filter_packetHolger Weiss2015-07-031-14/+26
| |/ / / | | | | | | | | | | | | | | | | | | | | Add a hook that allows for modifying or dropping MUC room messages. Closes #491.
* | | | Print content types more gracefullyEvgeniy Khramtsov2015-07-171-3/+10
| | | |
* | | | Set direction of resource and connection (#650)Badlop2015-07-151-4/+6
| | | |
* | | | If local guide.html file not found, redirect to the online guideBadlop2015-07-151-2/+2
| | | |
* | | | Update links to Guide in WebAdminBadlop2015-07-151-11/+9
| | | |
* | | | Support RTL page direction in WebAdmin for Hebrew (#650)Badlop2015-07-151-10/+14
| | | |
* | | | Set LTR direction to Raw links (#650)Badlop2015-07-151-2/+2
| | | |
* | | | Add links in WebAdmin to ejabberd and ProcessOne (#661)Badlop2015-07-151-2/+5
| | | |
* | | | Support more mod_muc_admin translation strings (#659)Badlop2015-07-151-7/+7
| | | |
* | | | Add "complete" attribute to MAM final response (EJABS-2484)Evgeniy Khramtsov2015-07-101-36/+55
| | | |
* | | | Fix mod_mam compatibility with RSM for other backendsEvgeniy Khramtsov2015-07-101-15/+17
| | | |
* | | | Fix tests related to MAMEvgeniy Khramtsov2015-07-101-2/+3
| | | |
* | | | Fix mod_mam compatibility with RSM (only with odbc backend).Jerome Sautret2015-07-101-37/+57
| | | |
* | | | During halt only shutdown MUC rooms on local nodeEvgeniy Khramtsov2015-07-081-1/+2
| | | |
* | | | Changed mam iq get to set as specified by XEP-0313 v0.3Jerome Sautret2015-07-081-43/+56
| |_|/ |/| |
* | | Fix dialyzer warningMickael Remond2015-07-051-1/+2
|/ / | | | | | | The function is call with binary format from ejabberd_c2s.
* | Send notifications on MUC configuration changesHolger Weiss2015-07-021-0/+34
| | | | | | | | | | | | | | Notify clients when the room configuration changes, as mandated by XEP-0045, #10.2.1. Closes #623.
* | Do not rely on behaviour info when doing config validationEvgeniy Khramtsov2015-06-291-6/+3
| |
* | Omit warning message regarding non-anonymous roomHolger Weiss2015-06-281-25/+0
| | | | | | | | | | | | | | | | | | | | | | | | Previous versions of XEP-0045 suggested sending a warning message to new occupants of a non-anonymous MUC room. The current revision (1.25) says that a status code of "100" must be returned with the user's initial presence, instead. We already do this (in addition to generating the warning message). Receiving the warning message each time the client joins the room can become annoying, especially when reconnections occur frequently (e.g., on mobile devices). So, we omit it, now.