summaryrefslogtreecommitdiff
Commit message (Collapse)AuthorAgeFilesLines
* Update NEWS for 1.29.3 release1.29.3Ondrej Holy2016-06-201-0/+8
|
* smb: Set always G_FILE_ATTRIBUTE_ACCESS_CAN_WRITE for filesOndrej Holy2016-06-141-2/+2
| | | | | | | | | | G_FILE_ATTRIBUTE_ACCESS_CAN_WRITE is set for regular files only when S_IWUSR is not set currently (i.e. FALSE only). It seems to me that it is reliable to set this to TRUE for regular files in other cases. Applications should not rely on those attributes, however it is crucial for some apps which rely on them still (e.g. File Roller). https://bugzilla.gnome.org/show_bug.cgi?id=724497
* Updated Brazilian Portuguese translationRafael Fontenelle2016-06-141-231/+257
|
* Updated Turkish translationMuhammet Kara2016-06-141-205/+230
|
* Updated Spanish translationDaniel Mustieles2016-06-131-239/+266
|
* afc: Fix several memory leaksOndrej Holy2016-06-101-0/+9
| | | | | Memory is not released on several places, when handling errors mostly. Add missing g_free calls to avoid those memory leaks.
* afc: Add missing trailing newline charOndrej Holy2016-06-101-1/+1
| | | | GVfs uses custom log handler, which requires trailing newline char.
* afc: Fix error when renaming files in DocumentsBastien Nocera2016-06-091-1/+7
| | | | | | | When renaming a file in Documents, the callers expect the new path in the global scheme, not the path inside the HouseArrest AFC itself. This fixes a warning in nautilus when renaming files.
* afc: Tweak HouseArrest warning outputBastien Nocera2016-06-091-5/+10
| | | | | | Only throw a warning when the second attempt at setting up HouseArrest fails, not when we're going to retry. This makes it easier to see real warnings compared to normal run-time behaviour.
* afc: Fix GStreamer playback of native filesBastien Nocera2016-06-091-10/+17
| | | | | | | | | | | | | | GStreamer gets very confused when the offset after a seek isn't what it expected, and told us that the "Stream contains no data." when trying to play back using giosrc. But we could play the video just fine going through fuse. FUSE hides bugs with relative seeks, such as our absolute offset being wrong when doing absolute seeks. To make sure that the offset we return is correct, call tell() and use that as the new offset. https://bugzilla.gnome.org/show_bug.cgi?id=767405
* afc: Fix mount failing after trusting deviceBastien Nocera2016-06-091-0/+8
| | | | | | | | After having trusted the device, if we are too quick at clicking the "Try Again" button, the creation will fail with an unknown error. Try again silently up to the maximum number of retries after a short sleep. https://bugzilla.gnome.org/show_bug.cgi?id=767399
* afc: Better lockdown client creation debugBastien Nocera2016-06-091-7/+13
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=767399
* afc: Fix volume init with a locked deviceBastien Nocera2016-06-091-22/+10
| | | | | | | | | | | | | | | | | The first time we connect the device, we won't be able to create a lockdown client with a handshake, so the volume won't be created at all and will not show up in nautilus. Furthermore, we were checking the OS version with a paired lockdown client (which is not needed, an unpaired one will do just as well) and checking whether HouseArrest was supported even if the service was different. So, we never needed to have a paired lockdown client in the first place, and we only need to check for the OS version if HouseArrest is what was requested. https://bugzilla.gnome.org/show_bug.cgi?id=767399
* afc: Add more debug when initialisation failsBastien Nocera2016-06-091-1/+4
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=767399
* afc: Add specific errors for trust dialogue errorsBastien Nocera2016-06-091-1/+9
| | | | | | | | Add specific errors for when the user refuses trust on the device, or repeatedly tries to connect to it but the device is locked, or the dialogue still showing on the device. https://bugzilla.gnome.org/show_bug.cgi?id=763605
* udisks2: Fix "format not a string literal" warningsBastien Nocera2016-06-091-9/+8
| | | | https://bugzilla.gnome.org/show_bug.cgi?id=767349
* afc: Add pairing dialogue specific messageBastien Nocera2016-06-091-7/+20
| | | | | | | When a pairing dialogue is shown on the device, don't tell the user that the device is locked, rather that they need to click "Trust". https://bugzilla.gnome.org/show_bug.cgi?id=763605
* afc: Fix iOS 7 trust dialogue handlingBastien Nocera2016-06-091-1/+8
| | | | | | | There are new iOS 7 related error codes that we need to handle to catch cases where a user action is required on the device. https://bugzilla.gnome.org/show_bug.cgi?id=767297
* afc: Limit the number of HouseArrest services we startBastien Nocera2016-06-091-21/+131
| | | | | | | | | | | | | To avoid being unable to navigate within applications because we ran out of HouseArrest services, keep track of file operations within each application, and disconnect from unused ones when trying to access an unconnected one. The tracking only covers file operations, and not enumerations, and the garbage collection algorithm is quite naive, but could easily be extended. https://bugzilla.gnome.org/show_bug.cgi?id=676188
* afc: Indicate whether to retry to setup HouseArrest for an appBastien Nocera2016-06-091-1/+5
| | | | | | | | If setting up the HouseArrest service for an app fails, we should clean up unused services, and try again. This tells us whether we should try again in the first place. https://bugzilla.gnome.org/show_bug.cgi?id=676188
* afc: Lock apps mutex when setting up HouseArrestBastien Nocera2016-06-081-8/+12
| | | | | | So that we don't have cases where the AppInfo disappears from under us. https://bugzilla.gnome.org/show_bug.cgi?id=676188
* afc: Better debug error message againBastien Nocera2016-06-081-1/+1
| | | | | | For starting the HouseArrest lockdownd service this time. https://bugzilla.gnome.org/show_bug.cgi?id=676188
* afc: Add guards against the operation mode not being setBastien Nocera2016-06-081-3/+15
| | | | | | | | | | | We usually did: if mode == AFC; <foo> else <bar> To guard against the mode not actually being HouseArrest in the else branch, change this to: if mode == AFC; <foo> else if mode == HOUSE_ARREST ; <bar> ; else ; <baz> https://bugzilla.gnome.org/show_bug.cgi?id=764099
* afc: Fix handling of new HouseArrest pathsBastien Nocera2016-06-081-36/+34
| | | | | | | | | | | | | | In b8ad223, we re-rooted the HouseArrest paths to always include Documents/ as anything closed to the root isn't accessible. This caused 2 problems: - first, it meant that checking for the new path being "/" wasn't correct, as the path would always include at least a "Documents" path component - secondly, it meant that the display name for the root directory for an application was incorrect as it was always "Documents" instead of the app name. https://bugzilla.gnome.org/show_bug.cgi?id=764099
* build: Fix all statfs() tests failingBastien Nocera2016-06-071-2/+2
| | | | | | | | | | The current statfs() compilation tests all fail because statfs() expects the first argument to be non-null. Pass a dummy path instead of NULL to satisfy the compiler. The patch is taken from GLib, see Bug 764574. https://bugzilla.gnome.org/show_bug.cgi?id=767331
* test: Be sure that gvfs is used by gioOndrej Holy2016-06-071-0/+2
| | | | | | | | | | | | Currently the tests may fail with: Error mounting location: volume doesn't implement mount This may happen if GIO_USE_VFS=local, which is currently set by gnome-desktop-testing as defaults. Set GIO_USE_VFS=gvfs for our process to fix it. https://bugzilla.gnome.org/show_bug.cgi?id=734370
* test: Skip sftp tests if ssh key is not availableOndrej Holy2016-06-031-3/+1
| | | | | | | | | | | Currently sftp tests fail if ~/.ssh/id_rsa is not available with the following error: AssertionError: False is not true : This test needs an existing ~/.ssh/id_rsa Sftp tests should be skipped if the key is not found to avoid test suite failures. https://bugzilla.gnome.org/show_bug.cgi?id=734370
* test: Do not fail if umockdev is not installedOndrej Holy2016-06-021-1/+1
| | | | | | | Change introduced by commit 6d77394 results in ValueError exception if umockdev is not installed. This exception is not handled and thus the test is unexpectedly terminated. Handle this exception and set have_umockdev accordingly.
* build: Fix srcdir != builddirOndrej Holy2016-05-311-6/+6
| | | | | | | | | | | Commit 0188305 broke build, because doesn't expect srcdir != builddir and consequently the build fails with the following error: make[3]: *** No rule to make target '../../daemon/org.gtk.vfs.Daemon.service', needed by 'all-am'. Stop. make[3]: *** Waiting for unfinished jobs.... Use builddir instead of srcdir to fix the problem. https://bugzilla.gnome.org/show_bug.cgi?id=734370
* Updated Occitan translationCédric Valmary2016-05-301-48/+48
|
* test: Use system paths in installed testsOndrej Holy2016-05-301-3/+8
| | | | | | | | Installed tests are still using buildroot paths, which causes test failures if buildroot is not available. System paths has to be used instead for the installed tests. https://bugzilla.gnome.org/show_bug.cgi?id=734370
* test: Skip smb tests if smbd is not installedOndrej Holy2016-05-301-0/+2
| | | | | | | | | | | Currently smb tests fail if smbd is not installed with the following error: FileNotFoundError: [Errno 2] No such file or directory: 'smbd' Smb tests should be skipped if smbd is not installed to avoid test suite failures. https://bugzilla.gnome.org/show_bug.cgi?id=734370
* test: Skip ftp tests if twistd is not installedOndrej Holy2016-05-301-0/+2
| | | | | | | | | | | Currently ftp tests fail if twistd is not installed with the following error: FileNotFoundError: [Errno 2] No such file or directory: 'twistd' Sftp tests should be skipped if twistd is not installed to avoid test suite failures. https://bugzilla.gnome.org/show_bug.cgi?id=734370
* fuse: Add GVFS_DEBUG_FUSE to enable debug outputOndrej Holy2016-05-252-87/+75
| | | | | | | | | | It was not possible to enable debug output from the fuse daemon without rebuilding. This patch replaces custom debug handling in the similar way as it is done in other daemons (using GLib logging functionality). Environment variable GVFS_DEBUG_FUSE can be specified in order to enable debugging and SIGUSR2 signal can be sent also to toggle debug output. The debug output will be printed at stdout, so the debug output can be seen on gvfsd output thanks to the commit 85bacbf.
* fuse: Do not suppress gvfsd-fuse outputOndrej Holy2016-05-241-2/+1
| | | | | | Fuse daemon is spawned with G_SPAWN_STDOUT_TO_DEV_NULL and G_SPAWN_STDERR_TO_DEV_NULL, which suppress its output. Use G_SPAWN_LEAVE_DESCRIPTORS_OPEN instead to simplify debugging.
* tests: Suppress unnecessary output from find_alternativeOndrej Holy2016-05-241-1/+1
| | | | | This suppress the following output from find_alternative: which: no apache2 in ($PATH)
* tests: Explicitly require version for GUMockdev to avoid warningOndrej Holy2016-05-241-0/+2
| | | | | This avoids the following deprecation warning: PyGIWarning: UMockdev was imported without specifying a version first.
* Post release version bumpOndrej Holy2016-05-231-1/+1
|
* Update NEWS for 1.29.2 release1.29.2Ondrej Holy2016-05-231-0/+8
|
* Updated Turkish translationMuhammet Kara2016-05-201-359/+370
| | | | (cherry picked from commit aad085b27e6abfb19ecaa4439cb88fd99f45f327)
* Always add mount prefix in cached fuse pathsJens Georg2016-05-201-34/+29
| | | | | | | | | | | | | If a fuse path is encountered the first time, the fuse path is re-created taking the mount prefix into account. When the file is created for the path the second time from the cached mount info, the mount prefix is silently ignored, causing the new file not to be equal to the old one (it is even invalid): This change just looks up the mount info in the cache and then continues constructing the path as it would without having the info cached. https://bugzilla.gnome.org/show_bug.cgi?id=766294
* monitor: Add g_drive_is_removable() supportOndrej Holy2016-05-205-4/+45
| | | | | | | | | | | | | | | Nautilus wants to show entries in the sidebar only for removable devices. It uses currently sort of conditions to determine which devices should be shown. Those condition fails in some cases unfortunatelly. Lets provide g_drive_is_removable() which uses udisks Removable property to determine which devices should be shown. It should return true for all drives with removable media, or flash media, or drives on usb and firewire buses. Add support for this property also in gvfs-mount tool. Bump GLib version accordingly. https://bugzilla.gnome.org/show_bug.cgi?id=765457
* dav: Fix IPv6 address handlingOndrej Holy2016-05-201-2/+15
| | | | | | | | | It is not possible to mount WebDAV share with IPv6 address currently. GMountSpec host contains brackets around the IPv6 address, however SoupURI doesn't. This patch handles this when converting GMountSpec to SoupURI and vice versa. https://bugzilla.gnome.org/show_bug.cgi?id=766631
* Updated Occitan translationCédric Valmary2016-05-091-207/+197
|
* gphoto2: Use G_FILESYSTEM_PREVIEW_TYPE_NEVERPierre Ossman2016-05-041-2/+2
| | | | | | | | | | | PTP/MTP devices can only transfer complete files, which means that generating a thumbnail can often take multiple minutes, and any other operation is blocked until this is done. In practice this can often prevent meaningful use of the device. So change the hint to applications, advising them it is a bad idea to generate thumbnails for these devices. https://bugzilla.gnome.org/show_bug.cgi?id=765843
* afc: Do not use G_FILESYSTEM_PREVIEW_TYPE_IF_LOCALOndrej Holy2016-05-041-3/+0
| | | | | | | | Commit fe83e09 removed thumbnailing support, so we should not using G_FILESYSTEM_PREVIEW_TYPE_IF_LOCAL anymore. Let's remove it to be treated as other remote filesystems. https://bugzilla.gnome.org/show_bug.cgi?id=749639
* udisks2: Fix wrong types to avoid warningsOndrej Holy2016-04-292-2/+2
| | | | | | | Commit beea21e introduced incompatible-pointer-types warnings, because incompatible types was used. Fix the wrong types to avoid such warnings. https://bugzilla.gnome.org/show_bug.cgi?id=763890
* Fix misleading if-statementOndrej Holy2016-04-291-2/+4
| | | | | | | GCC 6 introduced -Wmisleading-warnings. It reveals this misleading if-statement. "@" should not really be added if username is not specified, however I am convinced that this function is never called without username...
* localtest: Fix indentation to avoid warningsOndrej Holy2016-04-291-4/+4
| | | | | GCC 6 introduced -Wmisleading-warnings. Fix wrong indentation to avoid misleading-indentation warning.
* Updated Portuguese translationGNOME Translation Robot2016-04-281-285/+288
|
View Lorry Controller Status