| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
|
|
| |
https://bugs.freedesktop.org/show_bug.cgi?id=55100
|
| |
|
|
|
|
|
|
|
|
| |
by _dup_
New transfer and naming policy has been discussed in
https://bugs.freedesktop.org/show_bug.cgi?id=39189 and is
documented there: http://telepathy.freedesktop.org/wiki/Style/TelepathyGLib
|
|
|
|
| |
We can use g_list_free_full and (_tp_)g_list_copy_deep in more places
|
|
|
|
|
|
|
| |
When ObserveChannels comes in we are given all the immutable properties
for the statisfied requests straight away, so lets use them as well.
This prevents poor application authors from being confused about their
hints not coming through in their observer.
|
|
|
|
|
|
|
| |
When HandleChannels comes in we are given all the immutable properties
for the statisfied requests straight away, so lets use them as well.
This prevents poor application authors from being confused about their
hints not coming through in their handler.
|
|
|
|
|
| |
This fixes a crash in empathy-chat:
https://bugs.freedesktop.org/show_bug.cgi?id=51444
|
| |
|
|
|
|
| |
https://bugs.freedesktop.org/show_bug.cgi?id=49372
|
|
|
|
| |
We deprecated this in 0.11.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
| |
GAsyncResult guarantees to call your callback from the main loop.
For historical reasons (basically "5 years ago I didn't know any
better"), TpProxy does not: if the interface is missing or the proxy
has been invalidated, the callback is called re-entrantly. I'm going
to fix that in Telepathy 1.0, but for now, let's give GAsyncResult
the correct semantics.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=45514
Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
|
|
|
|
|
|
|
|
| |
fd.o#48363 was opened which complains about lots of documentation
problems (which g-ir-scanner finds). While I was at it I did some
grepping to find yet more problems.
Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
|
|
|
|
|
|
| |
It is not always a good idea for bindings to always rely on GObject::get_property(),
it can be slow to access properties that way everytime instead of keeping
a proxy object in native language.
|
|
|
|
|
|
|
|
| |
In contrast to the previous commit, I'm just using NULL here -
telepathy-glib has an explicit dependency on GLib 2.30.
Bug: https://bugs.freedesktop.org/show_bug.cgi?id=46523
Reviewed-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
Replace g_(ptr_)array_free (foo, TRUE) and g_hash_table_destroy
with respectively g_(ptr_)array_unref (foo) and g_hash_table_unref.
I used this command to generate this patch:
for f in `find -name "*.c"`; do sed -i $f -re 's/g_ptr_array_free \(([^ ,]+), TRUE\)/g_ptr_array_unref \(\1\)/'; done
See Danielle's blog for explanation of possible bug _free can do:
http://blogs.gnome.org/danni/2011/11/16/mistakes-with-g_value_set_boxed/
|
| |
|
| |
|
|
|
|
|
|
|
|
|
| |
A factory is what we really need, forcing to create a TpAccountManager
also means that a simple channel handler will always introspect
all accounts even if it needs only one.
With a factory, TpBaseClient can create only needed accounts and still
share them with a TpAccountManager if one exists.
|
|
|
|
| |
features are set on TpSimpleClientFactory now
|
|
|
|
|
|
|
| |
desired features
There are lots of code duplication for those proxy creation and preparation,
this will be fixed later.
|
|
|
|
| |
Use instead tp_simple_client_factory_ensure_account()
|
|
|
|
| |
This is much nicer C API, and TpBaseClient had those helpers too.
|
|
|
|
|
|
| |
Tp{Simple,Automatic}ClientFactory
Add internal wrappers for deprecated functions so we can still use them for compatibility
|
|
|
|
|
| |
For compatibility, still use TpAutomaticProxyFactory if one is explicitely set,
but use TpAutomaticClientFactory otherwise.
|
| |
|
|
|
|
| |
To match the name of the function that will be included in GLib 2.30
|
| |
|
| |
|
| |
|
|
|
|
| |
Actually if that var did not receive a value, it already does a goto out anyway
|
| |
|
|
|
|
| |
org.freedesktop.Telepathy.ChannelRequest.DelegateToPreferredHandler support (#38249)
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
|
|
|
| |
Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Setting the argument recover=TRUE or recover=FALSE should do the
sensible thing according to the documentation of the function, not
just always set it to TRUE regardless.
Fixes: fd.o#32551
Signed-off-by: Jonny Lamb <jonny.lamb@collabora.co.uk>
|
|
|
|
|
|
|
| |
For the channels hash in TpBaseClient the key is owned by the value in the
hash. The replacement behaviour of g_hash_table_insert always keeps the
existing key, which is not be valid anymore if the old value freed. Use
g_hash_table_replace instead to get the expected behaviour.
|
|
|
|
|
|
|
|
|
|
|
| |
TpProxy has the bad habit of emitting signals from dispose (Booo!).
TpBaseClient in various cases will be the last holder of a reference and tries
to remove the channel from its hash-table on invalidated..
So in case the proxy gets unreferenced because it is removed from the hash
table (because HandleChannels is called again with the same path), the
proxy is disposed, it signals invalidated (Boo!) and during the removal it gets
removed again and thus unreferenced again...
|
|\ |
|