Previously the values where 0, 1 but now we have several more
The logic is don't send push only if preference is 'never'
Change-Id: I7a1d263da23d79d18d4dd89bc5bf1f4567e52996
* proper activation of virtual host was only done when "server-started"
event was triggered, so only on start
* ngcp-panel [de]activates the domain via console using hosts:[de]activate()
* hook to "host-[de]activated" to [de]activate the necessary subdomains
Change-Id: I3a94a600ed2b485e069b77ec9310e34a2d5194bf
* sipwise_groups:
changes in arguments of roster-load event
* sipwise_lastactivity:
integrate improvements from prosody 0.11
* sipwise_vhosts_sql:
enable muc_mam if mam is enabled
Add README.md with some manual checks I do to test the
functionality in profanity
Change-Id: Ia3df562b35dbe37e4d8da7661a7a7634d2c61768
* sql module was moved to util package
* no need to set encoding manually, is done in engine:connect()
* reuse engine if available, seems to be the way other modules
behave
Change-Id: I39afeb5a08ca5199e3f096fb43a4e8c3eac779ea
* No need to keep patches for our own modules. Patches are just
for changes We need on upstream modules. That helps us to upgrade
faster and keep our changes in a clean way
Change-Id: I0c006b758e0a891aa67cc0216635a0342282969b
- mod_carbons
- mod_csi
- mod_man
- mod_websocket
* debian/control:
we want prosody 0.11 version
remove ngcp-system-tools from dependences, not called
since removal of postinst
remove modules merged upstream from description
Change-Id: I83f5fd48ba17fe4c91763177fed7dee69ca02db1
squashed commit of
commit 4a38d9ad2b (HEAD -> master)
Author: Victor Seva <linuxmaniac@torreviejawireless.org>
Date: Mon Mar 26 14:15:41 2018 +0200
[sipwise_redis_sessions] clean previous local sessions on start
commit a514e6b60d
Author: Victor Seva <linuxmaniac@torreviejawireless.org>
Date: Mon Mar 26 14:08:31 2018 +0200
[mod_shard] clean redis info on shard/send error
if We can't connect to a shard node, clean that info from
redis depending on the stanza
MUC, persistent rooms are always defined in muc module on start, so don't
check for them locally, set as local in redis if no other node has it
closes#3
Change-Id: I33b32dacb6aa4010d520c1d04ef81736d6555370
commit d207cf102b
Author: Victor Seva <linuxmaniac@torreviejawireless.org>
Date: Wed Feb 21 21:23:58 2018 +0100
[mod_shard_client] detect loop on shard messages
Add 'via' attribute to shard stanza in order to detect loops
closes#2
Change-Id: I1e5f947a7b6fdcbd0e0d773b790929dc0681fbcd
* sipwise app logic needs the presence of the nick attribute
always to map jid and nick on MUC room
* FTR, nick is at presence from attribute
Change-Id: I710e98f1a4c5b3ea524d980dea99e3bd70deb072
* If a muc room has no description set, send its room name
instead of "prosody chatroom" to provide proper context
in push messages.
Change-Id: Id8259d69a3ebefa63da97ae0c71e39df0d558a08
* migrate from files to dirs as upstream
* refresh debian/control list of modules
* refresh mod_auth_sql.patch
* update_plugins.sh for automate the process
* keep the list of prosody modules at
prosody-modules.list file
* keep the upstream mercurial revision at
prosody-modules.revision file
Change-Id: Id4a0a2c4e88dae64641270ee79c6ee206f5bb695
> mod_c2s: Traceback[c2s]: /usr/lib/prosody/util/sql.lua:189:
> Error executing statement parameters: Data too long for column '_uuid' at row 1
Change-Id: I9e7d6a310ee5e5e8f83cbc97cb2f7ebcf4f534b5
* reduce debug output
* fire_offline_message():
- assert that we have a from nick
- check the original from jid too for is_pushd_blocked()
Change-Id: I0718fe71b9ea06e16de14aeb6017ce3986f16a71
avoid errors like:
> mod_c2s: Traceback[c2s]: /usr/lib/prosody/modules/mod_sipwise_pushd.lua:145:
> attempt to index field '?' (a nil value)
> stack traceback:
> /usr/lib/prosody/modules/mod_sipwise_pushd.lua:145: in function 'get_muc_caller'
Change-Id: Ie32195f49cef6b2145deba6ea78b8b1c43ea7d41
* dataform with domain field, if field value is '*'
result is the list of domains served by the server
Change-Id: I5ea272178a183c633cf38a2f439252d8af168681