commit 7693b5d8176a6af8c1ce937d228049363c40d09f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-24 13:14:07 +0000

    jingle/outgoing-basics: disregard MembersChangedDetailed flag
    
    In 0.15, 65c8b8f makes this test always expect the
    Members_Changed_Detailed flag to be present in the channel's group
    flags. But 0.14 doesn't depend on a new enough telepathy-glib to be sure
    of that; so here we just mask that flag.

commit 266d8346b0b195a9c924cdd3ba1ba3295c9eb8ea
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-24 09:50:04 +0000

    Version 0.14.1

commit bb9ea185579d089ca87234a27e7ed971640a5da4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-24 09:49:55 +0000

    NEWS for 0.14.1

commit 4452a4c807b3fa6ad0b37342fc35d1f1049f101c
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-11-22 17:26:17 +0000

    Fix missing capabilities.h and caps-channel-manager.h
    
    The problem was introduced by this commit:
    
    |commit 587da15932d178ad289e81b039dcabb4e90984c5
    |Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
    |Date:   Thu Sep 1 14:19:52 2011 +0100
    |
    |    gabble: add capabilities.h and caps-channel-manager.h as public API
    |
    |    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 566c0f72fe54d44752c3704537e42c3e3640488c
Merge: ebcb135 7524ce7
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-15 15:34:27 +0000

    Merge branch 'telepathy-gabble-0.12' into telepathy-gabble-0.14

commit 7524ce77ef3989418d2e8b0a2af47e472c7c982d
Merge: 9cf653e 6cb2b5d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-15 15:34:11 +0000

    Merge branch 'local-pending-flicker2' into telepathy-gabble-0.12

commit 6cb2b5d0f9e4a164537e3861f787e603d02bb8e4
Author: Cosimo Alfarano <cosimo.alfarano@collabora.com>
Date:   2011-11-02 13:24:09 +0000

    add case for "cancel subscription" to test_local_pending

commit 31c2cd55912c72c82cba28202f220a80c8b4faca
Author: Cosimo Alfarano <cosimo.alfarano@collabora.com>
Date:   2011-11-02 13:20:15 +0000

    expect MembersChanged AND ContactsChanged on sub-request
    
    We should explicitly expect the relevant state transition on the publish
    channel; and when forbidding events, we just forbid them from occurring
    on the publish channel.
    
    (fix by Will Thompson, thanks Will)

commit 284779859a67e2dfceb4d9930df81e452d5a4b3a
Author: Cosimo Alfarano <cosimo.alfarano@collabora.com>
Date:   2011-11-02 11:39:46 +0000

    fix comment: Alice is just requesting subscription

commit 273654bfb07e5ff593b162dd455f739f697b71b8
Author: Cosimo Alfarano <cosimo.alfarano@collabora.com>
Date:   2011-11-02 11:39:21 +0000

    Remove useless symbol in test_local_pending

commit 8fcf944c1afbcd98a5cbf582e33d88d877a6980f
Author: Marco Barisione <marco@barisione.org>
Date:   2011-08-31 17:31:24 +0100

    Roster: don't hide local pending contacts from stored

commit ebcb1350c3b316f8fe510caacacf0b2a59b95458
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-10 16:40:09 +0000

    conn-olpc: handle activities with NULL ids.
    
    Other places in the code seem to handle GabbleOlpcActivity.id being
    NULL; and <https://bugs.freedesktop.org/show_bug.cgi?id=32050> is a
    crash on this strcmp. There are no logs, and the first argument to
    strcmp is not on record, so this suggests to me that this might be the
    cause!

commit 5677e158d6518e53bc06c131e3b2c8b9c779a74a
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-11-09 17:46:04 +0000

    Remove double definition
    
    gabbletestsdir is defined in configure.ac
    
    https://bugs.freedesktop.org/show_bug.cgi?id=42679

commit bdf5c1535079d7f69e8b0d50def2bf0acd2a7d1a
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-11-09 08:20:31 +0000

    connection: fix typo when indexing pointer array using wrong counter
    
    D'oh!
    
    Fixes: fd.o#42706
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit ed197893f4e303276480649330b9c362c60f6442
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-07 20:34:04 +0000

    Nano-version bump to 0.14.0.1

commit 2c134e18d24d1daebba6963f73c1bff9b960b46c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-07 19:59:19 +0000

    Version 0.14.0

commit 152c12a2e393e6c135a975df847f55990fc8d0f5
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-07 19:58:40 +0000

    Hey, let's call this 0.14.0

commit 0f86ad63a641a780b0e812725f7a7bd24d57a467
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-07 19:20:28 +0000

    NEWS for 0.13.8

commit df3882d34528ccb320835fafbad74034a589dc64
Merge: 6ef3ca9 6895340
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-07 17:46:04 +0000

    Merge branch 'facebook-own-message'
    
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 6895340badaff9f97683230848ae8e0f9f6f99d8
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-04 08:58:45 +0000

    ImFactory: handle Facebook's own-message extension
    
    Facebook's XMPP server sends us IQs containing messages which we send to
    contacts, whether on this connection or via another connection (such as
    the web interface). We can expose these as Delivery_Status_Accepted
    delivery reports, allowing UIs to log them and show them up or whatever.
    This could help keep things coherent if the user hops between chatting
    in the browser and/or Empathy, say.
    
    https://bugs.freedesktop.org/show_bug.cgi?id=41417

commit 6ef3ca940f0b4a4ae1aedb2c03ddc794fe659582
Author: Danielle Madeley <danielle@madeley.id.au>
Date:   2011-11-07 22:18:17 +1100

    Use commit ids that actually exist for Wocky
    
    What happened here was that git-bz was rewriting my commit ids when I
    submitted the patch to bugzilla for review, but my Gabble was still pointing
    to the original commit I had made. SURPRISE!

commit e61dc0318e7a7c00c268105599ea3eada2c1491b
Author: Danielle Madeley <danielle@madeley.id.au>
Date:   2011-11-07 21:45:02 +1100

    Return the appropriate value in the case of an unknown timestamp
    
    This updates Gabble for the latest spec.
    
    Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=42652

commit 465b3aef2330ae73c2f643eeec4676da657c2222
Merge: ed09e89 a04b9e5
Author: Danielle Madeley <danielle@madeley.id.au>
Date:   2011-11-07 20:54:44 +1100

    Merge branch 'wocky-muc-gdatetime'

commit 4312e1f6789227e658ee55349b91b0980128ab51
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-04 16:21:04 +0000

    IMChannel: reindent receiving/delivery report code
    
    This patch just gets rid of the funky formatting mentioned in an earlier
    patch; there are no code changes.

commit 574d1468432bd066dfe7a50fed8c2bbae5d71975
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-04 12:22:09 +0000

    IMChannel: split incoming message and report handling
    
    As a nice side effect, the new _gabble_im_channel_report_delivery()
    function can do successful delivery reports if send_error ==
    GABBLE_TEXT_CHANNEL_SEND_NO_ERROR. This will become useful
    momentarily...
    
    The funky formatting is to make clearer that this patch is mostly moving
    code around unmodified; a subsequent patch will clean it up.

commit 46da670d5c0fa51b92a168d454d23e9b119e349a
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-04 12:11:42 +0000

    im_channel_receive: split out building basic message
    
    This is about the only part of this function which is shared between the
    delivery report code path and the normal message code path.

commit deb5a15cb02344891f5871cd5f8dcf8faa464d4a
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-11 17:01:13 +0100

    ImFactory: refactor creating unrequested channels
    
    We'll need similar logic for handling incoming facebook:own-message
    notifications.

commit 214c0b08245fc264985e4504309abb7b7e82f68c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-03 23:50:24 +0000

    exec-with-log.sh.in: set WOCKY_DEBUG.
    
    I'm delighted that 3a1fe026 made us compatible with hypothetical shells,
    but it also reduced the amount of debug output in gabble-testing.log.

commit b34a0207287a3976e1b880c5b958b9215c5930ad
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-11-03 23:27:24 +0000

    im_channel_receive: remove redundant 'sender' parameter
    
    If we're passing an incoming message to a 1-1 IM channel, we don't need
    to tell the channel who sent the message: it knows perfectly well. In
    fact, we were telling it twice: once in handle form (which is
    redundant), and once in full JID form (which is more useful: it lets the
    channel update its resource binding).

commit 9b81a8da064954b714474208d2fb4e7d5cb2d6b4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-11 15:54:47 +0100

    ImFactory new_im_channel: remove redundant typecheck
    
    By the time we get this far we can be sure that the factory really is a
    factory.

commit 54104dac1cbfa6a13e8aba199a2043eb061e9a87
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-11 15:53:26 +0100

    ImFactory: remove GET_PRIV anti-idiom

commit 34b600adcb8894615fc2b40400ce6e740230f7c2
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-11 15:50:24 +0100

    IMChannel: implement get_object_path_suffix
    
    I think this is simpler than the factory passing the path in. Though
    it's actually longer, which offends me on some level.

commit b3d5848804b8128b0962f4860efcd758e19d7fc3
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-11 15:42:42 +0100

    ImFactory: simplify new_im_channel's arguments
    
    We can infer from @request_token whether the channel was requested
    locally or is being created in response to a incoming event; hence, we
    don't need a separate @initiator argument.

commit 7a9200ca2a426f4f794a4fd43291bcd511a9420f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-11 15:34:57 +0100

    ImFactory: document new_im_channel more usefully

commit 33cd9527f7c0d1f4a2b72e8cfadc417b69590336
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-04 01:15:32 +0100

    Remove unneeded im_channel_closed_cb prototype

commit eb0dc2a0694451453c40c0e75da3ba9093837339
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-04 00:24:40 +0100

    Document _gabble_im_channel_receive usefully

commit 933af070afdc7a3a1df333c318743e0ba115f3a5
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-04 00:20:14 +0100

    Remove redundant CHECK_STR_EMPTY macro

commit 2941264d571f3bd23739f7bbdbb350c53355398b
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-04 00:08:02 +0100

    IMFactory: use constructed, not constructor

commit aa4b8472a09fafd5bc8046c01f2163d6d7e4495c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-04 00:05:22 +0100

    Remove obsolete DBUS_API_SUBJECT_TO_CHANGE defines

commit d3607abee80d237ea1d00aca55e14eadc7ff5bb4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-30 12:16:16 +0100

    muc-factory: use a dispatch table for handling requests

commit 517cce66448836d6f3de0eef34b9c36d7a4a2bf9
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-30 11:45:40 +0100

    muc-factory: use fully-qualified property constants

commit 45b54e41a49f60b4e0c620546492a91e0c1b1b4c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-27 12:09:52 +0100

    MucFactory: remove stupid priv anti-idiom

commit ed09e896b3c249568192e92b5195462e626918fb
Merge: 08da102 f5f7cc6
Author: Andre Moreira Magalhaes (andrunko) <andre.magalhaes@collabora.co.uk>
Date:   2011-10-31 14:05:12 -0200

    Merge branch 'memory-leaks'
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit f5f7cc6b3407af00eb594aca4c2ffc797f987846
Author: Andre Moreira Magalhaes (andrunko) <andre.magalhaes@collabora.co.uk>
Date:   2011-10-31 13:55:18 -0200

    roster: Do not leak the roster stanza.

commit 1d61ece67ffac336e4beb430cb1da6e2b4f971cb
Author: Andre Moreira Magalhaes (andrunko) <andre.magalhaes@collabora.co.uk>
Date:   2011-10-31 13:54:40 -0200

    connection: Remove unnecessary extra copy of caps when updating capabilities.

commit cb97cf5832904d9ce05c93318f43d8b83fb5aa37
Author: Andre Moreira Magalhaes (andrunko) <andre.magalhaes@collabora.co.uk>
Date:   2011-10-31 13:48:15 -0200

    conn-presence: Properly free priv member.

commit 08da102662f74cf084e4d2bdfc86f2b15e33ca66
Merge: f9529b7 9cf653e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-31 09:38:35 +0000

    Merge branch 'telepathy-gabble-0.12'

commit 9cf653e93ba9beb8ff2df73627cd2b1f5bedf6bc
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-28 18:34:25 +0100

    roster: ignore multiple replies to roster query.
    
    The XMPP server running on vk.com is buggy, and replies to our roster
    query twice. Gabble just blindly assumed that any <iq type='result'>
    with a roster in it was the singular reply to our singular query. This
    is a pretty reasonable assumption but this buggy server violates it, and
    Gabble calls tp_base_contact_list_set_list_received() more than once,
    which triggers a critical in that function.
    
    A more invasive fix would use wocky_porter_send_iq_async() rather than
    scraping all the roster IQs out of everything using the LM API, but I
    wanted to quickly fix a bug on a Friday afternoon.
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=42186>
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit f9529b7d27b0cc9edc9b7c39573c533963d87079
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-10-21 15:20:03 +0100

    Update .gitignore with installed tests

commit f4b4768caabf44f0c2955e97bf8c8af592ff23b1
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-10-12 18:45:34 +0100

    Install the test plugin for the installed tests

commit 49925bb583375e19b2d0fa0aeda7902f79fca370
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-09-29 16:55:06 -0400

    tests: add --enable-installed-tests
    
    This configure option makes it possible to install the twisted tests in
    ${libdir}/telepathy-gabble-tests and run them without the sources.
    
    The dbus configuration is duplicated in tests/twisted/servicedir/ for the
    installed tests and tests/twisted/servicedir-uninstalled/ for the uninstalled
    tests.

commit 3a1fe026d8a43f9bad8c6820c73310245929fb4c
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-10-21 15:36:18 +0100

    exec-with-log.sh.in: POSIX shells, avoid "export var=foo"

commit eec2603bd4f8644cc165ce5c8ce42d21c4c73f66
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-10-21 15:11:03 +0100

    Use autoconf's $(MKDIR_P)

commit fa01616db7028029832cb6da6fa898dac0605a82
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-10-21 14:19:01 +0100

    tests/twisted/Makefile.am: Move additional tests in the main list

commit b1d83dcd017bac58e00461ed5cc2b1c967d426fe
Author: Alban Crequy <alban.crequy@collabora.co.uk>
Date:   2011-10-21 13:54:06 +0100

    tests/twisted/Makefile.am: order lists in alphabetic order

commit d92500bcd3911dbddad5fff2d40cfbc674663b41
Author: Guillaume Desmottes <guillaume.desmottes@collabora.co.uk>
Date:   2011-10-26 16:19:58 +0200

    actually implement ContactBlocking interface
    
    https://bugs.freedesktop.org/show_bug.cgi?id=42281

commit a8f305d39ee09748ac258ce577dd9738278f9561
Author: Guillaume Desmottes <guillaume.desmottes@collabora.co.uk>
Date:   2011-10-26 15:50:31 +0200

    Advertise ContactBlocking if supported
    
    https://bugs.freedesktop.org/show_bug.cgi?id=42281

commit e1973c15669a6043cd474cb48b89baafc4bb0d45
Merge: bf1c8dd 3cdbae8
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-24 11:15:41 +0100

    Merge branch '41743-show-contacts-as-offline-even-after-a-really-early-message'
    
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 3cdbae82cb09dd8218e1c783b7bad1f84b84498e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-13 11:33:08 +0100

    Fix offline contacts showing up as unknown, not offline
    
    Due to a weird interaction between the presence cache, IM channels, and
    scraping nicknames out of <message/>s, receiving a message from an
    offline contact before the roster is received would cause their status
    to erroneously show up as unknown, not offline.
    
    This fix is a bit of a hack, but it is much smaller than refactoring to
    make the IM channel store the alias (which would allow us to expunge
    keep_unavailable).
    
    This regressed as a side-effect of e0cda61.
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=41743>

commit 3b79e21684aa31b568aeb3969d034e772ac9e610
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-14 12:00:50 +0100

    Test grabbing <nick/> from <message/>
    
    If a contact is not on your roster, you typically have no idea what
    their nickname is: no roster, no PEP, no vCard (assuming the server
    doesn't let random people fetch your vCard). In this situation, contacts
    who message you out of the blue can include <nick/> in the message
    itself.
    
    This is implemented in a kind of dodgy way in Gabble at the moment: the
    IM channel forcibly retains an entry for the contact in the presence
    cache, and then the presence cache stashes the nickname as if it came
    from presence… It was also previously untested, so I thought it worth
    adding a test before I even thought about fixing how it's implemented.

commit a04b9e505e6a5baec65fbb52522f1b52656ca72e
Author: Danielle Madeley <danielle@madeley.id.au>
Date:   2011-10-18 13:47:55 +1100

    [muc-channel] WockyMuc now signals timestamps as GDateTime instead of time_t
    
    Fixes assumptions in the code where timestamp was assumed to be gint64.

commit bf1c8dd299d02ad433af2c31c879601d712e8d3b
Author: Danielle Madeley <danielle@madeley.id.au>
Date:   2011-10-17 23:04:00 +1100

    [muc-channel] timestamp is a time_t not a gint64
    
    Cast to the correct type so that Gabble compiles correctly on 32-bit.
    
    I propose fixing this properly by using either gint64 or GDateTime in Wocky,
    but not tonight.

commit 009bec00c4d83f401179fd7fb442158875238a34
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-12 17:46:33 +0100

    Update Wocky snapshot for glib/gtypes.h fix.

commit fca53f10536c4e758c24227c956296534ce2a4a3
Author: Cosimo Cecchi <cosimoc@gnome.org>
Date:   2011-10-12 12:33:33 -0400

    test-resolver: don't include glib/types.h directly
    
    Use the glib.h header instead.

commit f065fb6165f18441b00cbfc263acd50c3eb05b38
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-12 16:20:54 +0100

    nano version bump to 0.13.8.1
    
     •_________________________•

commit 9f1ddd0ae7edf1686dedb7c12d373c99b7444221
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-12 15:55:33 +0100

    Version 0.13.7

commit 51f01ab2f802ba5f40a10f5cab207d73d9914dcd
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-12 15:51:16 +0100

    Truncate generated ChangeLog at version 0.12.0
    
    It used to be 2.4M, which is almost as big as the entire contents of the
    'src' directory! Truncating it trims it down to 68K.

commit b109b58411387c2139704f13923453bbeaf15a21
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-12 15:28:08 +0100

    NEWS for 0.13.7

commit 3f1091e0f6b576f266472fbf96d271332e7686ad
Author: Xavier Claessens <xclaesse@gmail.com>
Date:   2011-10-01 20:33:17 +0200

    Do not assume we can authenticate with a password
    
    If Wocky does not understand any of the authentication mechanisms
    supported by the server, then we cannot authenticate with just a
    password; hence, we should not claim to support the X-TELEPATHY-PASSWORD
    pseudo-mechanism. (This is the case for the Windows Live Messenger XMPP
    server, for example.)

commit 9f64fd21ede78e42396c93c81892738011045404
Merge: e435637 aff2d84
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-12 12:21:28 +0100

    Merge branch 'moar-room'
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=32611>

commit aff2d8453d577020658ce47d06f07864f6194b1d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-15 18:40:40 +0100

    MUC: ref properties_being_updated

commit 12e4a9ac245617b43f3fbd3e69ad55ddb1a88fcd
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-12 18:03:08 +0100

    MUC: add more error-checking to SetSubject

commit d9160a13c236bcf8a1b789c70dee5497ca17ee4b
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-06 17:48:29 +0100

    muc/subject: fix obsolete FIXME about discoing MUCs
    
    fd.o#21152 was fixed in 2009. We do have to make the test echo back our
    unavailable presence if we want to re-use the JID, but that's fine, we
    can do that.

commit 92810af8b77c45981ff736050667161085abb099
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-06 14:35:45 +0100

    MUC: update for RoomConfig.update_async API change
    
    I decided it was clearer for TpBaseRoomConfigClass.update_async to take
    a TpBaseRoomConfig as its first argument, rather than a TpBaseChannel.
    
    This makes it reasonable to implement the update_async vfunc in
    GabbleRoomConfig! Yay, no more monkey-patching.
    
    I think it would be even better to move all the configuration-frobbing
    code to GabbleRoomConfig. But this branch is already plenty long enough.

commit 4cd04e5366889a5cf8be84d9b096023cba999598
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-05 19:51:07 +0100

    RoomConfig: move the base class to tp-glib

commit 8d298302188f33ab9336c186b3e03b458d56d629
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-05 18:29:35 +0100

    muc/room-config: test disconnecting with requests in flight

commit 9efe5f730a4ba61f8b28b0350d4b567069e7a7c7
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-05 16:19:05 +0100

    muc/room-config: expand test coverage
    
    This tests one more field of room configuration, and also checks the
    transitions to and from being a room owner.

commit 1b3e6d52e2f2969ab944460fb5c7e2b8350ec416
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-05 15:21:22 +0100

    muc/room-config.py: refactor form building/parsing

commit 983d327aa1731781d74334716adcf5baaac15f80
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-05 14:47:56 +0100

    muc/subject.py: fix disco reply
    
    <feature var='muc_public'/> should be a child of the <query/> element,
    not of the <x/> element.

commit b620985fb445fc7d54a90080232531248f7899dc
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-02 14:27:32 +0100

    MUC: emit ConfigurationRetrieved

commit dff8ddbc721ef5791bb089ed9aed19b918553a4d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-02 14:27:26 +0100

    RoomConfig: implement ConfigurationRetrieved

commit 5ef367af72b53e4c7d115be037139410f2c03a3e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-02 11:30:49 +0100

    subject: test for PropertiesChanged
    
    Removing the Telepathy.Properties interface from GabbleMucChannel
    miraculously makes this work. \o/

commit 675367323b8617b88f9f156545e0ea63b21d88e6
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-02 11:26:02 +0100

    Rename muc/test-muc-properties to muc/room-config
    
    This more accurately reflects its current nature!

commit e3b9601367429f2272949995c71b8fd425c28e93
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-02 11:16:21 +0100

    MUC: emit PropertiesChanged for RoomConfig

commit ede43dda5a6fbed9c7a90b735b24f59fc50a1589
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-02 11:15:19 +0100

    RoomConfig: track changed properties
    
    With this patch, GabbleRoomConfig tracks which properties have changed,
    and grows a method to emit PropertiesChanged for those properties.
    
    I am not very happy with the macroification of the set_property
    implementation, but it's the best I could come up with.

commit 4c05dc0910cc1d58f71047fa53162d750fc8c6fb
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-01 18:48:04 +0100

    RoomConfig: refactor validating property types

commit 413157a5b3a66343a7b811a3fd72b516ef86bf2e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-01 18:17:02 +0100

    MUC: remove Telepathy.Properties
    
    \o\ /o/ \o\ /o/
    
    Now I can get stuck into implementing change notification.

commit 16a480c39631481c17fd788a197f7445a7b1effe
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-06 17:01:22 +0100

    muc/test-muc: neaten up interface assertions

commit 59ee55354fee1a9425790886ea98f342340e390b
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-01 14:35:59 +0100

    MUC: hook up UpdateConfiguration
    
    This simultaneously un-hooks-up SetProperties. I could have refactored
    to make it possible to keep both, but I don't think it would buy us
    anything since we're just about to delete old properties anyway.

commit 3a7eadb2ce77968034832ce7d0f039d15fdb3256
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-01 13:39:20 +0100

    RoomConfig: add a vfunc for updating configuration

commit e70dccee1faf596ff8d00613d7651d3ae751c342
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-30 12:57:09 +0100

    MUC: mark appropriate properties as mutable
    
    This is a *lot* simpler than in the old days. We no longer have to
    do something per-property on every permission change, because whether or
    not we are currently an owner of the MUC has no bearing on which
    properties would be mutable if we were.

commit 26dc9c3df3340bb2662a870301f233274d5fdc47
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-30 12:32:18 +0100

    RoomConfig: track ostensibly-mutable properties
    
    This isn't hooked up in MucChannel yet, so we currently report all
    properties as immutable.
    
    It occurs to me that “immutable” means something else elsewhere in
    Telepathy so we should probably call this something else.

commit 4d5b16af21c1dfde2426456409182603b6063ee0
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-30 11:42:31 +0100

    RoomConfig: implement CanUpdateConfiguration

commit 757718b7993f17d95184795634519968553ed8eb
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-29 16:39:46 +0100

    RoomConfig: hook up reading most fields
    
    This doesn't cover emitting PropertiesChanged (impossible until we
    expunge old-style properties), nor the current room password, nor
    changing any fields.

commit 8d686b421db6a1b7b38ab30b0d3739f768ee3232
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-29 14:09:47 +0100

    Add the start of a RoomConfig object.
    
    I'm implementing this as a separate object rather than as part of
    GabbleMucChannel in the hope of sticking it into tp-glib.

commit 54dd251e25734eb097e2a4ba278f1d8ff1dde2e6
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-30 12:55:08 +0100

    test-muc-properties: make clearer assertions
    
    The formatting of these assertions was unhelpful, and using assertEquals
    gives us nicer messages when the assertions do fail.

commit 82c09a7a031ff2726d56a6d28c01d7c911ef3c23
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-30 12:34:08 +0100

    MUC: remove half-baked allowinvites support
    
    Previously, we allowed the user to configure whether or not occupants
    should be allowed to invite others to the room, but didn't actually
    provide any way to retrieve the current setting.
    
    This is a pretty useless MUC configuration flag anyway, given that
    occupants can always send direct (as opposed to mediated)
    invitations … so I have no qualms about expunging it.

commit 59e35681f2762383e05e696d176a21c6c3b81344
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-30 10:32:44 +0100

    MUC: wockify submitting configuration form

commit 1edd10efc2c7f16025a2de56b34740b37f849bf4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-30 09:56:29 +0100

    MUC: fill in config form using a lookup table
    
    Again, this is shorter than the huge chain of ifs, and I think it's more
    readable.

commit a9f7e51872bdf63a421cb4c26b6e61a59c4c992f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-29 17:46:47 +0100

    MUC: correctly send default configuration to the MUC
    
    Ahem. Previously we were sending it to our own server. Not so useful.

commit bd60fe08cb47fd30d5471c08eed8464bf19646a2
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-29 17:42:14 +0100

    MUC: remove faux error handling when accepting default config
    
    _gabble_connection_send_with_reply always returns TRUE. I guess the only
    case this doesn't handle is when the channel (and hence the WockyMuc)
    has somehow miraculously outlived the connection—I don't think this
    happens.

commit 9e2d76bc981fae2b1c309d11c52a6c369c9d0c57
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-29 17:20:16 +0100

    MUC: wockify one use of muc#owner.
    
    We request this and parse the reply in two-and-a-half places … this is
    the simpler one of the two places where we actually parse the reply.

commit 0d63603fe7835b18cc12b9e3d101dd30e700d8fa
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-29 15:39:06 +0100

    MUC: refactor handling forms in disco reply
    
    This only works because we only handle one field. As and when we handle
    more, we should switch to WockyDataForm anyway.

commit 772c8df4a54228b08f1d58ccd8de7d688a374a47
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-29 15:18:15 +0100

    MUC: simplify feature mapping with a lookup table
    
    This is shorter than the previous chain of if-else, and (I believe) much
    clearer.

commit d7e694b31c191bbe9d2dc56922cb171e8ed2dbd8
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 17:27:06 +0100

    MUC: remove redundant role and affiliation enums
    
    Wocky has these. No need to duplicate them.

commit f4b4a4b00011c8f40fc7e245d842e09526095d3c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 17:08:56 +0100

    MUC: Remove subject from old Properties

commit b06ab5b631006a512d1d811cb306aaa2f3702edf
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 16:49:09 +0100

    MUC: Wockify parsing Subject-setting errors

commit f53172e06c29586a32989b11cf66aa2d28368ad0
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-06 16:42:08 +0100

    MUC: use id='' to catch SetSubject errors
    
    Servers don't strictly have to echo the <subject/> element in their
    reply if they reject our subject change message. But they do have to
    preserve the id='' attribute.
    
    This patch makes Gabble set an id='' attribute on <message/>s sent to
    change the subject, and look for that ID when it receives a <message
    type='error'/> from a MUC. (It continues to look for <subject/> errors,
    too.)
    
    Prosody is an example of a server that behaves like this.

commit 2f4d98398c24b95b1178871f62cf89a7bf467dd8
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 16:38:19 +0100

    MUC: return asynchronously from SetSubject

commit 23451bbeb0d06f14d5d2726d92b79a1f618baffe
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 16:35:05 +0100

    MUC: handle errors when setting subject
    
    So this hasn't worked since the WockyMUC port. If you tried to set the
    subject, and got an error back, the SetProperties() call would never
    terminate (and you wouldn't be able to call it again for the lifetime of
    the channel).

commit 1e94a3c37c1e134dfb0f88bff1cfe4ad5b0628c7
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 16:34:46 +0100

    MUC: remove unused argument from _handle_subject

commit 2f3df38c90f06698a16fa95d6d4c676c07b28efe
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 16:04:25 +0100

    muc-channel.h: clean up stale prototypes
    
    Most of these no longer exist. Those which do are only used in
    muc-channel.c, and one of those three is a one-liner which I just
    inlined at the single point it's called.

commit a1463a4f488b62a8891dd3d1e98ffb91f1014f1c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-03 14:17:03 +0100

    Wockify SetSubject implementation

commit f40f170f058fd00a46f74fafca761cbbe6affa14
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-02 18:35:52 +0100

    Subject: use tp_svc_dbus_properties_emit_properties_changed

commit 5f3afbcb08a3e686ff536dc024aff91a50782549
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-02 18:22:06 +0100

    MUCChannel: simplify password flags implementation
    
    We only have one flag, so we may as well just store a boolean.

commit c7d28425eefc77826d3235647f0370c8813cd180
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-02 12:52:55 +0100

    Build against Room & Subject from tp-glib
    
    telepathy-glib 0.15.8 was technically the first release that had Room
    and Subject, but 0.15.9 was released 10 minutes later. :)

commit 6f2c9640bb568e6672989cee9ea078661afd541e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-31 23:03:48 +0100

    Update to Jonny's last Subject draft
    
    I had to comment out the change notification tests because this object
    implements both Telepathy.Properties and DBus.Properties; both of them
    have a signal called PropertiesChanged, and dbus-glib does not allow
    the two to co-exist.

commit 082d84aeef688809fe255098bac75cbc40ef30f9
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-04 09:01:45 +0000

    muc-{channel,factory}: update to latest draft spec
    
    Split the subject stuff from Room into a different interface and
    rename RoomID to RoomName.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 81bbbae26a884771b0cd7be5da742dc5b7a2b4e7
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-04 08:49:52 +0000

    extensions: update to latest Room draft and add Chan.I.Subject
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 5fe7678d2bd0c5af56adc753d7eb97b47b840b1a
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 17:09:43 +0000

    muc-channel: XMPP always supports room subjects
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 160e92fd1d4987019f63b27d1392e97fb4ec270c
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 16:51:33 +0000

    muc/subject.py: make more assertions when setting the subject
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 1b934c6f47bdbf5b53f670f9f55794afc65630a9
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 16:50:31 +0000

    muc/subject.py: replace magic numbers with text
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 636e252a569e231d3ec1db4eeb2ba49a88938428
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 16:48:45 +0000

    muc-channel: add a convenience emit_subject_changed method
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 0d54d8155d2f02d106dab6f45aa58a05f22fa3e2
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 12:45:22 +0000

    muc/subject.py: test calling SetSubject
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 9a3b1476583ab2726f40a917924cbdb7164827f2
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 12:34:16 +0000

    muc-channel: implement SetSubject
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit b4f1b66d4916144c6155653678592318037b0095
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 12:33:41 +0000

    muc/subject.py: test for subject flags and SubjectChanged signals
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 72b0a0a4dce061c12cbf406a54059c31beb1a6f7
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 12:33:19 +0000

    muc-channel: emit SubjectChanged where appropriate
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit e785bb839a8d9e0af6314497a28641c1e2a1c615
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 11:40:31 +0000

    extensions: update to unmerged Room draft
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit e435637b78212c6604489bf031d8bf0c56d549d5
Author: Mikhail Zabaluev <mikhail.zabaluev@nokia.com>
Date:   2011-10-11 20:55:05 +0300

    Disable STUN server tests that exercise Google relay, if that is disabled

commit 3e3bf4b79322ec5352c21f5e12019decb56ac44f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-10-03 14:47:20 +0100

    ServerTLSManager: remove useless DEBUG output
    
    Reviewed-by: Sjoerd Simons <sjoerd.simons@collabora.co.uk>

commit a64cd72472d48ad8e88bfebe5e6046f02d4b5c93
Merge: 5f1ad2c 9c86f44
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-28 18:58:10 +0100

    Merge branch 'moar-caps'

commit 5f1ad2c76f43842856fb01a2dc51648a42f24284
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-28 18:45:23 +0100

    Nano-version bump to 0.13.6.1

commit 9c86f446c6a00142c373aae1fa357f5c00f0f2c6
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-28 18:45:08 +0100

    connection: document why data forms might cause warnings to appear
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 30315eacd21cd322d0b7adaf743fa57b373f6417
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-28 18:26:55 +0100

    Version 0.13.6

commit ca01a7df1a6d7efa06021dba1594ca27380a9129
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-28 18:20:46 +0100

    More NEWS for 0.13.6

commit 7fb0ccab3ce97fabcb2009faaa50f5c3e7d04019
Merge: 6211097 c3b6d21
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-28 18:10:41 +0100

    Merge branch 'telepathy-gabble-0.12'
    
    Conflicts:
    	NEWS
    	configure.ac

commit c3b6d2101aa5257a6f1ec927f3b0f3b3911cd2a6
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-28 17:53:39 +0100

    nano-version bump! 0.12.7.1

commit 8acd2b0e95fa4b732e1b82eacbc85f0fddf05fee
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-28 17:29:32 +0100

    Version 0.12.7

commit c48f3da4949aac74b2c455983ca982166ff031db
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-28 17:29:23 +0100

    NEWS for 0.12.7

commit 498920f1f69cdccbcc3c5f2146a1b1b9b5cd5976
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-28 14:17:15 +0100

    connection: drop invalid or duplicate data forms from channel managers
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit f670d0082e850835b93405eac32da24d94496975
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-28 15:07:02 +0100

    Revert "connection: deal with not being able to hash self presence"
    
    This reverts commit 9f8f181fd943f1e96f898c736ee4977d310b94a1.

commit 6d7811ac1c1e0b7b98431d567626bff1f4161d3b
Author: Jussi Kukkonen <jku@linux.intel.com>
Date:   2011-09-01 23:48:35 +0300

    mail notification: ensure google notification setting is correct
    
    Google XMPP server has an (extension) setting that can be used to
    enable mail notifications, see
    https://code.google.com/apis/talk/jep_extensions/usersettings.html
    
    https://bugs.freedesktop.org/show_bug.cgi?id=40565

commit 5e27d123f8a24509a619010cf2dedbbd632bbf0b
Author: Jussi Kukkonen <jku@linux.intel.com>
Date:   2011-09-01 23:45:54 +0300

    mail notification: recognize "google:setting" namespace

commit 55b092e583fd16ac18440c989bfc603176dbd610
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-27 14:59:21 +0100

    connection: log what form types clients contribute
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit b9c8a79c4f562a6971eb051848685b69d6b986d9
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-27 14:42:05 +0100

    presence-cache: refactor data form replacing into its own function
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 1edf9a4dbc1bbbd24a9141b676eb310d6e4be919
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-27 14:39:20 +0100

    dataforms.py: be sure to call UpdateCapabilities with arguments
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit c005728f1cab3ae10aeae2cc51c4691f9aa03593
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-27 14:38:23 +0100

    presence-cache: use hash table iters instead of foreach
    
    This is nicer and can terminate early.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 084793e6ec80af7d7c6b891a6bdb896ab9f03d1a
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-27 14:34:32 +0100

    presence-cache: clean up code to replace data form arrays
    
    This was broken in the case of data_forms == NULL as it left
    info->data_forms pointing to a freed GPtrArray. Good catch, Will!
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 3d42307058b3a1bc31138b53b21e7ed8bf681e1b
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-27 11:58:50 +0100

    connection: fix comment placement in an if/else block
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 6211097e8b2df75b79313746b6f7c4e081921eb5
Merge: 060acfb c4bfa12
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-26 16:38:48 +0100

    Merge branch 'soup-is-optional'
    
    Conflicts:
    	tests/twisted/jingle/google-relay.py

commit c4bfa12099858594001c15a2b1750977aaef814d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-26 15:26:21 +0100

    tests: when built without soup, don't test google relay

commit 060acfb55bdafcafff595eff6e2ed3ae24877c04
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-21 13:47:21 +0100

    Rename test-initial-aliases to match Makefile

commit c84617178fd951ece907cec4eb5c1da44c2123bb
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-20 15:58:37 +0100

    Call: handle google relay reply after channel dies
    
    Previously, if the Google relay server replied to our HTTP request after
    the Call channel had already gone away, we'd crash.
    
    Fixes: <http://bugs.freedesktop.org/show_bug.cgi?id=39768>

commit e7cc7f9b9b41923ab57b284776911f82102d06b3
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-20 15:58:37 +0100

    Call: test google relay reply after channel dies
    
    It's kind of funny: the bug fixed by the previous commit repeatedly
    surfaced for StreamedMedia channels in a variety of situations (you
    closed the channel, you removed yourself from Members, you disconnected
    the connection), and hence there's a lovingly-written exhaustive set of
    test cases for those scenarios for incoming and outgoing channels. But
    of course they weren't run for Call channels…
    
    This patch only makes the bare minimum number of changes to the test to
    at least slightly exercise the Call code. It's not as exhaustive as it
    is for StreamedMedia, and works almost by accident, but hey...

commit 06d98c31504139136336160aac4b1f094a8ccc88
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-20 15:58:37 +0100

    Call: handle google relay reply after channel dies
    
    Previously, if the Google relay server replied to our HTTP request after
    the Call channel had already gone away, we'd crash.
    
    Fixes: <http://bugs.freedesktop.org/show_bug.cgi?id=39768>

commit f1ece52985d24d734462f96e8a33db8ced771fdb
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-20 15:50:28 +0100

    jingle/google-relay: use functools.partial

commit 114a25a2cb9a1f57560cca30bddfeaeaeec0e3b6
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-20 15:21:50 +0100

    Roster: rename updates_nicknames to updated_nicknames
    
    This is a typo!

commit b070c914d5f59f2f5b7a5cd1ad476b9c4729eaaf
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-20 15:19:17 +0100

    Roster: don't emit nicknames-update with no nicknames.
    
    This was triggering an assertion failure in one of the signal handlers,
    which asserts that the array is non-empty.
    
    Reviewed-by: Marco Barisione <marco@barisione.org>

commit 2413a026e53c42fa0973b0a861587dff12c1e65e
Author: Derek Foreman <derek.foreman@collabora.co.uk>
Date:   2011-05-16 12:19:44 -0400

    Remove duplicates from source file lists
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit acaf2d4c7f54ecad880450987f716aface11d741
Author: Derek Foreman <derek.foreman@collabora.co.uk>
Date:   2011-05-16 12:17:50 -0400

    Check for presence of gmodule
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 3967b4ccf908c70be173c1b27617ebdc19fd6c48
Merge: bc40410 40b5b45
Author: Marco Barisione <marco@barisione.org>
Date:   2011-09-20 12:44:00 +0100

    Merge branch 'grouped-alias-updates'
    
    Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=40943
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit 40b5b45203098ce308a51b6c88c4ab31d40e047d
Author: Marco Barisione <marco@barisione.org>
Date:   2011-09-16 18:57:56 +0100

    roster: emit a single signal for all the aliases updated at the same time

commit d4cfda9cd64923c81d87b8f875b5f20658023a17
Author: Marco Barisione <marco@barisione.org>
Date:   2011-09-16 16:39:02 +0100

    conn-aliasing: add a plural version of the nickname-update signal handler

commit bc40410024314a700ef70702502f868630e02e8b
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-15 16:05:03 +0100

    with-session-bus: import --also-for-system from Salut
    
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit e829ba1c802f8b06bf678772046833d0a1c6fd0b
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-12 18:03:27 +0100

    with-session-bus.sh: retain newline when forking dbus-monitor

commit 30a7ac413f66178c3cceae63baab487365ef8845
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-07 17:02:18 +0100

    connection: add more public helper functions for getting caps
    
    This is useful so in a Gabble plugin you can get the TpBaseContactList
    and then look at each contact's caps for whatever reason.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 485c9d8ad1365cbee11b8c7769425d5d5c02fc27
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-07 15:25:14 +0100

    connection: fix typo when checking whether a client can do anything
    
    Naughty Simon.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 9ca541ae76b6dc1838f11eaea21d6e284d5d7965
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-07 15:18:40 +0100

    connection: stop trying to print data form titles as debug messages
    
    Approximately no data forms in entity capabilities have titles set
    properly anyway, so outputting:
    
      (null)
      (null)
    
    was getting dull.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit c0e658ed4e99555466e5d1c09140f43cb4faf898
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-06 14:21:51 +0100

    connection: add pick_best_resource_for_caps function
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit d96a1a927959b43586a888c56d45ab8b23b60de0
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-06 14:21:31 +0100

    connection: add get_jid_for_caps utility function
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 70580a776d7b54a950042ea6c08b4949d96f2332
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-06 14:17:44 +0100

    connection: add get_session public function
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 003e6186d70640c32b85cdf5bc920cbd94882e43
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-06 14:17:01 +0100

    connection: make get_full_jid part of the public API
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 7436f9a3144929ffb10e6c9b1c240d153529a5a3
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-06 14:15:57 +0100

    presence: implement WockyXep0115Capabilities interface
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit e34fde96b5e3c5a766f66380e1e09a3f9a9074d2
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-06 14:15:30 +0100

    presence-cache: add get_handle utility function
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit d63ca76093a79a630a69fc2d49384004f3f5fe33
Author: Jussi Kukkonen <jku@linux.intel.com>
Date:   2011-09-01 23:48:35 +0300

    mail notification: ensure google notification setting is correct
    
    Google XMPP server has an (extension) setting that can be used to
    enable mail notifications, see
    https://code.google.com/apis/talk/jep_extensions/usersettings.html
    
    https://bugs.freedesktop.org/show_bug.cgi?id=40565

commit dc7ba1b82823a8e637f1bf3e8257ff44132f44b1
Author: Jussi Kukkonen <jku@linux.intel.com>
Date:   2011-09-01 23:45:54 +0300

    mail notification: recognize "google:setting" namespace

commit 9b3b88ab8a7fdeeeb98faae23e6b53b1051e3415
Merge: 9b3e977 97175d8
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-05 18:53:26 +0100

    Merge branch 'telepathy-gabble-0.12'

commit 97175d86d43f0d9ebe7fa3f3f0ed96e825ae8094
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-03 08:32:16 +0100

    MUC: add password test
    
    I left this out of 2d7c9db :(

commit 9b3e977b7b08c6cfa055183f1914ab1b1b891c4b
Merge: e3d4b85 6c102c0
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-05 18:49:55 +0100

    Merge branch 'CAPS-LOCK'
    
    https://bugs.freedesktop.org/show_bug.cgi?id=40390

commit 6c102c09557e2988386751ffb1ee7186f0747578
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 11:00:49 +0100

    Connection: assume all handles have ContactCaps
    
    Previously, some places in Gabble which called
    gabble_connection_get_handle_contact_capabilities() did not assume that
    it never returns NULL. But (due to the implicit text caps for everyone)
    it is guaranteed always to return a non-NULL array of capabilities, so
    we can simplify the places that call it.

commit 00de9b2849013d7bc011ef34d3a3cb5a2730a83c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 10:56:49 +0100

    CapsChannelManager: correct English in a comment

commit 29d365c626445bd3f0c30d9cd6a1103cc87c3d96
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-26 10:55:20 +0100

    Connection: simplify getting ContactCaps without presence
    
    If we don't have presence for a contact, we do not have any capabilities
    for them. But we always want to ensure that all contacts are capable of
    IM. So, currently there is a special-case in the caps-gathering code: if
    we have no presence, we only call _get_contact_capabilities on the IM
    factory: its implementation of this virtual method, unlike all others,
    doesn't care if it's passed a NULL set of caps. Otherwise, if we do have
    a set of capabilities, we call gabble_connection_build_contact_caps(),
    which is essentially a foreach over all channel managers, including the
    IM factory.
    
    I think it's neater to avoid the special case, and treat "no presence"
    as equivalent (for capabilities purposes) to "presence with no caps".

commit e3d4b85839b465e160b91ac7b581dabd1097fc6e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-03 11:23:33 +0100

    Remove support for legacy Presence.
    
    This was Olli's idea. He claims that building the old-school signals
    causes measurable CPU usage on accounts with lots of contacts (think:
    Facebook). It certainly causes D-Bus spam.
    
    Neither Telepathy-GLib nor TelepathyQt4 use this old interface. Empathy
    doesn't, either. I think this should be safe.
    
    Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=40598
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit b16296db19ecbca6cfaf8d669c8997dbedb8e5fc
Merge: 8569bab 2d7c9db
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-09-05 18:39:02 +0100

    Merge branch 'telepathy-gabble-0.12'

commit 2d7c9db4766394d39ff3b3fc42cf7e928e4f4751
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-03 08:32:16 +0100

    MUC: don't forget password when handling nick conflicts
    
    WockyMuc has a property, :password, representing the current password
    being used to join the MUC. GabbleMuc previously had a private variable,
    'password', which was used for this. In the port to WockyMuc, setting
    the private variable was removed, but it was still used when
    re-attempting to join after a nick conflict.
    
    (I think the password should be a parameter to a hypothetical
    wocky_muc_join_async() which does all the nick conflict crap for you.
    Having this as state that kicks around on the WockyMuc for ever is
    bizarre—once you're in the room, you usually don't use the password,
    unless you're the owner, in which case you can retrieve the current
    password *which may be different*!)
    
    This patch expunges the zombie private variable, and ensures
    WockyMuc:password is only set when the user provides a password, not at
    every join attempt. It adds a test for this case, and some of the basic
    functionality of Password (which subsumes some incidental testing of the
    Password interface in muc/presence-before-closing: the only test that
    touched Password at all!).
    
    Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=39790
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 25c3f4dd8b9360cb531f0fc3a653094ce1e9b042
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-01 14:30:15 +0100

    test plugin: make TestChannelManager implement CapsChannelManager
    
    ...and give back a new data form in the represent_client function.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 9f8f181fd943f1e96f898c736ee4977d310b94a1
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-02 16:29:04 +0100

    connection: deal with not being able to hash self presence
    
    If a data form is bad we might not be able to, so we shouldn't crash.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 00a7c2fb7195374c39f2b9c9be8254822e5484e8
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-02 16:38:43 +0100

    caps_helper: give data forms back to disco utility functions
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit adea0ce7c2e48190fcf18c11717380e551432e04
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-02 16:35:48 +0100

    jingle-share tests: also consider data forms in disco replies
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 6dfe751e17618a5f29f6611b649e7a1bbe0969d3
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-02 16:31:41 +0100

    caps_helper: split out code to read disco replies
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 3968fd5da0e56006fde7bf6832e605ef8f544572
Author: Marco Barisione <marco@barisione.org>
Date:   2011-09-02 11:24:30 +0100

    Stable Gabble was accidentally using the master branch of wocky

commit 587da15932d178ad289e81b039dcabb4e90984c5
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-01 14:19:52 +0100

    gabble: add capabilities.h and caps-channel-manager.h as public API
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 1afd2d8e2fab78542467b2f36f08fc6e84523f91
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-09-01 12:13:16 +0100

    caps_helper.py: add initial data form parser
    
    We want to be able to check that disco replies with data forms still
    have the correct hash.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 45dc8f926f18be249f24f66deb589907bda47f98
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-08-31 16:29:43 +0100

    connection: send data forms in disco query replies
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 7c6971a75a6eb32c77b64a7e9754e1fdb2a694af
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-08-31 11:40:45 +0100

    presence-cache: store data forms from disco replies and the caps cache
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit d887fdc2d23bf00322c9923e8eead7c5f0607302
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-08-31 08:58:55 +0100

    presence: start holding a list of data forms
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 28b64163e8bbca02de308612f6653ed291e2b1df
Author: Mikhail Zabaluev <mikhail.zabaluev@nokia.com>
Date:   2011-09-01 13:22:42 +0300

    Don't allow the stub for GabbleGoogleRelayResolver to be empty
    
    g_slice_alloc() behavior for zero-sized blocks is undocumented and
    leads to obscure code paths taken.

commit 39fc20df8c978ebf072998cc1077a48ec6b41593
Author: Mikhail Zabaluev <mikhail.zabaluev@nokia.com>
Date:   2011-09-01 11:32:08 +0300

    Ignore the relay info in Jingle stanza if Google relay support is disabled

commit d2b3153f84e1f010b02ce4b084f3711ab95060e8
Author: Mikhail Zabaluev <mikhail.zabaluev@nokia.com>
Date:   2011-08-31 16:14:09 +0300

    Make Google proxy support optional
    
    The intent is to make optional the dependency on libsoup (fd.o #40537).

commit 60b5bc0cf55ee60b0c66384e544a9f35d3b3688d
Author: Mikhail Zabaluev <mikhail.zabaluev@nokia.com>
Date:   2011-08-31 13:56:10 +0300

    Isolated the Google relay resolution code in a separate source file

commit 8569bab0c49f88e397aebc9fcf2d045a21b5be75
Author: Olivier Crête <olivier.crete@collabora.com>
Date:   2011-08-29 17:46:58 -0400

    Announce the camera-v1 caps bundle so that the Google UIs put a little camera for us
    
    https://bugs.freedesktop.org/show_bug.cgi?id=40471

commit 5dcf95eaf432b5af1edf82e87f40f795a3265a35
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-08-30 11:42:50 +0100

    connection: save data forms created from UpdateCapabilities
    
    These aren't used yet.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit d9ecbbc1bbe573049e8cbb3348c600ec7512c84e
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-08-30 11:18:45 +0100

    caps-channel-manager: add data form argument to represent_client
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 81176c13d19530ac37401e7b9e597ebfbd2a2904
Author: Olivier Crête <olivier.crete@collabora.co.uk>
Date:   2011-07-19 21:01:10 +0300

    presence-cache.c: Add Android presence cache bundle
    
    Let's add another "well-known" bundle
    
    https://bugs.freedesktop.org/show_bug.cgi?id=36996

commit 78015e5dc12c37d6a99b417d8e1abea9e975e607
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2010-02-05 13:41:29 +0000

    Add a summary of options to the end of configure

commit 068445479b27c036e01e6e2f7134f7cfd2f48d6f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 16:25:07 +0100

    nano version bump

commit 6c86d9bc863af7504bfafeb1212360cfd0f2b70d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 16:22:35 +0100

    bump nano-version to 0.13.5.1

commit 6f7803086bdce403b2b32e83d16a083bce9d0ee9
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 16:18:11 +0100

    0.13.5

commit 881c1802539f5ee4a8d095d0c97abc6088b301bc
Merge: 7d56957 45254c4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 16:17:42 +0100

    Merge branch 'telepathy-gabble-0.12'
    
    Conflicts:
    	NEWS
    	configure.ac

commit 45254c4fd2bd13f6ddb7cdf0e26f2d35cb4ba8ff
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 16:13:07 +0100

    Version 0.12.6

commit 7d569578494b20abd7889bfd7d3fd4defef07568
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 15:54:44 +0100

    nano version! 0.13.4.1

commit a94dbcdd0d3eed3cecc84e2c825b2002ec61d60f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 15:30:41 +0100

    version 0.13.4

commit ddea8e83665ae5e5fa9f256091992f3065c77797
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 15:30:16 +0100

    NEWS for 0.13.4

commit e4590e849801ff4b0f2ddb9ad23a7da210212395
Merge: a89eb63 ac4485a
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 13:12:57 +0100

    Merge branch 'telepathy-gabble-0.12'
    
    Conflicts:
    	NEWS
    	configure.ac

commit ac4485aa2b217b622fca2529f456b39ec4a0e0b7
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 13:09:20 +0100

    nano-version bump!

commit 7ae6e69ddc96f5d8fb38208e46a2f44c7daca245
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 12:41:09 +0100

    Version 0.12.5

commit 1bc901f50cf3d5bb7dec0729e018ba00e24f6f06
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 12:40:43 +0100

    NEWS for 0.12.5

commit a89eb63f14b55f5e291490c8cbaa114e2a7a2b53
Merge: b183b4f b32a81d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-17 12:27:14 +0100

    Merge branch 'old-bugs'

commit b183b4f20361f96cc7326da62b1705bb4758f09e
Author: Olli Salli <ollisal@gmail.com>
Date:   2011-08-10 21:53:26 +0300

    Update NEWS

commit 5c201dffc0c71d1a0500a341709b4930d6187f97
Merge: df2e4bd 5a62326
Author: Olli Salli <ollisal@gmail.com>
Date:   2011-08-10 21:43:13 +0300

    Merge branch 'tube-caps'
    
    Reviewed-by: Simon McVittie (smcv) <simon.mcvittie@collabora.co.uk>

commit 5a62326a6d3e91820d6ccec78635151ec403dd1d
Author: Olli Salli <ollisal@gmail.com>
Date:   2011-08-10 19:22:02 +0300

    tube-caps: Test for a bidirectional (no Requested property) stream tube filter

commit bd22f64177ae9532b5dd3f6697355545bccc8625
Author: Olli Salli <ollisal@gmail.com>
Date:   2011-08-10 19:20:07 +0300

    Drop an extra == TRUE in a conditional

commit fb81504e14cbd02ae051f06bf39ed7b9f2e85280
Author: Olli Salli <ollisal@gmail.com>
Date:   2011-08-10 19:19:36 +0300

    private-tubes-factory.c: Use TP_PROP constants

commit c44f4f1c601b2f71d08567c94f349f8c719a1a36
Author: Olli Salli <ollisal@gmail.com>
Date:   2011-08-10 18:24:02 +0300

    tube-caps test: Test that Requested=True filters don't make tube services advertised

commit 8ca532d73b98c828b0620bd2ded172fd90a4b893
Author: Olli Salli <ollisal@gmail.com>
Date:   2011-08-10 17:52:48 +0300

    Don't advertise being able to receive tubes if we can only initiate them

commit df2e4bd487e26bf785dbdc85fe50db7a9d35c989
Author: Marco Barisione <marco@barisione.org>
Date:   2011-08-05 15:24:57 +0100

    Update wocky to fix PEP when power saving is enabled

commit cee0ee6d6a8daa2c32b139e28a88b07e3746853f
Author: Marco Barisione <marco@barisione.org>
Date:   2011-08-05 15:22:02 +0100

    Update wocky to fix PEP when power saving is enabled

commit 18b99be3be0c0b3ed31e52c25ac545ad6152e85d
Author: Marco Barisione <marco@barisione.org>
Date:   2011-08-01 14:20:34 +0100

    Update the Wocky snaphost and adapt the changes in WockyPing
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit 77155451e3145d0c856d4bd3d2e2ae53923463ad
Merge: fbb15c3 82d3fbe
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-03 10:21:10 +0100

    Merge branch 'telepathy-gabble-0.12'

commit 82d3fbe13b68caa242727ad7cc5d60e53926f7c9
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-03 10:11:36 +0100

    Connection: correctly install power-saving property.
    
    Previously :power-saving was actually mistakenly implemented as a
    synonym for :decloak-automatically. This only showed up when (on master)
    I changed the default for "decloak-automatically" to be TRUE (not FALSE,
    as is the default for :power-saving).

commit fbb15c3cb82e6c8944a8797e6f3b7a4a9c24b427
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-08-02 17:46:05 +0100

    Turn on DecloakAutomatically by default
    
    There should be UI to turn this off, but since people can (in general)
    IM you even if they're not subscribed to your presence, they should be
    able (by default) to call you too.
    
    Reviewed-by: Sjoerd Simons <sjoerd.simons@collabora.co.uk>

commit d5d3107396225812841078dc8f0e3661b916f506
Merge: 7db8678 d0707d6
Author: Marco Barisione <marco@barisione.org>
Date:   2011-08-01 17:38:23 +0100

    Merge branch 'telepathy-gabble-0.12'

commit d0707d61636e16b6acfbb84c0846b9373681b747
Merge: cbe0592 3b87f9d
Author: Marco Barisione <marco@barisione.org>
Date:   2011-08-01 17:36:23 +0100

    Merge branch 'initial-invisibility' into telepathy-gabble-0.12
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=38706>

commit 3b87f9d655f5ecdaca7fffa7de0394037fbec3df
Author: Marco Barisione <marco@barisione.org>
Date:   2011-07-26 14:46:32 +0200

    presence/shared-status.py: test invisibility works with updated servers

commit c2e977cf73bb961d3130428e2b5a68f1d54a42fe
Author: Marco Barisione <marco@barisione.org>
Date:   2011-07-26 14:39:20 +0200

    conn-presence.c: make shared invisibility work after the server update
    
    The new version of the Google Talk servers supports a newer version of
    shared status. Because of an unknown version string we were assuming
    invisibility was not supported. We now just fall back from "hidden" to
    "dnd" if another resource doesn't support shared invisibility.
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=38706>

commit 7db86780825ed1c3459c380fa5b0b209c6287fd3
Merge: 2ec472c 22f6849
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-29 18:54:28 +0100

    Merge branch 'null-caps-39464'
    
    Fixes: http://bugs.freedesktop.org/show_bug.cgi?id=39464

commit 2ec472c9deb1bbca578c4f3cb939ef8ae3065bd8
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-28 15:57:31 +0100

    test bytestreams: also only listen on localhost

commit 885d9cdc175f58c5490c64d38e585678e8e10ebb
Author: Paul Seidler <pl.seidler@gmail.com>
Date:   2011-04-03 15:16:37 +0200

    tests: use localhost
    
    https://bugs.freedesktop.org/show_bug.cgi?id=35968

commit 22f684987e53d0043a7739c6c1b1929f639c659f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-28 15:28:53 +0100

    Add a regression test for fd.o#39464

commit 147ebf5404c5818aba122e25cf1c7844cbd99a59
Merge: 5b423d6 49606ef
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 17:36:27 +0100

    Merge branch 'rudolph'
    
    Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>

commit 5b423d671523b7f3b28755d0de4a064540d8e814
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 16:25:57 +0100

    bump nano-version to 0.13.3.1

commit f50a1e625217d528c6a6406d6ce610773910d1d8
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 16:22:36 +0100

    make-release-mail: detect headers more robustly.
    
    I just broke this by having a line beginning 'telepathy-gabble' which
    was not actually a version header.

commit 7f64afa396c7e639e7920a60d46b980a12d5c2ee
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 16:01:51 +0100

    Version 0.13.3

commit b099c2a269804857cf41a393cca22f8b82704ec6
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 15:58:16 +0100

    NEWS for 0.13.3

commit b5ec55b7056634a679ce277f3575f09214d5372a
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 15:42:41 +0100

    Update Wocky submodule to fix fd.o#36077

commit a0ad5e286229d90b79942bf32e29e2874d278a31
Merge: 2b4881b cbe0592
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 15:40:16 +0100

    Merge branch 'telepathy-gabble-0.12'
    
    Conflicts:
    	lib/ext/wocky

commit cbe0592fbfbd2706f3c306da860b88e61e33a82e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 15:20:01 +0100

    bump nano-version

commit ecf29e6a17441829bd369bd2d25245c8dcb7aa73
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 15:07:59 +0100

    Version 0.12.4

commit 3ad658923c085f7ecad5cca06ab0de77fa7eadf4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 14:49:52 +0100

    Update Wocky submodule to fix fd.o#36077
    
    No other changes are included: this is the gabble-0.12 branch of Wocky.

commit 49606ef2f8637792e0b4fc5cfc77851259867fd5
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 14:08:18 +0100

    JingleFactory: stop including fake loudmouth!

commit 50aa29eceeaa7eefa336a5dee2641a02b2c8e846
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 13:40:44 +0100

    JingleFactory: wockify google:jingleinfo callback
    
    This squashes the spurious "ignoring jingleinfo from '%s', not ourself
    nor the server" debug messages that showed up on pretty much every
    incoming IQ set, because some idiot (me) put that check before the check
    for a <query xmlns='google:jingleinfo'/> element. Using
    _register_handler_from_server() and pattern-matching makes this a *lot*
    neater, I think—and objectively, it's half as much code.

commit 319c851c095dea4a4425f80c4db67947cb1abdbd
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-27 13:33:32 +0100

    JingleFactory: wockify jingle action callback
    
    This could be better: we could match dialects up-front. But hey.

commit fa61e14c10296b945cbb75585ad87d673b577426
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-26 15:37:54 +0100

    JingleFactory: use constructed, not constructor.
    
    This also expunges a weird FIXME!

commit 138cea6ca12d5a929d52aa2e86b9f0e34919e35a
Author: David Laban <david.laban@collabora.co.uk>
Date:   2011-07-22 02:28:15 -0400

    presence-cache: don't crash if computed_hash == NULL

commit 2b4881bb7df570c4840dc4cfdea8ee3ca9176c92
Merge: 4080a24 05faadf
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-13 17:13:09 +0100

    Merge branch 'encryption'
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=38719>

commit 4080a244d604207bac505073295b7ed1f1556538
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-11 17:30:04 +0100

    test-debug: use ProxyWrapper
    
    Obsessive? Moi?

commit 133c168d3a2838293c2d63b7796a6a942e67ea39
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-11 17:29:35 +0100

    ProxyWrapper: add a sensible ‘no other interfaces’ default

commit 818528a2f36a74c0a493fe77daa9f12f35de9285
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-11 17:20:45 +0100

    Test Debug object's absence with --disable-debug
    
    38a6178 changed test-debug.py to verify that the Debug object is present
    but non-functional when --disable-debug is passed to configure. Gabble
    actually doesn't have a Debug object at all in that situation, which I
    think is reasonable—so this now tests that calling GetMessages fails.
    The test doesn't make any particular assertions about the error. Right
    now it happens to be 'org.freedesktop.DBus.Error.UnknownMethod' but it
    really ought to be some kind of ‘unknown object’ method. So.

commit feed269a8d964c61fa6b271591ed2311e6fba75e
Merge: ebb6963 1e3b981
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-11 16:19:15 +0100

    Merge branch 'debug-stubs'
    
    https://bugs.freedesktop.org/show_bug.cgi?id=39046
    Reviewed-by: Guillaume Desmottes <guillaume.desmottes@collabora.co.uk>

commit 1e3b9812fce185efb830b5aec5f5912472e49cd4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-08 14:13:15 +0100

    Update Wocky for corresponding --disable-debug fixes

commit 99ee60ef205efa3d6b77134bd2003d373477f9cb
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-08 10:46:36 +0100

    Define a static inline no-op DEBUG with --disable-debug

commit ebb6963dc788faa9683f2b976b1db19187819351
Merge: 38a6178 74e4699
Author: Marco Barisione <marco@barisione.org>
Date:   2011-07-07 14:23:16 +0100

    Merge branch 'extra-certificate-identities'
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=38749>
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit 74e46998689e8049faa9eb87eca0886ab1f72481
Author: Marco Barisione <marco@barisione.org>
Date:   2011-06-28 15:39:33 +0100

    server-tls-channel.py: test that gabble verify hostnames
    
    Gabble should accept certificates for hostnames if they match the
    hostname in the JID or one of the hostnames in the
    extra-certificate-identities parameter.

commit 2672b6c7e03a303aed2649f925dac486ab9fb469
Author: Marco Barisione <marco@barisione.org>
Date:   2011-06-28 15:39:07 +0100

    Move the extra domains to trust from the TLS channel to the manager

commit ed99546dc9ee402f42727702305b3125a73724af
Author: Marco Barisione <marco@barisione.org>
Date:   2011-06-28 15:38:29 +0100

    TLSManager: check the connection before the non-interactive verification

commit 86726ddce8ad38cfca1a753d3b7b8940bfe5f82b
Author: Marco Barisione <marco@barisione.org>
Date:   2011-06-28 15:37:45 +0100

    ServerTLSChannel: make reference_identities a GStrv and not a GPtrArray
    
    In this case the GPtrArray doesn't give us any advantage and doesn't
    make clear that the content is a NULL terminated array of strings.

commit a6c42b97d4e1984451c919048cc04df72b4c454d
Author: Marco Barisione <marco@barisione.org>
Date:   2011-06-28 15:36:53 +0100

    Add an extra_identify argument to the TLS verification functions
    
    Update the Wocky snaphost and adapt the TLS manager to the changes.
    
    Now the TLS verification functions have an extra argument for the
    additional hostnames that we have to consider valid.
    This is useful, for instance, when using Google apps. Rather than
    providing a certificate for the JID’s domain part, the server provides a
    certificate for talk.google.com; if the user has explicitly configured a
    ‘Google Talk’ account, it's reasonable to accept certificates for this
    domain.

commit 38a6178a7529e3b1bf1ab8439185541e01c18dd5
Author: David Laban <david.laban@collabora.co.uk>
Date:   2011-05-31 16:10:16 -0400

    test-debug.py: actually check that no debugging happens
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit f7cd57841f39ca7d007623745ea2c33acaa97c88
Author: Vincent Penquerc'h <vincent.penquerch@collabora.co.uk>
Date:   2011-05-23 12:29:52 +0300

    Fix tp-yell includes being read from /usr/include instead of the submodule
    
    Since tp-yell gets installed in ${PREFIX}/telepathy-1.0, adding
    include paths for some other telepathy library (such as tp-glib)
    would automatically make tp-yell headers discoverable.
    
    Fixed by moving the tp-yell submodule paths before the ones for
    any telepathy library, so these will be found first.
    
    This fixes the build if you have a version of tp-yell installed
    which does not have a symbol that tp-gabble needs.
    
    https://bugs.freedesktop.org/show_bug.cgi?id=37492
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit c356abc2de2b33cb250b143bb5a8b28068971f4d
Author: Vincent Penquerc'h <vincent.penquerch@collabora.co.uk>
Date:   2011-05-23 12:29:52 +0300

    Fix tp-yell includes being read from /usr/include instead of the submodule
    
    Since tp-yell gets installed in ${PREFIX}/telepathy-1.0, adding
    include paths for some other telepathy library (such as tp-glib)
    would automatically make tp-yell headers discoverable.
    
    Fixed by moving the tp-yell submodule paths before the ones for
    any telepathy library, so these will be found first.
    
    This fixes the build if you have a version of tp-yell installed
    which does not have a symbol that tp-gabble needs.
    
    https://bugs.freedesktop.org/show_bug.cgi?id=37492
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit 5c06d57058f169401449039b7148470ca1784699
Author: Thomas Flueeli <tflueeli@gmail.com>
Date:   2011-02-02 22:59:56 +0100

    gibber: fix TCP connection establishment on Windows
    
    https://bugs.freedesktop.org/show_bug.cgi?id=31621
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit 9086cd4c9be315aa91a6284ffecb65178c89c8c1
Author: Thomas Flueeli <tflueeli@gmail.com>
Date:   2011-02-02 21:02:54 +0100

    gibber: add workaround for GLib bug on Windows
    
    https://bugs.freedesktop.org/show_bug.cgi?id=31621
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit 04763e4bc8ed32c208c75209b66221ab46290005
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-07-06 18:40:13 +0100

    Set the DBus_Property flag on DecloakAutomatically
    
    Without this, MC won't update the property on running connections.
    
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 51ddcba7fde8c2f9db5cd01930652318093c0032
Merge: f54614c 0acb80d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-28 14:00:59 +0100

    Merge branch 'more-alias-and-vcard-tweaks'
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=11321>
    Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>

commit 05faadffcce2573f788c979b9d85bf78721f84c4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-28 13:49:21 +0100

    Fix Make dependencies for write-mgr-file
    
    All the parameter information lives in protocol.c these days.

commit 0acb80d4c7669d26fe543e515473ad3f7dfbe11d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-28 10:59:20 +0100

    Correctly clear our own alias when asked to.
    
    Previously, Gabble would include an empty <NICKNAME/> node in your
    vCard, which kind of offends my inner purist: it should just leave it
    out.
    
    En passant, this should also fix clearing the name='' attribute on the
    roster for yourself (if you're on your own roster) but this is untested.

commit 0f30bcdcaab39a2ebcaafb42272056d79f126c11
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-28 10:57:41 +0100

    gabbletest: fix expect_and_handle_[sg]et_vcard
    
    I wrote these, and they've *always* been broken:
    
    • _get_ would include two <vCard/> elements in the reply: the first
      (from the query) would always be empty.
    • _set_ tries to update the global 'current_vcard' variable, but because
      it wasn't declared 'global' it did not update it.
    
    Astonishingly all the tests pass...

commit 4080addd6c2fa879b69e1e9b4567ccac0cf15697
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-28 10:30:24 +0100

    test-set-alias: clean up vCard check.
    
    This makes the test more thorough: it actually checks the structure of
    the vCard…

commit 22382935be99e027da46303107926a15cd8f3017
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-28 10:19:12 +0100

    Test setting our own alias via PEP
    
    Astonishingly, this was untested...

commit 17d99f5e7ed7b4837d5367950a99773fb61bf646
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-28 10:11:08 +0100

    Aliasing: handle SetAlias({contact: ''}) better
    
    Previously, if the user set a contact's alias to '', Gabble would change
    its alias on your roster to the empty string… this is not really ideal.
    
    This patch makes Gabble switch to using a cached remote alias for the
    contact, if any; if there is none, it simply removes the name=''
    attribute from the roster entirely, and tries to make a request for a
    better nickname.

commit abd99988770e2fdbd8bd811c8e08662fa21e7d0d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-28 09:57:20 +0100

    Aliasing: split out cached remote alias lookup
    
    There are no code changes in this patch: it simply moves the middle
    section of _gabble_connection_get_cached_alias(), which deals with
    looking for cached aliases that the contact specified for themself (on
    PEP, in their vCard, in their MUC JID, etc.) into a separate function.
    
    It also adds a docstring to _gabble_connection_get_cached_alias().

commit 928ef53affeb52d6ed380078d1acabd7f1b4dfec
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-22 00:27:42 +0100

    test-save-alias-to-roster: ack and echo roster pushes
    
    I was adding some more stuff to this test and realised that it's
    technically misbehaving as a server: the server MUST (of course) ack the
    roster update the client says, but it also MUST send roster pushes to
    all connected resources, event to the resource who made the change.
    Gabble relies on this, and only respects changes to the roster when the
    server pushes them to it. (I think this is fine, it means there's only
    one code path.)
    
    This test happened not to ever update a roster item more than once, so
    not acking the pushes didn't matter so much. It also never depended on
    reading back stuff that it added to the roster. But I'm about to add
    a section which will.

commit e4450664ff14d65b76f869aeb0aeb7cc9fe51438
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-22 00:26:25 +0100

    Aliasing: remove a hateful g_hash_table_foreach

commit 7bea4d961ec7361455258ed01750df28a7207c26
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-22 12:34:42 +0100

    ContactInfo: don't claim PHOTO is unknown
    
    Previously this code would log “unknown vCard node in XML: PHOTO” for
    every vCard that flew past. This is misleading: PHOTO is perfectly
    well-known, it's just handled by a different bit of code. So this patch
    adds the notion of an ignored field, which the ContactInfo code knows
    about but does nothing with.
    
    I didn't add any of the other fields mentioned in the comment at the end
    of the field table: they're not handled anywhere, so it may arguably be
    worth logging something about these.

commit b32a81d5436cb6f6d0c6435674fc424cabfed9e3
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-27 18:49:00 +0100

    MucChannel: fix coding style violations

commit 12ea43016a73e853a1a6ad8227f2a6f109474abf
Author: Michael Scherer <misc@mandriva.org>
Date:   2011-06-27 18:41:35 +0100

    Support XEP-0012: Last Activity
    
    Patch largely rewritten by Will, since Gabble has changed a lot in the
    two years since he rejected it on (in retrospect, shaky and
    unnecessarily hostile) grounds.
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=11688>
    Signed-off-by: Will Thompson <will.thompson@collabora.co.uk>

commit 901577fd23a6e7f7e639be6c8d8b086a3531a091
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-27 18:03:36 +0100

    Remove TpBaseConnection.connected callback.
    
    This OLPC-specific function can just be called from the ::status-changed
    signal handler. I've called it at the end of that function because the
    'connected' vfunc is called just after ::status-changed is emitted, so
    this shouldn't change the order of operations in the OLPC code.
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=12041>

commit 0f00b08488ef0459fdb6d8f85d55b049d4dc0245
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-20 12:51:40 +0100

    Enable require-encryption by default
    
    This seems like a reasonable thing to do in 2011.

commit f54614c40756f382d1a608054f1b711ff6390d5c
Merge: 7d58d83 11ddfc2
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 20:12:12 +0100

    Merge remote-tracking branch 'origin/master'

commit 7d58d83eee6624c66001582d1f56caadedaf82ab
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 20:11:30 +0100

    Bump version to 0.13.2.1

commit ea425adba81a2681af4c4ed7184972deae378331
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 19:57:18 +0100

    Version 0.13.2

commit ac2df5eacd34852d6d6ddf584bd2394812e66ad6
Merge: cf9ed52 fcfbc06
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 19:55:03 +0100

    Merge branch 'telepathy-gabble-0.12'
    
    Conflicts:
    	NEWS
    	configure.ac
    	src/presence-cache.c

commit fcfbc06d3f679a8266f13921ff3771bd07cf7f24
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 19:50:08 +0100

    bump version to 0.12.3.1

commit a1d33d82c8645303e279e20033d8f96bc6f6e567
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 18:48:03 +0100

    Version 0.12.3

commit 18a8b11833af6b8b2e62e4ee3bd3037a8b5b2c2f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 18:47:45 +0100

    NEWS for 0.12.3

commit 712bbe2b59ed5746cee50604956b0cb842ea9cc4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 19:35:23 +0100

    FTChannel: don't assert about handle-type in setter
    
    telepathy-glib's default for TpChannelIface:handle-type has changed in
    0.15.2: it used to be NONE, and now it's UNKNOWN. This broke both Salut
    and Gabble…
    
    But as the comment says, it's not meaningfully writeable. It's not at
    all obviously why this assertion is there.

commit 276abd5ff84d635eb5f8cd3eee56978f75631722
Author: Colin Walters <walters@verbum.org>
Date:   2011-06-24 18:42:23 +0100

    Honor NOCONFIGURE for compatibility with gnome-autogen.sh
    
    See also:
    http://people.gnome.org/~walters/docs/build-api.txt

commit 11ddfc2e24bbb2fc5297c714dbea88c377a1911e
Merge: cf9ed52 1ff589c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 15:48:44 +0100

    Merge branch 'telepathy-gabble-0.12'
    
    Conflicts:
    	src/presence-cache.c

commit 1ff589c1e89cf9bb90f59a961c23f2ffa9c3dff0
Merge: cdff268 f7cdb73
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-24 15:31:29 +0100

    Merge branch 'fd.o-38603-initial-contact-presence' into telepathy-gabble-0.12

commit f7cdb735461e81b428f9a5d776e2b2487263c8aa
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-23 16:53:23 +0100

    Cite fd.o#38603 (presence-before-roster doesn't work)

commit fa0990a9810b9629d32dd53840aa415e2d5dbe7d
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-23 16:39:59 +0100

    Roster: don't emit presence updates for available contacts
    
    Previously, when we received the roster we would announce that the
    presence of everyone on it we're subscribed to has changed, just for the
    side-effect of signalling that offline contacts really are offline, not
    just unknown.
    
    This offended me. :)

commit d1f7941c3343b1c498e55543c66ae72fa2b880f5
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-23 16:28:20 +0100

    initial-contact-presence: reduce duplication of presences

commit e0cda613a166e9c538ccdcd8cf5d5feae72a2596
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-23 16:19:11 +0100

    GabblePresence: start in state Unknown
    
    This has the effect of ensuring that receiving <presence
    type='unavailable'/> for a contact not previously in the presence cache
    will not cause a change from Unknown to Unknown to be signalled.

commit 66788076de9fb0062d685346d4bb5c444befd3ea
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-23 15:58:11 +0100

    PresenceCache: discard UNKNOWN presences in maybe_remove
    
    In various places, gabble_presence_cache_maybe_remove() is called to
    allow the cache to relinquish its grip on content-free offline
    presences. It should also be able to drop unknown presences: they're
    basically equivalent, based on whether or not the contact is on our
    roster.
    
    Specifically, this means that UNKNOWN presences for contacts received
    before the roster has arrived do not linger in the cache, and hence
    these contacts are correctly signalled as OFFLINE when the roster
    arrives.
    
    Really, I think a better fix would be to change the presence cache to
    not have this distinction between Unknown and Offline: it's completely
    based on whether or not the contact is (known to be) on the roster, and
    this is reflected in the way that the handler for incoming <presence/>
    passes a different value into the presence cache based on whether or not
    the contact is on the roster. But this would be a much more invasive
    change, so I'm not prepared to make it just yet.

commit 01d600f1f9f52ac7f021fed2a75e6aa8e7168274
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-23 15:46:28 +0100

    Presence: debug GabblePresenceId names where possible
    
    (I may have missed a few places, but these are the two I came across.)

commit b2616cbb5e25fcd28dbeaad68055b18a399db8b5
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-23 13:48:27 +0100

    Build GEnum/GFlags for enums in connection.h
    
    Specifically, I want a GEnum for GabblePresenceId so I can include the
    stringy versions in debug strings.

commit a6c0404419c7c84924e1b98783860f8ad99aa18f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-23 12:51:58 +0100

    initial-contact-presence: split expect from assertion
    
    I think this makes things clearer: we're expecting some presences to
    change in response to the roster arriving, and we're expecting the
    changes to be these.
    
    This means that the test failing manifests itself as an assertion
    failure, not a timeout.

commit cf9ed5291015fd97b2121dcf38933e8bf885013d
Merge: 0b6f5f5 cdff268
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-21 14:24:00 +0100

    Merge branch 'telepathy-gabble-0.12'

commit cdff268c3965046baa37aac100cceda3e9aafe37
Merge: 63c4d27 6df83a1
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-21 14:21:26 +0100

    Merge branch 'aliases' into telepathy-gabble-0.12
    
    Reviewed-by: Sjoerd Simons <sjoerd.simons@collabora.co.uk>

commit 6df83a17ff312dc3ad99c79bb40163eba687449c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-21 12:25:05 +0100

    Look for better aliases if roster says name=jid.
    
    When we discover that a contact doesn't have an alias at all, we fall
    back to their JID, and write this to the roster (as we do for any other
    alias we use, to spare our users' precious bandwidth).
    
    This is fine and all, but it means that we ignore any subsequent updates
    from the contact to say that they have a better alias. For instance, we
    might get a PEP nickname update, or fetch their vCard for some other
    reason (getting an avatar, ContactInfo, &c &c). In those cases, we would
    like to have any better alias obtained passively to replace the JID
    fallback. This is only ever passive: we still don't query for a better
    alias if the roster says it's the JID. (Clever users can set the alias
    to the empty string, which has the effect of forcing a refresh.)
    
    Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=27361

commit 7a221f03f2a8796895bd48febe1b04168d1b5a1a
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-20 17:00:56 +0100

    servicetest: Stringify events more nicely

commit e619592d8bfbba0d0679f918f514fdfc08b829cb
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-20 16:51:08 +0100

    gabbletest: make stanza events subclasses of Event
    
    This looks neater to me.

commit a85a67c99ab9085eafc8afe17437253a082e0e08
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-20 16:13:49 +0100

    Ignore empty name='' attributes on the roster.
    
    Previously, if a roster item had an empty name='' attribute, Gabble
    would be content to show this to the user, and never try PEP or their
    vCard for a better nickname.
    
    The empty alias is never useful.

commit 6ba62283b7052ff49cc223b65d4ac06937ec1043
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-20 16:09:50 +0100

    rostertest: support name='' attrs in roster pushes

commit 7ac26b27921b9217066c68c2a3333ff616c69a31
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-20 13:26:58 +0100

    test-save-alias-to-roster: use contact list fetching helper

commit c033c160dfd91a9253aaaac7804922067116cb4e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-20 13:17:55 +0100

    test-save-alias-to-roster: expect AddMembers to return
    
    This is an ancient FIXME. I think it's safe to say that we depend on a
    new enough tp-glib.

commit 0b6f5f529ac72a479956663175cb6205c41ea394
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-16 18:02:03 +0100

    Bump nano-version to 0.13.1.1

commit 17e3e817f5faa349f08b8b541c1ecd65f83234a4
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-16 17:46:27 +0100

    Version 0.13.1

commit bd88fdcc79243df03d3fa31259696e1743bd7b78
Merge: a2d6254 63c4d27
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-16 17:45:23 +0100

    Merge branch 'telepathy-gabble-0.12'
    
    Conflicts:
    	NEWS
    	configure.ac

commit 63c4d275abab3aad53923f556473ee902965e2b6
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-16 17:37:36 +0100

    National novel versus ionization

commit 5e59111c54577650316749a262d6c891f415e603
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-16 17:22:32 +0100

    version 0.12.2

commit 052ca8131c8e145714372db249efd9bd04c70f94
Merge: e8e2d23 a88dd55
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-16 17:07:55 +0100

    Merge branch 'gtalk-jingle-workarounds' into telepathy-gabble-0.12
    
    These aren't perfect, but they'll do for now.
    
    https://bugs.freedesktop.org/show_bug.cgi?id=38352

commit a88dd5529d038f32b3c01b9eb6109eaa10e88604
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-15 18:08:08 +0100

    Add a test for working around GMail

commit 714e066f8b566f107cef819d4ae6fb7786c8739f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-15 17:53:17 +0100

    JingleSession: special-case that GMail can't modify contents.
    
    In theory, it can, because it now speaks modern Jingle which has this
    concept. But in practice it cannot, at least for now.

commit b1e774762806b1e0446aaccd42657437e96b01a2
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-15 17:44:40 +0100

    MediaChannel: clean up setting ImmutableStreams

commit bbf7da0ca74749cfec6701e9cd16f8d7595c4bb8
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-15 17:38:31 +0100

    RTP: use video_rtp/_rtcp component names with GMail

commit fcc5df48e59c4141151fe6c9d8348a1b676cc88c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-15 17:13:12 +0100

    JingleContent: work around GMail omitting content creators

commit a29379f10e8ff96bfcd206cd92a12cff25bc5c72
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-15 16:51:33 +0100

    PresenceCache: set Omits Content Creators quirk for GMail

commit 519504984f9ec70657fdf0bff0eb580bee253237
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-15 16:50:12 +0100

    JingleSession: refactor quirk-checking code.

commit cd4a003443ff6f9b6e022e9a21146e59c2c56463
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-13 20:15:09 +0100

    TransportIface: correct a contents vs. candidates thinko

commit a2d6254646182c10d55290f8b3ce53d7a399ca96
Merge: 0a7c31e e8e2d23
Author: Olivier Crête <olivier.crete@collabora.com>
Date:   2011-06-13 15:16:07 -0400

    Merge branch 'telepathy-gabble-0.12'

commit e8e2d23490846ebabb7c640bb905b7549883b3af
Author: Olivier Crête <olivier.crete@collabora.com>
Date:   2011-06-13 14:20:53 -0400

    jingle-media-rtp: Codec encoding names are not case sensitive

commit 0a7c31ea51c5096e86111f73f157bb87f0f46d0e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-02 19:46:38 +0100

    Post-release nano-version and NEWS bump to 0.13.1

commit c4a118850bcaccd6502b6a19cbb61238de1d157e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-02 19:20:59 +0100

    Version 0.13.0

commit 3fbd382bf99ce3584d9e167f60c47b9f973cd8c3
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-02 19:20:36 +0100

    NEWS for 0.13.0

commit 9f1fa8b567e7f792013ab8a624ad46aa0735e606
Merge: 8d5a2f5 828f2e7
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-02 17:10:08 +0100

    Merge branch 'telepathy-gabble-0.12'
    
    Conflicts:
    	configure.ac
    	lib/ext/wocky

commit 828f2e79137acfd1079a6d1efcc527d8bec724b1
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-02 17:06:55 +0100

    Post-release nano-version and NEWS bump

commit fe551853438ca44f6f28e5ce334a53e9b9c6b693
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-02 16:11:37 +0100

    Version 0.12.1

commit dbc9a937f335ac403436c6fddbd0129693247e49
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-02 16:08:51 +0100

    NEWS for 0.12.1

commit b756564ad0a27e1d210b38bf7583ec41d7007574
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-01 14:54:45 +0100

    httptest: chain up from HTTPFactory constructor.
    
    This fixes jingle/google-relay.py with recent versions of Twisted Web.

commit 8f19db01eac316b81d76f4533399fac2d05d72ac
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-02 14:59:09 +0100

    Update Wocky snapshot to fix self avatar in MUCs
    
    (That is: fd.o#32017)

commit 8d5a2f593c4f6c3b2db161be74a03bb3f05270b9
Merge: 5bfe3ed 34b9066
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-01 18:12:51 +0100

    Merge branch 'fd.o-32017-muc-self-avatar'

commit 34b90665e4860ff5dc022cc7549f070d1de645de
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-31 15:07:24 +0100

    Add a regression test for avatars in MUCs.

commit ef057a2fb297ce639618e89dae4b6cdbd732ff3c
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-31 17:42:51 +0100

    Update Wocky snapshot for WockyMuc changes
    
    This update changes the signature of various signals to include MUC
    status codes as bits in a guint rather than in a hash table. It also
    fixes WockyMuc stealing <presence/> stanzas before Gabble's handlers
    have a chance to get stuff out of them; test for that to follow.
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=32017> but I am not
    very happy with the fix.

commit 4f80d8e0aa86a0b8759c29f0e9e91aaeb909e76e
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-31 15:14:46 +0100

    mucutil: return event from try_to_join_muc

commit 0fb508d1b269069aa25f701f82e14060b90051ef
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-06-01 14:54:45 +0100

    httptest: chain up from HTTPFactory constructor.
    
    This fixes jingle/google-relay.py with recent versions of Twisted Web.

commit 5bfe3ede2c17b7dd10d258e5661892967cabcc60
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-24 19:41:00 +0100

    Explicitly cast to guint64 when passing varargs
    
    Without any additional type information, integers used in varargs lists
    can get passed as the wrong type. In this call to tp_value_array_build()
    in gabble_muc_channel_constructed(), we should be passing 0 as a a
    gint64, but I think in practice we're passing it as a pointer-sized
    thing. (I'm not a C language lawyer, I'm sure someone could correct me
    on this.) If the type we're passing it as is not the same size as a
    gint64, this will cause tp_value_array_build() to read off the end of
    its varargs, probably crashing in the process (because what it reads
    will probably not be a valid GType).
    
    The other call to tp_value_array_build() added in ec8ca6e (for the Room
    interface) does cast correctly.
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=37165>
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 9b81db6ba12edb6e5c8f9b901f83b9a92d9e23fa
Merge: e27aa6f 04f044f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-24 14:17:52 +0100

    Merge branch 'telepathy-gabble-0.12'

commit 04f044fe0ce68dd73d6bb92477106a09fd43dbdd
Merge: 53401b9 487dce0
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-24 14:17:17 +0100

    Merge remote-tracking branch 'origin/telepathy-gabble-0.12' into telepathy-gabble-0.12

commit e27aa6f6fd58533c22b0a2c30579f969378bad3e
Merge: efaebdd 53401b9
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-24 14:16:31 +0100

    Merge branch 'telepathy-gabble-0.12'

commit 53401b9d16dc6ed2c2e15b7af5f7c2ad254aac1f
Author: Chandni Verma <chandniverma2112@gmail.com>
Date:   2011-05-10 20:25:43 +0530

    Filter out google-rbc-announcement messages
    
    They are server generated containing no interesting information
    and appear in clients as if sent by peer
    
    Fixes: https://bugs.freedesktop.org/show_bug.cgi?id=36647

commit efaebdd96fdfec40aec3e257176e9d16faf44112
Author: Debarshi Ray <rishi@gnu.org>
Date:   2011-05-24 01:49:08 +0530

    Update wocky (fixes build failure with GCC 4.6)
    
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=37518>

commit 51fdf722717a5646116b554f5feb860d82a26566
Author: Siraj Razick <siraj.razick@collabora.co.uk>
Date:   2011-05-23 19:43:57 +0530

    fixes : telepathy gabble build break
    
    build breaks due to my previous two patches when doing make check
    this commit fixes those errors
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit 1794df70dd67d70e6c8649e47a4f3719fee268d7
Merge: 987722a 4087b2a
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-23 14:18:37 +0100

    Merge branch 'error_reports'
    
    Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>

commit 4087b2a87a9ceb6553030f3514771e7e8ffaf2f8
Author: Siraj Razick <siraj.razick@collabora.co.uk>
Date:   2011-05-20 23:52:14 +0530

    gabble_message_util_build_stanza: clearer renaming of variables
    
    The commit renames INVALID_ARGUMENT to RETURN_INVALID_ARGUMENT,
    and msg to stanza renamed msg to stanza, the difference between
    message (a TpMessagerepresenting an abstract message) and msg
    (a WockyStanza representing an XMPPXML stanza) would be clearer.
    (based on the comments by:Simon McVittie on f.d.o BUG#33460)

commit 1c86beea6be1cb0b5b946da2345d98139072a3ea
Author: Siraj Razick <siraj.razick@collabora.co.uk>
Date:   2011-04-28 18:53:23 +0530

    Fixes : fd.o#33460  by using wocky_porter_send_async()
    
    Bug - 33460 - IM and MUC channels should use wocky_porter_send_async(),
    and emit failing delivery reports
    
    url : https://bugs.freedesktop.org/show_bug.cgi?id=33460

commit 5aa9d93472ba3a90410496f453d861ce16f6bd5a
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-19 16:47:11 +0100

    Smoke-test MUC delivery report flags
    
    Reviewed-by: Simon McVittie <simon.mcvittie@collabora.co.uk>

commit 5ba21edcf88ef60be5558acc0b0259b5f29e98e9
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-05-18 16:06:52 +0100

    muc-factory: don't use a string after freeing it
    
    Uh oh!
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 987722a87f929035c4c6dcbcd069e0ba414b25ee
Merge: ba3c764 487dce0
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-17 16:42:29 +0100

    Merge branch 'telepathy-gabble-0.12'

commit 487dce00f8f54ea30edb807b1e1c8eb3c717a85d
Merge: e704ea0 a776a13
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-17 16:33:19 +0100

    Merge branch 'shared-presence' into telepathy-gabble-0.12
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=37283>

commit a776a13404c9af8f29a0e52d8645bd66622c27d4
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-17 11:41:47 +0100

    presence/shared-status.py: test that remote presence doesn't override away

commit ba640e0f9d1fd0ab7aa42e21ec5ad2f2bee5622b
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-17 15:51:34 +0100

    presence/shared-status.py: don't use a weird and avoidable "or" construct

commit ca600286bd3f42771a2b8bd35ca2e5d278644d45
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-17 11:39:40 +0100

    conn-presence.c: don't override away when the remote status changes

commit ba3c764bf634fad3af4ca0a8f35b180ca9ba221d
Merge: 52b2d76 e704ea0
Author: Simon McVittie <simon.mcvittie@collabora.co.uk>
Date:   2011-05-16 10:12:01 +0100

    Merge branch 'telepathy-gabble-0.12'

commit e704ea03ac1cc6a4e38670e3b8f89386154f8894
Author: Simon McVittie <simon.mcvittie@collabora.co.uk>
Date:   2011-05-10 18:52:07 +0100

    set_status_to_connected: do nothing if we already disconnected
    
    Now that we're making more use of GAsyncResult, we can easily get into
    this situation:
    
    * all but one of the preconditions for being CONNECTED have happened;
      the remaining one uses GAsyncResult (currently that can only be
      conn_presence_set_initial_presence_async)
    
    * conn_presence_set_initial_presence succeeds, and schedules a call
      to its callback in an idle; we are now ready to be CONNECTED, but because
      of GAsyncResult calling conventions, we won't call the callback right now
    
    * someone calls Disconnect(), and we do so, synchronously; we are now
      DISCONNECTED
    
    * the idle goes off and connection_initial_presence_cb is called,
      with success, while DISCONNECTED!
    
    Bug: https://bugs.freedesktop.org/show_bug.cgi?id=37078
    Reviewed-by: Vivek Dasmohapatra <vivek@collabora.co.uk>

commit 52b2d7691404457a559fc3fbb09e6f064d7dc33b
Merge: 02ad9c4 8f08873
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-12 16:55:32 +0100

    Merge branch 'telepathy-gabble-0.12'

commit 8f088732c364f1b77f805d65590e30abaf61a707
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-12 16:40:38 +0100

    conn-presence.c: remove a line forgotten when fixing a merge conflict
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>

commit 02ad9c4dcb40e047eab8fb8777ddf1ccb5e35dce
Author: Debarshi Ray <rishi@gnu.org>
Date:   2011-05-12 10:10:23 +0300

    Update the release mail automation script to contain the Git URL

commit 44c5d0ec3de8014efa79f15d85c630ee120db0c9
Merge: d4ddbf7 f99b6f4
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-12 14:20:11 +0100

    Merge branch 'telepathy-gabble-0.12'

commit f99b6f43f92299c7eb99ccbfef7b806f5b52c1de
Merge: fe15601 d5421d1
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-12 13:07:21 +0100

    Merge branch 'shared-presence' into telepathy-gabble-0.12
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=37069>

commit d5421d18d1fcb03a86ecfb81e37029c3f015c791
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-10 18:46:05 +0100

    conn-presence.c: handle the transition from hidden to away on GTalk

commit d895028ef03577768e19476e263455e7f7cbe4a7
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-10 18:33:12 +0100

    conn-presence.c: add a function to know if a GabblePresenceId is away

commit c5fc80a4d42c794d43332b3626e9ec50ea03580a
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-10 14:33:22 +0100

    conn-presence.c: leave an away status using <presence/> too
    
    When setting an away status on a server that supports shared status, we
    need to use <presence/> instead of shared status. When we then leave the
    away status, we want to use the shared status, but GTalk also expects us
    to also send a <presence/>.

commit 1ee540de1b939c676a76cbcf046a4cfb7b77deab
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-10 18:47:06 +0100

    presence/shared-status.py: test the transition from hidden to away

commit 951991f33fea6525469cb712d205ffd1ba2b1eec
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-10 14:29:06 +0100

    presence/shared-status.py: test the transition from away to non-away

commit d669b203f4c09396b3ff04c6dd392c314211ee59
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-10 14:04:34 +0100

    presence/shared-status.py: fix a typo (even → event)

commit cb9807eb74a70df762d173b57ee04368572c1113
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-10 13:57:36 +0100

    presence/shared-status.py: don't unforbid events too early

commit d4ddbf79d3b4b845a01cdfcfb14b024d69105dc4
Merge: 6ad9e1e b85a525
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-05-06 11:34:42 +0100

    Merge branch 'room'
    
    Conflicts:
    	src/muc-channel.c
    	tests/twisted/constants.py
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 6ad9e1e3d6a553405394032ae7258beff37c6f2a
Merge: d467f27 fe15601
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-05 15:26:02 +0100

    Merge branch 'telepathy-gabble-0.12'

commit fe156016db44b3cfd8bb3276d3c5a8944e7ff22f
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-04-15 14:27:07 +0100

    Don't assume google:queue support from google:roster
    
    Dave Cridland asked the Google Talk server team to advertise a
    google:queue stream feature, and as of this month some time, they do! So
    we can stop looking for google:roster as well.
    
    Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=36260>

commit 5d2927d9155d2388145406c189fdf9d9f8dfcac6
Merge: cacd814 e4ae0fe
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-05 15:01:59 +0100

    Merge branch 'shared-presence' into telepathy-gabble-0.12
    
    Reviewed-by: Will Thompson <will.thompson@collabora.co.uk>
    Fixes: <https://bugs.freedesktop.org/show_bug.cgi?id=36058>

commit e4ae0fe527dbbffe403026b25eb4c44b3b325715
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-03 18:43:22 +0100

    presence/shared-status.py: test that "chat" falls back to "available"

commit d010bb78008681707fc6cd4b4c2393c15affd442
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-03 18:41:05 +0100

    conn-presence.c: don't allow the "chat" status when using shared status

commit 42ca8dd32fa930c85fcdddc16b64220d45429f4f
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-03 16:19:06 +0100

    conn-presence.c: simplify the non-connected case in status_available_cb

commit 49633baeb861f9798cf87809f7421a087e9c9cff
Author: Marco Barisione <marco@barisione.org>
Date:   2011-04-18 18:32:39 +0100

    presence/shared-status.py: always expect an initial <presence/>

commit c2b8e2d64edfec296554f72a2d59cca26cbacfe4
Author: Marco Barisione <marco@barisione.org>
Date:   2011-04-18 18:29:20 +0100

    conn-presence.c: send an initial <presence/> when using shared status

commit c0ff90f6850ff80e682a86d45a4a5bda27e75425
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-04 11:54:25 +0100

    presence/invisible_helper.py: don't handle privacy lists in XmppXmlStream

commit 6bc9759f7140d0ac544659a4a6776c57e405699f
Author: Marco Barisione <marco@barisione.org>
Date:   2011-05-04 11:53:14 +0100

    presence/shared-status.py: don't allow privacy list stanzas

commit d04b628113fb93300f6dc82e88addbcf1e58b80e
Author: Marco Barisione <marco@barisione.org>
Date:   2011-04-15 17:13:11 +0100

    conn-presence.c: don't try privacy lists when using shared status

commit 06cc941cbada06c876e922df097d93956d88e00f
Author: Marco Barisione <marco@barisione.org>
Date:   2011-04-07 18:44:14 +0100

    presence/shared-status.py: test the correct behaviour with away and xa

commit 2a53547a92ec30a31cf4442e60f3f318269952d6
Author: Marco Barisione <marco@barisione.org>
Date:   2011-04-07 18:41:30 +0100

    conn-presence.c: don't use shared presence for away statuses
    
    Away and extended away are mapped to idle in GTalk and are per
    connection, so they are not set through shared presence.

commit 2fb53a5a09c7dea0eee5bf7f54ab0d6c24e8b246
Author: Marco Barisione <marco@barisione.org>
Date:   2011-04-07 18:40:43 +0100

    conn-presence.c: fix indentation

commit c49ea89ae115ad8a0c537a1a491a79604c819072
Author: Marco Barisione <marco@barisione.org>
Date:   2011-04-07 18:38:53 +0100

    conn-presence.c: don't send <presence/> if we already used shared status

commit d467f276329264be0b24f5af6c854fd7a182b531
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-05-05 12:05:33 +0100

    Bump micro version for unstable branch.
    
    The next release from master will be 0.13.0

commit 160289f855182e9e2d00a91b38541d8e989eee17
Merge: 4722b73 0d865ef
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-05-02 09:32:54 +0100

    Merge branch 'caps-hash'

commit 0d865efce5ca0840925abdca673a3f171258f767
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-05-02 09:11:57 +0100

    caps-hash test: removed
    
    This has moved to wocky.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 4722b731516f98a810b97d1be0c5762cb87d9432
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-05-02 09:32:02 +0100

    update wocky snapshot
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit cacd8147f3d0c0f4b0b851731b76dcacb4a9fb35
Author: Will Thompson <will.thompson@collabora.co.uk>
Date:   2011-04-21 18:48:03 +0100

    naaaa nanananano version katamari damacy

commit 8b00635e0c048bae7219c06f5e18ac8b33a9dacf
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-02-04 11:16:15 +0000

    caps-hash: misc fixes to cleanup code
    
    wocky_caps_hash_compute_from_lists allows dataforms to be NULL if
    there aren't any and will no longer assert on dataforms=NULL.
    
    Also, we no longer dup the feature strings as we free the pointer
    array in the same function so we know what the memory allocation of
    the array is.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit d34197fb00196aff94142d38216acf4efc36d2af
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-02-04 10:21:09 +0000

    caps-hash: use wocky_caps_hash_compute_from_lists
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit c7bba6807c7c236a3a7c7121e1e096f626c56dcb
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-02-04 10:19:31 +0000

    connection: clarify add_sidecar_own_caps' identity arg
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 0917818df48202cf979e5911c968966b6384a84a
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-02-04 10:19:14 +0000

    presence-cache: don't try and copy a NULL disco identity array
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit c330f99490d65b87698ecc3677cc7a521955a74a
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-02-03 09:32:28 +0000

    caps-hash: remove anything to do with dataforms
    
    We don't need it for self presence, and we're moving to Wocky.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit bfd15d09556a01079842e5f56b2fb3729f991cf8
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-24 13:05:41 +0000

    caps-hash: remove _from_lm_node
    
    We're slowly moving this into Wocky.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 969c73f63979dada565e9c36d51d34da0e5a68c3
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-24 13:04:21 +0000

    disco-identity: remove
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit df58879f27bced1367fe3c9011cce733952cfae5
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-24 12:48:17 +0000

    use WockyDiscoIdentity instead of the gabble version
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 926a07f594b303888ab98fe71735d288dfaf62cf
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-24 12:47:57 +0000

    use wocky_caps_hash_ functions
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit b85a52558e9260cf28dba25b60f894aa8075383c
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-13 09:20:14 +0000

    muc-factory: remove more assertions with side effects
    
    I forgot these last time, whoops.
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 0986e22bcbff90cc12d4bef5cf27e42d8bde9853
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-13 09:18:14 +0000

    muc-factory: improve debug messages when TargetID and room props conflict
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 9806dea0ec26d3803f6f05e12666296bbfa527b7
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-04 08:47:20 +0000

    muc/room.py: fix copy-paste error
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 921f6ee9813814baead9a0672af6b42beab654c9
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-04 08:46:18 +0000

    muc-factory: give temporary variables better names
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 558fd89929b0b4dc978f0e02b9f694ab7f93ffb5
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-04 08:42:33 +0000

    muc-factory: also note why asserting on decoding a JID
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 86f10810b3f244d34d2cc79ac3f46176fabe1c64
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-04 08:40:58 +0000

    muc-channel: re-add comment about valid TargetID
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 9ca2e3f94bd2ddb062d5eed6aff2f64fc929ad2b
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 16:45:05 +0000

    muc-factory: ensure RoomID and ServerID don't conflict with TargetID
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 35356cd6fda7dd10d223fd99b204c51c7aa3c28c
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 16:03:44 +0000

    muc-channel: check the result of gabble_decode_jid properly
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 23d4b72664d671374b6eaa48b703d7ea8e936f7d
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2011-01-03 16:01:51 +0000

    muc-channel: add newlines when calling g_object_get for clarity
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 4312f26c61c9a61ca007575d23a521500931bc8b
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2010-12-23 17:06:06 +0000

    tests: add a test for Room interface
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 637ba03783fff269933a7802728d680f5a8b82e6
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2010-12-23 17:04:55 +0000

    muc-factory: allow channel requests with RoomID + Server set
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 51c40a378b1b1434b6ad3705fbd2744216c35bd8
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2010-12-23 12:22:57 +0000

    muc/subject.py: test the Room Subject property is correct
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit a8053079c2fafe334684be6ee34b768f89abd8ab
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2010-12-23 12:22:34 +0000

    muc-channel: set Room properties as appropriate
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit ec8ca6ea5d945fae643b45cc5fdaa757cb070afa
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2010-12-23 10:30:08 +0000

    muc-channel: implement Room properties
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>

commit 7c67a8eeac4a65597beb860a9750e4cd639061b1
Author: Jonny Lamb <jonny.lamb@collabora.co.uk>
Date:   2010-12-23 09:54:43 +0000

    extensions: add Chan_Iface_Room draft
    
    Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
