Tristan Van Berkom [Tue, 19 Nov 2013 03:41:04 +0000 (12:41 +0900)]
test-client-custom-summary.c: Report specific error from add_contacts()
Daniel Mustieles [Mon, 25 Nov 2013 11:35:42 +0000 (12:35 +0100)]
Updated Spanish translation
Matthew Barnes [Sun, 24 Nov 2013 21:02:07 +0000 (16:02 -0500)]
CamelSession: Remove "check-junk" property.
Evolution uses it but Camel doesn't. Move it to Evolution.
Matthew Barnes [Sun, 24 Nov 2013 21:03:34 +0000 (16:03 -0500)]
Bump libcamel soname.
Some more API changes to follow.
Tristan Van Berkom [Fri, 22 Nov 2013 13:54:19 +0000 (22:54 +0900)]
ESexp: Check for NULL before freeing a GPtrArray
Tristan Van Berkom [Sun, 24 Nov 2013 07:10:54 +0000 (16:10 +0900)]
ESExp: Fixed memory leak of ESExp->error string
Tristan Van Berkom [Sun, 24 Nov 2013 07:11:33 +0000 (16:11 +0900)]
EDataBook: Fixed memory leak of every EContact succesfully added to the addressbook.
Piotr Drąg [Sat, 23 Nov 2013 17:13:59 +0000 (18:13 +0100)]
Updated POTFILES.in
Dimitris Spingos [Sat, 23 Nov 2013 16:57:25 +0000 (18:57 +0200)]
Updated Greek translation
Matthew Barnes [Sat, 23 Nov 2013 16:27:06 +0000 (11:27 -0500)]
ECalBackend: Use GThreadPool instead of g_io_scheduler_push_job().
Once the old-style backend API is gone we can port everything to GTask,
but for now be conservative with the changes.
Matthew Barnes [Sat, 23 Nov 2013 15:55:59 +0000 (10:55 -0500)]
EBookBackend: Use GThreadPool instead of g_io_scheduler_push_job().
Once the old-style backend API is gone we can port everything to GTask,
but for now be conservative with the changes.
Philip Withnall [Sat, 23 Nov 2013 13:41:59 +0000 (08:41 -0500)]
Bug 712323 - Substantially expand EVCard documentation
Philip Withnall [Sat, 23 Nov 2013 13:39:01 +0000 (08:39 -0500)]
e-vcard: Fix case sensitivity of string comparisions
The rest of the code (and the specification) uniformly users
case-insensitive comparisons of attribute names and parameter values.
Philip Withnall [Sat, 23 Nov 2013 13:37:48 +0000 (08:37 -0500)]
e-vcard: Fix a minor memory leak for multi-valued parameters
Matthew Barnes [Fri, 22 Nov 2013 22:13:05 +0000 (17:13 -0500)]
CamelFolder: Update "host-reachable" before connecting.
Matthew Barnes [Fri, 22 Nov 2013 22:08:45 +0000 (17:08 -0500)]
CamelStore: Update "host-reachable" before connecting.
Matthew Barnes [Fri, 22 Nov 2013 21:16:32 +0000 (16:16 -0500)]
Add camel_network_service_can_reach_sync().
Just wrappers g_network_monitor_can_reach() using the service's own
GSocketConnectable, but also updates the "host-reachable" property.
External callers can use this to synchronously update the property,
in case they need an up-to-date result to make a decision.
Also adds equivalent asynchronous functions:
camel_network_service_can_reach()
camel_network_service_can_reach_finish()
Colin Walters [Thu, 21 Nov 2013 19:49:36 +0000 (14:49 -0500)]
services: Demote g_print() -> g_debug() for bus names
This happens in normal operation, we don't need to announce it.
Developers wanting to trace things relating to DBus names can use
tools like gdbus monitor or d-feet.
https://bugzilla.gnome.org/show_bug.cgi?id=712841
Milan Crha [Fri, 22 Nov 2013 18:15:18 +0000 (19:15 +0100)]
Weather calendar: Be able to change temperature unit
The ESourceExtension had a 'unit' property, but it was not used
anywhere, thus I changed it to reflect possible values of current
libgweather and made it used in the background.
Matthew Barnes [Fri, 22 Nov 2013 17:46:25 +0000 (12:46 -0500)]
CamelNetworkService: Don't block first "host-reachable" notification.
Evolution had a similar problem recently, where an arbitrary choice of
default for a status property we don't yet know the real initial value
for could block the property's first "notify" signal after the initial
value is determined.
Daniel Mustieles [Fri, 22 Nov 2013 16:24:50 +0000 (17:24 +0100)]
Updated Spanish translation
Matthew Barnes [Thu, 21 Nov 2013 15:34:24 +0000 (10:34 -0500)]
UOA: Add support for IMAP access to Windows Live accounts.
Matthew Barnes [Fri, 22 Nov 2013 15:10:51 +0000 (10:10 -0500)]
UOA: Return IMAP/SMTP user names when collecting user info.
Can't assume the user names are the same as the collection identity.
Would be better to specify this in the .service files.
e.g. <imapuser>EMAILADDRESS / USERIDENTITY</imapuser>
But that can wait. Hard-coding by provider is ugly but works for now.
Matthew Barnes [Fri, 22 Nov 2013 14:38:26 +0000 (09:38 -0500)]
UOA: Miscellaneous cleanups.
Milan Crha [Fri, 22 Nov 2013 15:53:44 +0000 (16:53 +0100)]
ESourceRegistry: Ignore hidden files from .source folder monitoring
There used to be a critical warning on the source registry console:
** Failed to load key file at
'file:///home/user/.config/evolution/sources/.goutputstream-1P9V6W':
File must have a '.source' extension
There is nothing the source registry can do with these files (because
they are created by GIO), but ignore them during the folder monitoring,
instead of spreading useless runtime warning.
Piotr Drąg [Fri, 22 Nov 2013 15:09:37 +0000 (16:09 +0100)]
Updated POTFILES.in
Tristan Van Berkom [Wed, 20 Nov 2013 06:08:54 +0000 (15:08 +0900)]
Bug 712728 - Add --disable-migration option to evolution-source-registry
This allows one to run the EDS processes in relocated XDG user data
directories, without the registry stealing your user data from ~/.evolution
or previous versions.
Colin Walters [Thu, 21 Nov 2013 21:54:03 +0000 (16:54 -0500)]
Bug 712829 - Silence unnecessary migration chatter on stdout
Currently we log a message every time the migrater starts, even
if we have nothing to migrate. This clutters the systemd journal.
To fix this cleanly, rework this code so that error handling is
cleaner and more unified, which preserves the migration context.
Colin Walters [Thu, 21 Nov 2013 14:13:37 +0000 (09:13 -0500)]
registry-server: Demote g_print() to g_debug()
Avoid doing this sort of thing during normal startup.
Matthew Barnes [Thu, 21 Nov 2013 14:45:15 +0000 (09:45 -0500)]
UOA: Update google-calendar service to use OAuth-based CalDAV.
Colin Walters [Thu, 21 Nov 2013 13:59:56 +0000 (08:59 -0500)]
libebackend: Use g_debug() for normal operation message instead of g_print()
I'm trying to reduce log spew from simply booting. Now that dbus
connects bus-activated services to the journal, this g_print() is now
visible. Demote it to g_debug().
Matthew Barnes [Wed, 20 Nov 2013 13:13:56 +0000 (08:13 -0500)]
Bump minimum libgweather requirement.
Recent changes began using symbols introduced in libgweather 3.8 while
the minimum build requirement remained at 3.5. Update the requirement.
Shantha kumar [Wed, 20 Nov 2013 08:58:15 +0000 (14:28 +0530)]
Tamil Translations Updated
Milan Crha [Mon, 18 Nov 2013 20:16:47 +0000 (21:16 +0100)]
Make the Weather calendar work after changes for bug #712388
Matthew Barnes [Sun, 17 Nov 2013 16:42:18 +0000 (11:42 -0500)]
Bump AddressBook service name for new cursor API.
Matthew Barnes [Sun, 17 Nov 2013 16:40:50 +0000 (11:40 -0500)]
Update API documentation.
Matthew Barnes [Sun, 17 Nov 2013 16:17:00 +0000 (11:17 -0500)]
Remove undefined $(GNOME_INCLUDEDIR) / $(GTK_INCLUDEDIR).
Matthew Barnes [Sun, 17 Nov 2013 16:14:20 +0000 (11:14 -0500)]
Remove --enable-nntp configure option.
NNTP adds no extra dependencies; there's no need for it to be optional.
Matthew Barnes [Sun, 17 Nov 2013 15:53:53 +0000 (10:53 -0500)]
Makefile.am cleanups.
- Remove trailing tabs.
- Add terminators to multi-line lists.
Matthew Barnes [Sun, 17 Nov 2013 14:41:31 +0000 (09:41 -0500)]
Post-release version bump.
Matthew Barnes [Sun, 17 Nov 2013 14:03:27 +0000 (09:03 -0500)]
NEWS update for 3.11.2 release.
Matthew Barnes [Sun, 17 Nov 2013 14:21:11 +0000 (09:21 -0500)]
Fix distcheck error.
Daniel Mustieles [Sat, 16 Nov 2013 17:17:41 +0000 (18:17 +0100)]
Updated Spanish translation
Matthew Barnes [Sat, 16 Nov 2013 14:38:02 +0000 (09:38 -0500)]
CamelIMAPXStore: Chain up in the notify() method.
CamelObject provides a safety net for this method now.
Matthew Barnes [Sat, 16 Nov 2013 14:36:49 +0000 (09:36 -0500)]
CamelOfflineStore: Chain up in the notify() method.
CamelObject provides a safety net for this method now.
Matthew Barnes [Sat, 16 Nov 2013 14:34:13 +0000 (09:34 -0500)]
CamelObject: Implement the notify() method.
Just a placeholder so subclasses can safely chain up, since GObjectClass
does not implement its own notify() method.
Tristan Van Berkom [Sat, 16 Nov 2013 07:35:07 +0000 (16:35 +0900)]
EBookClient: Fix arrays passed to e_book_client_get_cursor() to be constant.
Minor API fixup, the arrays passed to e_book_client_get_cursor() should
be const.
Milan Crha [Fri, 15 Nov 2013 20:06:44 +0000 (21:06 +0100)]
Address some compiler warnings and Coverity scan issues in a different way
The previous Coverity scan commit used not so great way of dealing
with certain warnings, which also did not help with compiler warnings,
which this commit tries to change.
Piotr Drąg [Fri, 15 Nov 2013 19:02:46 +0000 (20:02 +0100)]
Updated POTFILES.in
Alexandre Franke [Thu, 14 Nov 2013 17:49:15 +0000 (18:49 +0100)]
Switch from g_timeout_add to g_timeout_add_seconds
https://bugzilla.gnome.org/show_bug.cgi?id=712244
Matthew Barnes [Thu, 14 Nov 2013 13:47:41 +0000 (08:47 -0500)]
Disable Nagle algorithm for IMAP connections.
I think I understand why this was done for the old CamelTcpStream:
the Nagle algorithm interferes with the pipelining of IMAP commands.
I do seem to notice some improved responsiveness after disabling this.
Tried to provide a clearer explanation in the code this time.
Tristan Van Berkom [Thu, 14 Nov 2013 08:40:59 +0000 (17:40 +0900)]
EBookBackendFile: Reduce uneeded warnings in server logs
With this patch, only unknown errors, which the user is not expected
to handle or recover from (such as internal SQLite errors or file system
errors), are warned about in the stderr.
Tristan Van Berkom [Wed, 13 Nov 2013 10:08:31 +0000 (19:08 +0900)]
test-client-custom-summary.c: Added exact email match test.
Tristan Van Berkom [Mon, 11 Nov 2013 04:37:28 +0000 (13:37 +0900)]
test-sqlite-cursor-set-sexp.c: Added test.
New test steps the cursor after setting a search expression.
Matthew Barnes [Thu, 14 Nov 2013 04:37:36 +0000 (23:37 -0500)]
CamelSession: Remove "network-available" property.
Too simplistic.
Use CamelNetworkService's "host-reachable" property instead.
Matthew Barnes [Thu, 14 Nov 2013 05:52:47 +0000 (00:52 -0500)]
CamelNetworkService: Disconnect as needed.
Disconnect immediately if the host is not reachable.
Matthew Barnes [Thu, 14 Nov 2013 04:30:37 +0000 (23:30 -0500)]
CamelOfflineStore: Prefer service's "host-reachable" status.
For some stores, the remote host machine may still not be reachable even
if the network is available. Common example is a host machine requiring
a virtual private network connection to reach.
If the CamelOfflineStore is also a CamelNetworkService, and the service
indicates the remote host is not reachable, then always return FALSE in
camel_offline_store_get_online().
Also emit "notify::online" along with "notify::host-reachable".
Matthew Barnes [Thu, 14 Nov 2013 04:40:56 +0000 (23:40 -0500)]
Remove "CamelDisco" classes from documentation TOC.
Matthew Barnes [Wed, 13 Nov 2013 23:09:19 +0000 (18:09 -0500)]
Remove CamelDisco{Diary,Folder,Store}.
These classes are no longer used.
Matthew Barnes [Wed, 13 Nov 2013 22:34:02 +0000 (17:34 -0500)]
NNTP: Move from 'disco' to 'offline' base classes.
NNTP was the last user of the old unmaintained 'CamelDisco' classes.
Matthew Barnes [Wed, 13 Nov 2013 21:16:09 +0000 (16:16 -0500)]
ECollectionBackend: Implement "proxy-resolver" property.
Defined the property but forgot to implement the "get" method.
Piotr Drąg [Wed, 13 Nov 2013 20:10:49 +0000 (21:10 +0100)]
Updated POTFILES.in
Tristan Van Berkom [Wed, 13 Nov 2013 12:09:34 +0000 (21:09 +0900)]
Migrated org.gnome.evolution.shell.network-config.gschema.xml.in to data/ directory.
The gsettings schemas need to all be stored and installed from the data/
directory, this is because we compile them locally in-tree and set the
GSETTINGS_SCHEMA_DIR to a local directory in order to run our tests
at make check time (tests dont work without this, instead GSettings wants
to abort for a missing schema).
Tristan Van Berkom [Wed, 13 Nov 2013 08:41:19 +0000 (17:41 +0900)]
Fixing cursor example build.
Enabled warnings in the Makefile.am and fixed them.
Milan Crha [Wed, 13 Nov 2013 08:40:46 +0000 (09:40 +0100)]
Fix/mute issues found by Coverity scan
This makes the code free of Coverity scan issues, except of the doc/
folder. It is sometimes quite pedantic and expects/suggests some
coding habits, thus certain changes may look weird, but for a good
thing, I hope. The code is also tagged with Coverity scan
suppressions, to keep the code as is and hide the warning too.
Also note that Coverity treats g_return_if_fail(), g_assert() and
similar macros as unreliable, and it's true these can be disabled
during the compile time, thus it brings in other set of 'weird'
changes.
Milan Crha [Tue, 12 Nov 2013 13:28:21 +0000 (14:28 +0100)]
Update libebook-contacts::test-untyped-phones test
The test_business has been disabled, because it was failing, but it
was failing even in time of 2.28. Looking briefly into RFC 2426
shows that the 'VOICE' is a default 'TEL' type, but there is no
indication that the left-alone 'WORK' is supposed to be the same
as 'WORK,VOICE' type. Due to this are done these changes.
Fabiano Fidêncio [Tue, 12 Nov 2013 12:28:20 +0000 (13:28 +0100)]
Fix 'may be used uninitialized' warnings found by cppcheck
Vadim Rutkovsky [Tue, 12 Nov 2013 12:27:44 +0000 (13:27 +0100)]
Suppress a false-positive cppcheck memleak warning
Make sure you run cppcheck with ''inline-suppr" to take it into effect.
Anish A [Tue, 12 Nov 2013 08:28:22 +0000 (13:58 +0530)]
Updated Malayalam Translation
Dimitris Spingos [Thu, 7 Nov 2013 08:10:34 +0000 (10:10 +0200)]
Updated Greek translation
Matthew Barnes [Wed, 6 Nov 2013 12:51:49 +0000 (07:51 -0500)]
imapx_untagged_fetch: Fix a possible divide-by-zero error.
Matthew Barnes [Tue, 5 Nov 2013 16:26:02 +0000 (11:26 -0500)]
CamelService: Make "password" property thread-safe.
New functions:
camel_service_dup_password()
Matthew Barnes [Tue, 5 Nov 2013 16:11:35 +0000 (11:11 -0500)]
CamelService: Make "display-name" property thread-safe.
New functions:
camel_service_dup_display_name()
Milan Crha [Tue, 5 Nov 2013 21:55:40 +0000 (22:55 +0100)]
Add missing CFLAGS/LIBS in the tests/ Makefile.am-s
If the libical is installed in a different prefix than the one shared
by other CFLAGS, then the test compilation breaks due to failed search
for the <libical/ical.h> include file.
Daniel Mustieles [Tue, 5 Nov 2013 17:23:25 +0000 (18:23 +0100)]
Updated Spanish translation
Matthew Barnes [Tue, 17 Sep 2013 16:11:40 +0000 (12:11 -0400)]
camel-maildir-summary.h: Remove unused enum.
Matthew Barnes [Sun, 3 Nov 2013 13:50:33 +0000 (08:50 -0500)]
CamelNetworkService: Return GIOStream instead of CamelStream.
Similar to recent CamelDataCache changes. Instead of wrapping a TCP
connection GIOStream in a CamelStream, return the GIOStream directly
and let callers wrap it if they need to with camel_stream_new().
Matthew Barnes [Sun, 3 Nov 2013 22:26:12 +0000 (17:26 -0500)]
CamelIMAPXFolder: Fix reference leak.
Matthew Barnes [Sun, 3 Nov 2013 18:02:23 +0000 (13:02 -0500)]
Remove unused camel_read/write_socket().
Matthew Barnes [Sun, 3 Nov 2013 15:28:32 +0000 (10:28 -0500)]
Add the other Camel services to API documentation.
For developer convenience, and to help me stay disciplined about writing
API documentation... even in provider modules.
Matthew Barnes [Sat, 2 Nov 2013 13:10:33 +0000 (09:10 -0400)]
CamelDataCache: Return GIOStream instead of CamelStream.
Another step toward dropping CamelStream. Instead of wrapping a local
file GIOStream in a CamelStream, return the GIOStream directly and let
callers wrap it if they need to with camel_stream_new().
Matthew Barnes [Fri, 1 Nov 2013 15:50:17 +0000 (11:50 -0400)]
Limit usage of legacy CamelObject class.
Only CamelFolder and CamelSession utilize what's left in CamelObject,
which is just binary state files to persistently store a flag or two.
All other Camel classes should inherit directly from GObject. This
requires a libcamel soname bump since the class structure sizes will
decrease by two pointers (no state_read() and state_write() methods).
Murray Cumming [Fri, 1 Nov 2013 14:32:23 +0000 (10:32 -0400)]
Bug 710989 - Replace most uses of strcpy()
Because strcpy() can overwrite its buffer. As with the previous
replacement of sprintf(), some of these could probably be replaced by
g_strdup(), but there could be some performance impact.
This also replaces a use of ICU's austrcpy() with austrncpy().
Fran Diéguez [Fri, 1 Nov 2013 12:07:43 +0000 (13:07 +0100)]
Updated Galician translations
Matthew Barnes [Fri, 1 Nov 2013 01:09:30 +0000 (21:09 -0400)]
Add --disable-examples to DISTCHECK_CONFIGURE_FLAGS.
Matthew Barnes [Fri, 1 Nov 2013 01:00:37 +0000 (21:00 -0400)]
Convert IMAPX back to a loadable module.
Since evolution-kolab is no longer active, convert the IMAPX provider
back to a loadable module so that necessary changes can be made without
affecting libcamel's public API.
The IMAPX chapter of Camel's API documentation will remain, however.
Matthew Barnes [Tue, 29 Oct 2013 19:07:19 +0000 (15:07 -0400)]
e-data-server-util.h: Fix a couple typos.
Matthew Barnes [Tue, 29 Oct 2013 16:30:04 +0000 (12:30 -0400)]
Remove duplicate "named timeout" docs.
Oopsie.
Bastien Nocera [Tue, 29 Oct 2013 16:18:32 +0000 (12:18 -0400)]
Bug 710808 - Name all the timeouts added with g_timeout_add()
Matthew Barnes [Tue, 29 Oct 2013 15:37:47 +0000 (11:37 -0400)]
Add macros/functions to help name timeouts.
The macros form a name for the timeout GSource implicitly.
New macros:
e_named_timeout_add()
e_named_timeout_add_full()
e_named_timeout_add_seconds()
e_named_timeout_add_seconds_full()
New functions (called by the macros):
e_timeout_add_with_name()
e_timeout_add_seconds_with_name()
Matthew Barnes [Mon, 28 Oct 2013 17:44:00 +0000 (13:44 -0400)]
Support migrating merged GConf trees.
Check for and parse a ~/.gconf/%gconf-tree.xml file during migration.
This is mainly for Debian, which I think accidentally merged some users'
GConf trees at some point.
This also adds a command-line program (evolution-scan-gconf-tree-xml),
which parses an arbitrary %gconf-tree.xml file as a recovery measure.
Matthew Barnes [Thu, 12 Sep 2013 22:23:30 +0000 (18:23 -0400)]
CamelIMAPXStatusResponse: Don't assume all values are present.
Unsolicited STATUS responses may not include all the values that we
normally request in a STATUS command, so just update the mailbox with
the values that are present in the response.
Matthew Barnes [Thu, 12 Sep 2013 15:58:39 +0000 (11:58 -0400)]
CamelIMAPXStoreSummary cleanups.
CamelIMAPXStoreSummary has no public or private members, so no reason
to cast pointers to that type. Stick with plain (CamelStoreSummary *).
Removed functions:
camel_imapx_store_summary_new()
Matthew Barnes [Thu, 12 Sep 2013 13:32:04 +0000 (09:32 -0400)]
Hack around default Courier namespaces.
By default, Courier lists (("INBOX." ".")) for its personal namespaces,
which excludes INBOX itself. This causes problems for us so check for
it specifically and hack around it where necessary.
Matthew Barnes [Wed, 11 Sep 2013 21:19:05 +0000 (17:19 -0400)]
CamelIMAPXStore: Remove public 'dir_sep' member.
No longer used.
Matthew Barnes [Wed, 11 Sep 2013 21:15:50 +0000 (17:15 -0400)]
Remove CamelIMAPXStoreNamespace.
No longer used.
Matthew Barnes [Wed, 11 Sep 2013 21:13:52 +0000 (17:13 -0400)]
Remove CamelIMAPXNamespaceList.
No longer used.
Matthew Barnes [Wed, 11 Sep 2013 21:10:02 +0000 (17:10 -0400)]
Remove camel_imapx_namespace_list_clear().
No longer used.
Matthew Barnes [Wed, 11 Sep 2013 21:04:44 +0000 (17:04 -0400)]
CamelIMAPXStoreSummary: Remove all public members.
The 'version' member was not used for anything and we no longer use the
'namespaces' member. Only the CamelIMAPXServer knows about server-side
mailbox namespaces. That information is not relevant when offline, and
therefore doesn't need to be cached.
This also removes imapx_fabricate_old_namespace_list() and similar hacks
in CamelIMAPXServer.
Matthew Barnes [Wed, 11 Sep 2013 20:31:04 +0000 (16:31 -0400)]
Remove camel_imapx_store_summary_namespace_find_by_mailbox().
No longer used.