summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2014-05-26Merge branch 'telepathy-glib-0.24'HEADmasterGuillaume Desmottes2-3/+7
2014-05-26test-connection-error: don't check errors twiceGuillaume Desmottes1-10/+2
We are already using g_assert_error() to check those. https://bugs.freedesktop.org/show_bug.cgi?id=79006
2014-05-26account-request: remove useless 'priv' variableGuillaume Desmottes1-11/+7
https://bugs.freedesktop.org/show_bug.cgi?id=79006
2014-05-26contact-list: don't crash if 'd' is NULLGuillaume Desmottes1-2/+6
https://bugs.freedesktop.org/show_bug.cgi?id=79006
2014-05-26base-client: fix potential uninitialized variable bugGuillaume Desmottes1-1/+1
2014-05-08stop hardcoding python's path in .py scriptsGuillaume Desmottes13-13/+13
https://bugs.freedesktop.org/show_bug.cgi?id=76495
2014-03-26nano versionSimon McVittie2-1/+6
2014-03-260.24.0Simon McVittie2-6/+8
2014-03-26Upload documentation to the 0.24.x locationSimon McVittie2-1/+3
2014-03-26Don't check documentation completeness in this stable branchSimon McVittie1-10/+2
2014-03-26Disable warnings about deprecated functions for stable branchSimon McVittie1-1/+2
Since this is a stable branch, we're unlikely to fix uses of deprecated functions, so we shouldn't warn about them.
2014-03-18NEWSSimon McVittie1-0/+7
2014-03-18Don't leak method call results (regression in 0.23.1)Simon McVittie2-3/+24
Commit 7ffbde3f used common code for two blocks that were not actually the same. dbus-glib method calls give the caller a new reference to the results, whereas dbus-glib signals do not. As a result, the method-call case needs to either free the results, or get them freed by giving ownership to the GValueArray. I chose to give ownership. Reviewed-by: Guillaume Desmottes
2014-03-18set nano flagGuillaume Desmottes1-1/+1
2014-03-180.23.3Guillaume Desmottes6-8/+34
2014-03-17tp_list_connection_managers_async: terminate properly if there is no CMGuillaume Desmottes3-0/+97
Fix fdo#68892.
2014-03-17connection-manager: factor out all_cms_prepared()Guillaume Desmottes1-20/+26
2014-03-14call-channel: pass our factory to TpCallStream objectsGuillaume Desmottes2-0/+5
Partially fix fdo#76168.
2014-03-14call-channel: pass our factory to TpCallContent objectsGuillaume Desmottes2-1/+4
Partially fix fdo#76168.
2014-03-13Merge branch 'telepathy-glib-0.22'Guillaume Desmottes1-4/+9
Conflicts: telepathy-glib/contact.c
2014-03-13fix file path leaksGuillaume Desmottes1-6/+14
g_file_get_path() returns (transfer full). Fix fdo#76119
2014-03-13Merge branch 'telepathy-glib-0.22'Simon McVittie3-5/+17
Conflicts: NEWS
2014-03-130.22 NEWSSimon McVittie1-1/+9
2014-03-13TpFileTransferChannel: Fix possible crashes, particularly with GLib 2.39Xavier Claessens1-21/+16
tp_file_transfer_channel_accept_file_async() and tp_file_transfer_channel_provide_file_async() operations are supposed to complete as soon as the CM returns from AcceptFile or ProvideFile. That means that we cannot call operation_failed() for streaming errors. We also have to keep a ref on self while streaming the file to avoid a crash in the callback when we dereference self. This means that the client app cannot cancel the ongoing streaming by unreffing the channel, replying on dispose calling g_cancellable_cancel(). It can still be doing using g_object_run_dispose() though. To make it cleaner we should probably add tp_file_transfer_channel_cancel(). The spec does not provide any way for the streaming client to inform the CM and other clients about the error occured while streaming. TpFileTransferChannel API does not even have a way to propagate that error to the user. [slightly more informative commit message for the 0.22 cherry-pick -smcv] Bug: https://bugs.freedesktop.org/show_bug.cgi?id=72319
2014-03-13TpProxy: finish_all_requests: don't leak copied GQueueSimon McVittie1-1/+1
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=76000 Reviewed-by: Guillaume Desmottes
2014-03-13TpContact: mime_file_written: don't leak the file's pathSimon McVittie1-4/+6
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=76000 Reviewed-by: Guillaume Desmottes
2014-02-28add tp_protocol_new_vardict()Guillaume Desmottes4-1/+58
2014-02-28protocol: add TpProtocol:protocol-properties-vardictGuillaume Desmottes4-0/+54
2014-02-28test creating a TpProtocol by passing its immutable propsGuillaume Desmottes1-0/+19
2014-02-28test-protocol-objects: factor out check_tp_protocol()Guillaume Desmottes1-28/+34
2014-02-26set nano versionGuillaume Desmottes1-1/+1
2014-02-260.23.2Guillaume Desmottes7-15/+40
2014-02-26add tp_account_channel_request_set_initial_inviteesGuillaume Desmottes4-0/+106
2014-02-26add tp_account_channel_request_set_initial_invitee_ids()Guillaume Desmottes4-1/+47
2014-02-26add tp_account_channel_request_set_conference_initial_channels()Guillaume Desmottes4-1/+56
2014-02-26account-channel-request: set TargetHandleType: TP_HANDLE_TYPE_NONE if neededGuillaume Desmottes2-5/+53
2014-02-25account-channel-request: factor out going_to_request()Guillaume Desmottes1-45/+32
2014-02-25add tp_account_channel_request_set_sms_channel()Guillaume Desmottes4-1/+35
2014-02-25add tp_account_channel_request_new_dbus_tube()Guillaume Desmottes4-0/+92
2014-02-25add tp_account_channel_request_new_stream_tube()Guillaume Desmottes4-0/+96
2014-02-25add tp_account_channel_request_set_file_transfer_hash()Guillaume Desmottes4-1/+49
2014-02-17TpBaseConnection: add "account-path-suffix" propertyXavier Claessens3-0/+57
2014-02-04nano versionSimon McVittie2-1/+4
2014-02-04maintainer-prepare-release: do "git tag" in srcdirSimon McVittie1-1/+1
2014-02-040.23.1Simon McVittie10-45/+74
2014-02-04TpProtocol: fix some suspicious memory managementSimon McVittie1-4/+4
It works fine as long as a preallocated GArray doesn't move its memory buffer for no reason, but it's good to be obviously correct. Spotted while working on the previous commit. Bug: https://bugs.freedesktop.org/show_bug.cgi?id=71093 Reviewed-by: Guillaume Desmottes <guillaume.desmottes@collabora.co.uk>
2014-02-04tp_connection_manager_param_dup_variant_type: addSimon McVittie4-0/+35
In order for this to work, TpProtocol now ignores parameters whose D-Bus signatures are not a syntactically valid single complete type. This is helpful for Mission Control to be able to migrate parameters from untyped to typed storage - it stores parameters in terms of GVariant. Bug: https://bugs.freedesktop.org/show_bug.cgi?id=71093 Reviewed-by: Guillaume Desmottes <guillaume.desmottes@collabora.co.uk> [added (transfer full) as requested in review -smcv]
2014-01-29Merge branch 'telepathy-glib-0.22'Simon McVittie1-1/+0
Conflicts: NEWS configure.ac
2014-01-29nano versionSimon McVittie2-1/+6
2014-01-29release 0.22.1Simon McVittie2-3/+5