Manuel Nickschas [Mon, 9 Mar 2015 20:53:00 +0000 (21:53 +0100)]
Add missing dependency to KWidgetsAddons
This is pulled in by some other frameworks modules anyway, but we should
list it explicitly.
Michael Marley [Fri, 27 Feb 2015 17:31:17 +0000 (12:31 -0500)]
Add back the SSL protocol selection dialog box for old cores.
Cores before 0.10 will default to SSLv3 if the user doesn't make a
selection. If a >=0.10 client is used with a <0.10 core to connect
to a server that has SSLv3 disabled, it is impossible to connect
to that server without upgrading the core or using an old client
to change the SSL protocol settings.
This also changes the SSLv2 and SSLv3 options to indicate their
insecurity and therefore discourage their use.
Cores from 0.10 and up use SSL autonegotiation and to not need the
protocol setting.
This partially reverts commit
e53fc69a91553b57932ba599b39999d550114588.
Michael Marley [Sat, 14 Feb 2015 01:12:18 +0000 (20:12 -0500)]
Don't require SSL for loopback connections with --require-ssl
Encrypting a loopback connection is a pointless waste of computing
power. This patch will allow people to use loopback connections
from things like quassel-webserver without encryption while still
requiring encryption for remote connections.
Michael Marley [Tue, 3 Mar 2015 01:31:02 +0000 (20:31 -0500)]
Add support for SHA2-512 hash on Qt4 builds
Qt4 doesn't support anything higher than SHA1 in
QCryptographicHash, so it is necessary to include a fallback. This
bundles sha512.c from PolarSSL/mbed TLS. It is only used for Qt4
builds and does not require a dependency on external libraries that
might be hard to compile on Windows® or Mac OSX®. It can be
removed once Qt4 is no longer supported.
Michael Marley [Sat, 21 Feb 2015 12:33:57 +0000 (07:33 -0500)]
Improve the message-splitting algorithm for PRIVMSG and CTCP
This introduces a new message splitting algorithm based on
QTextBoundaryFinder. It works by first starting with the entire
message to be sent, encoding it, and checking to see if it is over
the maximum message length. If it is, it uses QTBF to find the
word boundary most immediately preceding the maximum length. If no
suitable boundary can be found, it falls back to searching for
grapheme boundaries. It repeats this process until the entire
message has been sent.
Unlike what it replaces, the new splitting code is not recursive
and cannot cause stack overflows. Additionally, if it is unable
to split a string, it will give up gracefully and not crash the
core or cause a thread to run away.
This patch fixes two bugs. The first is garbage characters caused
by accidentally splitting the string in the middle of a multibyte
character. Since the new code splits at a character level instead
of a byte level, this will no longer be an issue. The second is
the core crash caused by sending an overlength CTCP query ("/me")
containing only multibyte characters. This bug was caused by the
old CTCP splitter using the byte index from lastParamOverrun() as
a character index for a QString.
Manuel Nickschas [Wed, 18 Feb 2015 21:09:54 +0000 (22:09 +0100)]
Bump version for release
Manuel Nickschas [Wed, 18 Feb 2015 21:07:59 +0000 (22:07 +0100)]
Update INSTALL
Manuel Nickschas [Wed, 18 Feb 2015 20:54:50 +0000 (21:54 +0100)]
Update ChangeLog
Manuel Nickschas [Wed, 18 Feb 2015 20:17:51 +0000 (21:17 +0100)]
Update AboutData
Kimmo Huoman [Mon, 2 Feb 2015 12:13:51 +0000 (14:13 +0200)]
Allows merging buffers over networks
Feature request at http://bugs.quassel-irc.org/issues/1334
> As stated in Issue 83 merge is only possible with chats on the
> same network.
> As the usage of Bitlbee etc has increased, it would be useful
> to be able to merge chats across all networks as well. The issue
> rises as people are now using alternative methods (Facebook, Lync
> etc) during the day and returning to IRC in the evening.
Closes #1334
Manuel Nickschas [Wed, 18 Feb 2015 20:03:51 +0000 (21:03 +0100)]
Merge pull request #107 from siduction-upstream/master
Set GIT_HEAD and GIT_DESCRIBE from environment
Chris Holland [Sat, 1 Feb 2014 20:54:40 +0000 (15:54 -0500)]
Toggle hiding of Join/Part/Quits with a single button
Alf Gaida [Tue, 17 Feb 2015 22:58:32 +0000 (23:58 +0100)]
Set GIT_HEAD and GIT_DESCRIBE via environmental variables
Set GIT_HEAD and GIT_DESCRIBE before run cmake in case no source repo is
available - booth values must be extracted at clone/pull time
Example:
git log -1 | grep commit | sed 's#commit ##' > debian/git_head; \
git describe > debian/git_describe
export GIT_HEAD=$(shell [ -f debian/git_head ] && cat debian/git_head)
Pierre-Hugues Husson [Mon, 26 May 2014 11:38:08 +0000 (13:38 +0200)]
Add "/print" command
This only displays an info message without sending anything to
the server.
Manuel Nickschas [Tue, 17 Feb 2015 21:16:23 +0000 (22:16 +0100)]
Change location and behavior of the Password Change action
This action fits better in the File menu than under Settings, so
move it there.
Also, to increase discoverability, we now show the option even if
the core is too old to support the feature. A popup message informing
the user is shown in that case.
Additionally, the changed password is now saved client-side (if the
account is setup to remember the password), thus avoiding having to
re-enter the new password on next launch.
Michael Marley [Sat, 14 Feb 2015 19:01:42 +0000 (14:01 -0500)]
Hide or disable the Change Password option when appropriate
Previously, the Change Password option would be shown all the
time, even the the client was connected to a core not supporting
the password change functionality or if a monolithic build was in
use. Now, the Change Password option will be disabled if the
client is disconnected or is connected to a core not supporting the
functionality and will be hidden if a monolithic build is in use.
Martin T. H. Sandsmark [Tue, 25 Nov 2014 00:43:18 +0000 (01:43 +0100)]
Implement password changing from client.
Add menu item and dialog for getting a new password from the user,
signal from client to core to give the new password, and wire up the
signal on the core to change the password in the database.
Also update the stored password on the client side.
Manuel Nickschas [Mon, 16 Feb 2015 22:05:55 +0000 (23:05 +0100)]
Migrate KDE-specific settings from KDE4 to KF5
This uses Kdelibs4ConfigMigrator to move over KDE-specific settings
(i.e. window state, shortcuts, notifications...) from the KDE4 config
dir to the one used by KF5. This should make the move seamless for
most intents and purposes.
Note that migration won't happen (again) once the KF5 config files
exist.
Manuel Nickschas [Mon, 16 Feb 2015 21:52:57 +0000 (22:52 +0100)]
Port remaining stuff in MainWin to KF5
We use KToolbar now and handle the full screen action properly in
both KDE4 and KF5.
Manuel Nickschas [Mon, 16 Feb 2015 20:56:48 +0000 (21:56 +0100)]
Fix shortcut saving on KDE, and use the system dialog for Frameworks
While re-enabling the system dialog for shortcut editing for KF5,
I finally figured out why shortcuts where never saved when using KDE;
a long standing bug we've had.
Turns out QDialog::exec() does not actually save things. All this time
I was looking into completely different directions; but it never
occured to me that I need to explicitly save after the dialog is
accepted (our own dialog does this automatically). This is not a
complaint; it was being stupid on my end :)
Fixes #1057.
Manuel Nickschas [Sun, 15 Feb 2015 21:48:09 +0000 (22:48 +0100)]
Use KNotifications if Frameworks integration is enabled
This brings back proper support for KNotify if KDE Frameworks
integration is enabled.
Note that quassel.notifyrc is installed in a different location
if KF5 as opposed to KDE4. If you don't have a system installation
of Quassel, you may have to manually copy that file from data/
to $HOME/.local/share/knofitications5 for KNotify to know about
Quassel.
Manuel Nickschas [Thu, 12 Feb 2015 21:24:34 +0000 (22:24 +0100)]
Adapt custom stylesheet path to new default install location
When switching from a KDE4 build to Qt5/KF5, the default install
location for data files switches from /usr/share/apps/quassel to
/usr/share/quassel, as is standard. Users who have selected one of
the pre-installed stylesheets will use that setting.
Thus, if a custom stylesheet is configured and points to /usr/share/apps,
in case it is not found we will now try and look into /usr/share as well.
If we find the stylesheet in the new location, we'll also update the
stored setting.
Manuel Nickschas [Wed, 11 Feb 2015 22:22:31 +0000 (23:22 +0100)]
Work around weirdly broken MOCs
Apparently, Debian Wheezy ships Qt 4.8.2 with a MOC that thinks it's older
than 4.6, or something. Not sure what's going on, but if we remove
that "#if QT_VERSION" that checks for versions older than 4.6 (and thus does
not make any sense now that we require 4.8 anyway), the MOC on (some?)
Wheezy installations fails to deal with the Q_INTERFACES call.
I guess it doesn't hurt to keep the guard around, but this issue still leaves
me scratching my head...
Manuel Nickschas [Sun, 8 Feb 2015 23:58:49 +0000 (00:58 +0100)]
Use KMainWindow if KF5 integration is enabled.
With this commit, Quassel uses a proper KMainWindow if KF5 integration
is enabled. This works pretty much like the KDE4 version, so we just had
to add the appropriate #includes and adapt the #ifdefs.
Manuel Nickschas [Sun, 8 Feb 2015 23:58:16 +0000 (00:58 +0100)]
Move KAboutData initialization before app initialization
Manuel Nickschas [Sun, 8 Feb 2015 23:55:41 +0000 (00:55 +0100)]
Add comment regarding the removal of X11 hacks to GraphicalUi
Turns out all window activation stuff just works with Qt5, even with
KF5 integration enabled. So there is no need to employ fugly X11 hacks
in order to bypass focus stealing prevention, as was required in KDE4.
Because GraphicalUi doesn't seem to require any more porting changes,
add a comment instead to later remove the hacks together with Qt4/KDE4.
Manuel Nickschas [Sun, 8 Feb 2015 23:54:07 +0000 (00:54 +0100)]
Require CMake 2.8.12 for KDE Frameworks integration
CMake support for KF5 requires at least 2.8.12, so enforce this.
Accordingly, move cmake_minimum_required() and friends below reading
the configure option.
Manuel Nickschas [Sun, 8 Feb 2015 23:50:53 +0000 (00:50 +0100)]
Make the AboutDlg layout slightly prettier
Some italic font and some more whitespace make things so much nicer.
Manuel Nickschas [Sun, 8 Feb 2015 18:52:55 +0000 (19:52 +0100)]
Improve data path handling
Originally I intended to switch to QStandardPaths for Qt5, however this
would also require migration code. We'll postpone this until we can
get rid of Qt4 and KDE4, so there are less combinations to handle properly.
Instead, this commit improves the handling of data paths in general by
fixing search order, supporting XDG_DATA_HOME, and sanitizing the paths
properly. Also KDE builds now check their own install prefix, in case
Quassel is installed in a non-standard location.
Manuel Nickschas [Sun, 8 Feb 2015 17:07:30 +0000 (18:07 +0100)]
Add missing index.theme to the bundled Oxygen icon theme
Otherwise, the icon loader may not find the bundled icons.
Michael Marley [Sun, 8 Feb 2015 18:44:33 +0000 (13:44 -0500)]
Convert the Storage::HashVersion names to CamelCase
Michael Marley [Sun, 8 Feb 2015 18:25:50 +0000 (13:25 -0500)]
Add support for a salted SHA2-512 hash
This new hash uses a SHA2-512 hash salted with 512 bits of data
generated using the C++11 std::mt19937 algorithm, seeded by
std::random_device. It only works with Qt5, so Qt4 builds will
keep using the old SHA1 hash.
Michael Marley [Sat, 7 Feb 2015 20:40:23 +0000 (15:40 -0500)]
Add support for password hash versioning
Previously, Quassel would always use an unsalted SHA1 hash. This
patch introduces a version field for the password, allowing for
using a better hash. (No new hash is added yet.) Users will be
automatically upgraded to the newest hash whenever they login or
change their passwords.
This is implemented by adding a new integer field to hold the hash
version in the DB. Additionally, the password field in PostgreSQL
has been expanded to a TEXT field, allowing larger hash values to
be stored in the future. The password hashing code has also been
refactored to make it easier to introduce new hashes without
fooling around in the DB driver code.
Manuel Nickschas [Sat, 7 Feb 2015 16:57:26 +0000 (17:57 +0100)]
Merge pull request #97 from Bombe/focus-host-input
Set initial focus to host input widget.
Manuel Nickschas [Sat, 7 Feb 2015 16:45:24 +0000 (17:45 +0100)]
Merge pull request #104 from mamarley/disconnectdesyncfix
Fix potential state desync in disconnectFromIrc()
Manuel Nickschas [Sat, 7 Feb 2015 16:34:02 +0000 (17:34 +0100)]
Remove dead assignments
This makes clang-analyzer happy.
Manuel Nickschas [Sat, 7 Feb 2015 16:33:32 +0000 (17:33 +0100)]
Remove unneeded conditionals
This makes clang-analyzer happy.
Manuel Nickschas [Sat, 7 Feb 2015 16:25:43 +0000 (17:25 +0100)]
Simplify session thread handling in Core
Whenever we want to get a session thread for a given user, we first
would check if we already have one and get that, or create a new one.
Instead of doing this several times in the code, this commit just
turns the former createSession() method into sessionForUser(), which
will create a thread if needed. This also removes a dead assignment
clang-analyzer complained about.
Manuel Nickschas [Sat, 7 Feb 2015 16:05:58 +0000 (17:05 +0100)]
Replace assertion by condition
While the condition should be always true, clang-analyzer still
complained about it. So turn this into a condition instead.
Worst case we'll just lose the item selection now.
Manuel Nickschas [Fri, 6 Feb 2015 21:30:44 +0000 (22:30 +0100)]
Remove support for obsolete Qt versions
Now that we rely on Qt 4.8+, we can get rid of a few workaround that
were in place for older versions.
Manuel Nickschas [Fri, 6 Feb 2015 21:22:21 +0000 (22:22 +0100)]
Remove support for network detection via Solid
Now that we rely on Qt 4.8 and up, we can use QNCM on all platforms.
Solid::Networking didn't do anything else, and is deprecated in
KDE Frameworks in favor of QNCM anyway.
Also improved the core connection settings dialog a bit.
Manuel Nickschas [Fri, 6 Feb 2015 18:59:58 +0000 (19:59 +0100)]
Filter -ansi from CMAKE_CXX_FLAGS
This was set by older versions of KDE 4 and negates -std=c++11.
Manuel Nickschas [Thu, 5 Feb 2015 23:02:36 +0000 (00:02 +0100)]
Add Yaohan Chen to AboutData
Manuel Nickschas [Thu, 5 Feb 2015 22:46:54 +0000 (23:46 +0100)]
Merge pull request #108 from hagabaka/master
Add QNetworkConfigurationManager-based connection detection
Manuel Nickschas [Thu, 5 Feb 2015 22:26:52 +0000 (23:26 +0100)]
Disable Clang for building against Qt4 in the CI
Qt4 does not support C++11 features when building with Clang, so
the Travis CI fails. This commit should disable this combination
in the build matrix.
Manuel Nickschas [Thu, 5 Feb 2015 20:47:59 +0000 (21:47 +0100)]
Merge pull request #102 from mamarley/qcaqt5
Add support for compiling with QCA on Qt5
Yaohan Chen [Sat, 24 Jan 2015 23:50:12 +0000 (18:50 -0500)]
Add Qt-based connection detection
A new network detection mode is added to use QNetworkConfigurationManager,
available in Qt4.7+ including Qt5. This becomes the default mode when KDE4
integration is not available.
Manuel Nickschas [Thu, 5 Feb 2015 00:08:46 +0000 (01:08 +0100)]
Merge pull request #105 from merlin1991/qdatastream
Set the version of the QDataStream in ClientAuthHandler
Manuel Nickschas [Wed, 4 Feb 2015 23:20:06 +0000 (00:20 +0100)]
Simplify the handling of the contributor list in AboutDlg
This introduces an AboutData class akin to KAboutData from KDE
Frameworks. Together with brace initialization, which we can now
use thanks to C++11, this makes specifying the (huge!) list of
contributors to Quassel much nicer, compared to editing a long
HTML string.
AboutData also can create a KAboutData object that can be used for
KF5 integration, and later for showing the native KAboutDialog
instead of our own if KDE integration is enabled.
The list of contributors has also been synced with the Git history
again. If you find your name missing, or want changes, please
let us know.
Manuel Nickschas [Wed, 4 Feb 2015 23:15:26 +0000 (00:15 +0100)]
Make UTF-8 the default encoding for QString and tr()
As all our source files are in UTF-8 (and Qt5 even requires them to
be), we can just tell Qt4 to treat them as that. So no need anymore
to use QString::fromUtf8() for literals.
This also makes it possible to use string literals in brace
initializers for QString.
Manuel Nickschas [Wed, 4 Feb 2015 23:14:05 +0000 (00:14 +0100)]
Add D-Bus domain to .desktop files
This is recommended for KDE software.
Manuel Nickschas [Mon, 2 Feb 2015 22:21:40 +0000 (23:21 +0100)]
Bump Qt4 version requirement to 4.8
Qt 4.8 gained some support for C++11 features such as brace
initializers, which make some things much more convenient. As
this version was released more than three years ago, it should
be safe to finally bump our requirements accordingly. The core
was even fine with Qt 4.4 until now, and that was released a
whopping 7 years ago!
Requiring 4.8 will allow us to remove a bunch of old workarounds
as well, besides allowing for proper use of C++11 in the future.
Manuel Nickschas [Sat, 31 Jan 2015 13:13:45 +0000 (14:13 +0100)]
Add support for Qt5's native command line parser
Qt 5.2 finally has grown a way for parsing command line arguments,
namely QCommandLineParser. According to the KF5 porting docs, this
is also the preferred choice for software using KDE Frameworks.
This commit adds a Qt5-based implementation for AbstractCliParser and
uses that instead of our own solution if Quassel is built against Qt5.
We also add support for naming the values that are shown in the --help
output, as QCommandLineParser can make use of that. Once we stop
supporting Qt4, the CLI parser abstraction shall be killed, simplifying
things a lot.
Alf Gaida [Sat, 10 Jan 2015 15:39:33 +0000 (16:39 +0100)]
set GIT_HEAD and GIT_DESCRIBE from environment
Manuel Nickschas [Sat, 3 Jan 2015 00:21:01 +0000 (01:21 +0100)]
Use KTextEdit when building against KF5
When building against Qt5 and WITH_KDE enabled, the input widget
now is a KTextEdit again, making use of features like spell check.
Note that we'll add support for a stand-alone Sonnet framework to have
spell checking without KDE integration later on.
Manuel Nickschas [Sat, 3 Jan 2015 00:19:10 +0000 (01:19 +0100)]
Use link dependencies between Quassel modules
Rather than expressing build dependencies between the Quassel
modules through add_dependency, we should treat those as link
dependencies instead. This has the added advantage that transitive
deps work properly (including properties like include dirs).
Manuel Nickschas [Fri, 2 Jan 2015 23:14:49 +0000 (00:14 +0100)]
Rename HAVE_KDE define to HAVE_KDE4
The build system now sets HAVE_KDE regardless if we're building
with KDE4 or KF5, and in addition adds HAVE_KDE4 and HAVE_KF5.
As many of the frameworks are source compatible with the old kdelibs,
this will allow switching HAVE_KDE4 back to HAVE_KDE later where
possible, without having to check for two defines everywhere.
Manuel Nickschas [Fri, 2 Jan 2015 21:08:12 +0000 (22:08 +0100)]
Prepare build system for KDE Frameworks support
Besides setting everything up for finding frameworks, this commit also
changes the handling of install paths. We now support the standard CMake
variables of the form CMAKE_INSTALL_<TYPE>_DIR, which can be overridden
on the command line. If KDE integration is enabled, the default paths set
by KDE (4 or frameworks) will be used.
Some additional cleanups were done in passing, too.
Manuel Nickschas [Fri, 2 Jan 2015 20:55:24 +0000 (21:55 +0100)]
Happy New Year!
Christian Ratzenhofer [Tue, 23 Dec 2014 22:38:47 +0000 (23:38 +0100)]
Set the version of the QDataStream in ClientAuthHandler
Michael Marley [Sat, 20 Dec 2014 20:22:19 +0000 (15:22 -0500)]
Fix potential state desync in disconnectFromIrc()
Previously, disconnectFromIrc() would assume that if the socket was
in any state other than ConnectedState that it was disconnected and
send the socketDisconnected() signal. However, if the socket was
in a different state, such as ConnectingState, this caused the
corenetwork state and the socket state to become desynchronized.
This patch causes disconnectFromIrc() to close the socket and
start the socketCloseTimer in the event that the socket is in a
state other than UnconnectedState or ConnectedState.
egs-me [Fri, 19 Dec 2014 13:27:31 +0000 (14:27 +0100)]
Merge pull request #103 from mamarley/psqlrefactor2
Improve reliability of PostgreSQL connections
Michael Marley [Fri, 19 Dec 2014 00:02:45 +0000 (19:02 -0500)]
Improve reliability of PostgreSQL connections
This patch adds support for detecting that a query failed due to
the DB connection being down and attempts to reconnect before
retrying the query. It does this for opening transactions,
safeExec(), and prepareAndExecuteQuery().
Michael Marley [Wed, 26 Nov 2014 17:41:43 +0000 (12:41 -0500)]
Add support for compiling with QCA on Qt5
The maintainer for QCA insists on adding a suffix to the QCA
libname, forcing us to make a few changes to support it.
Manuel Nickschas [Sat, 25 Oct 2014 21:28:54 +0000 (23:28 +0200)]
Adapt Oxygen import script to the new realities
Re-running the adapted version afterwards actually added a couple
new icons too, because the old scripts didn't catch them...
Manuel Nickschas [Sat, 25 Oct 2014 21:14:51 +0000 (23:14 +0200)]
Remove the obsolete Icon and IconLoader classes
Removing (hopefully) all traces of the now obsolete icon
loading support.
Manuel Nickschas [Sat, 25 Oct 2014 20:35:01 +0000 (22:35 +0200)]
Add command line parameter for overriding the system icon theme
Quassel uses many icons that are not part of the freedesktop naming
spec, but supported by KDE icon themes. If Quassel is run in
another environment, such as GNOME, which would select a non-KDE icon
theme, there may be missing icons.
In such cases, it is recommended to install the Oxygen icon theme and
start the Quassel client with '--icontheme=oxygen' to override the
system setting. Of course, you are free to choose any other theme as
well.
Note that oxygen is already selected as a fallback if the platform does
not tell Qt about its icon theme (for example, on non-Linux platforms).
Manuel Nickschas [Sat, 25 Oct 2014 20:03:34 +0000 (22:03 +0200)]
Use Qt's native icon theme support rather than our own
For the longest time, we've had our own icon loader (or used KDE's)
in order to provide rudimentary icon theme support. However, since
2009, Qt has had native support for loading icons from themes. In recent
version (due to platform integration etc.), proper icon theme support
has gotten much more complex too... things like theme inheritance,
proper fallback policies, specialized states and so on have never been
supported by our simple custom icon loader.
This commit switches all icon loading to QIcon::fromTheme().
The downside is that we have little control about Qt's icon loader,
which may cause issues with some of the non-standard icons we ship.
In case of the native Qt icon loader being used, we define fallbacks
that should find the extra icons. If you use Qt5 and have the
KDE Frameworks integration plugin for Qt enabled, these fallbacks won't
work unfortunately. If you see missing icons, you may have to install
Quassel's 'hicolor' icons into /usr/share/icons/hicolor so they get
picked up.
For non-Linux platforms, we recommend using the bundled Oxygen icon
set (this is the default).
Manuel Nickschas [Sat, 25 Oct 2014 14:14:12 +0000 (16:14 +0200)]
Rework how icons are handled in the build system
In preparation of moving to native Qt icon theme support, this
commit reworks the handling of icons in the build system:
* Move all vital icons (i.e. application, channel and nicklist icons)
into the hicolor theme
* Always embed the hicolor theme, so we always find the vital icons
as a fallback
* Move the bundled Oxygen theme to oxygen/ and include metadata and
index files, so QIconLoader can find this theme as a fallback
(a system-installed theme will be preferred in the future, however)
* Ship all icon sizes rather than just the requested ones, so Qt can
later pick the most applicable size
* Refactor the Oxygen import script accordingly
* Clean up the CMakeLists.txt
* Adapt resource initialization code
Manuel Nickschas [Sat, 25 Oct 2014 14:01:30 +0000 (16:01 +0200)]
Extend XDG_DATA_DIRS to include Quassel-specific directories
XDG_DATA_DIRS is used for example by Qt to find icon themes and other
application data. In order to avoid having to hack our own locations into
various places, we extend the variable at runtime to include them instead.
This is only relevant if KDE integration is not enabled; otherwise KDE will
specify a list of data dirs (this means that Quassel needs to be installed
properly for everything to work as expected if KDE integration is enabled).
Manuel Nickschas [Sat, 25 Oct 2014 19:59:51 +0000 (21:59 +0200)]
Change applicationName to 'quassel'
QCoreApplication::applicationName() should match things like our
data dir prefix and similar, so 'quassel' makes more sense than
'Quassel IRC'. This will be especially relevant in a KDE Frameworks
world.
Manuel Nickschas [Thu, 23 Oct 2014 21:48:57 +0000 (23:48 +0200)]
Merge pull request #96 from TheOneRing/qca_qt5
make quassel compile with qca enabled on qt5
Manuel Nickschas [Tue, 21 Oct 2014 19:20:07 +0000 (21:20 +0200)]
Check for invalid input in encrypted buffers
The ECB Blowfish decryption function assumed that encrypted input would
always come in blocks of 12 characters, as specified. However, buggy
clients or annoying people may not adhere to that assumption, causing
the core to crash while trying to process the invalid base64 input.
With this commit we make sure that we're not overstepping the bounds of
the input string while decoding it; instead we bail out early and display
the original input. Fixes #1314.
Thanks to Tucos for finding that one!
David ‘Bombe’ Roden [Fri, 10 Oct 2014 05:28:35 +0000 (07:28 +0200)]
Set initial focus to host input widget.
Patrick von Reth [Thu, 2 Oct 2014 13:32:11 +0000 (15:32 +0200)]
make quassel compile with qca enabled on qt5
Marcus Eggenberger [Tue, 30 Sep 2014 10:11:44 +0000 (12:11 +0200)]
Shipping Mac OS X packages with SQL plugins
For core and mono builds, required SQL plugins are now automatically
included in the bundle. Since the core uses a flat package, it
requires a special qt.conf. Here, the Prefix is set to ../ as the
default Prefix on Mac OS points to the Contents direcotry of the
bundle, which does not exist in the flat package.
Manuel Nickschas [Tue, 23 Sep 2014 22:04:08 +0000 (00:04 +0200)]
Post-release version bump
Manuel Nickschas [Tue, 23 Sep 2014 21:54:20 +0000 (23:54 +0200)]
Bump version for release
Manuel Nickschas [Tue, 23 Sep 2014 21:42:02 +0000 (23:42 +0200)]
Update ChangeLog
Manuel Nickschas [Tue, 23 Sep 2014 21:04:41 +0000 (23:04 +0200)]
Merge pull request #83 from m5/hotlist-fix
Fix buffer sorting for hot buffer hotkey
Manuel Nickschas [Tue, 23 Sep 2014 19:53:57 +0000 (21:53 +0200)]
Remove obsolete README.Qt5
Qt5 is fully supported now, and installation instructions have been
in INSTALL for a while, so the old README.Qt5 can go.
Manuel Nickschas [Mon, 22 Sep 2014 20:31:06 +0000 (22:31 +0200)]
Small optimization
No need to copy the list around.
Manuel Nickschas [Mon, 22 Sep 2014 20:32:50 +0000 (22:32 +0200)]
Merge pull request #84 from tecknojock/master
Fixed splitting of CTCP ACTION
Manuel Nickschas [Mon, 22 Sep 2014 18:39:48 +0000 (20:39 +0200)]
Merge pull request #93 from mamarley/quitonfaileddbconn
Quit on database connection failure
Manuel Nickschas [Mon, 22 Sep 2014 18:39:39 +0000 (20:39 +0200)]
Merge pull request #92 from mamarley/nodbnoemit
Do not emit messages if storing them in the DB failed
Manuel Nickschas [Mon, 22 Sep 2014 18:39:25 +0000 (20:39 +0200)]
Merge pull request #94 from mamarley/ssloidentdfix
Rename socketConnected() to socketOpen() to reduce confusion
Manuel Nickschas [Mon, 22 Sep 2014 18:14:20 +0000 (20:14 +0200)]
Merge pull request #95 from fire/travis-pull-request
Tell travis to create a matrix of gcc 4.7, clang 3.4, qt4 and qt5.
K. S. Ernest (iFire) Lee [Sun, 21 Sep 2014 19:57:34 +0000 (12:57 -0700)]
Tell travis to create a matrix of gcc 4.7, clang 3.4, qt4 and qt5.
tecknojock [Wed, 17 Sep 2014 05:28:42 +0000 (00:28 -0500)]
Updated Encode to not run additional times.
Fixed the issue of the message being encoded twice unnessesarily on
actions that did not require splitting.
Michael Marley [Wed, 17 Sep 2014 02:04:11 +0000 (22:04 -0400)]
Rename socketConnected() to socketOpen() to reduce confusion
Michael Marley [Tue, 16 Sep 2014 23:29:12 +0000 (19:29 -0400)]
Quit on database connection failure
Previously, if a Quasselcore was configured to use PostgreSQL but
the PostgreSQL connection failed on startup, it would come up
unconfigured with SQLite. This patch makes the Quasselcore quit
with a fatal error instead.
Manuel Nickschas [Tue, 16 Sep 2014 17:26:38 +0000 (19:26 +0200)]
Merge pull request #89 from mamarley/ssloidentdfix
Improve the odds of winning the oidentd race when using SSL for IRC
Michael Marley [Sat, 6 Sep 2014 02:16:05 +0000 (22:16 -0400)]
Do not emit messages if storing them in the DB failed
If a message is not stored in the database, then some of its IDs
(networkId at least) will be invalid. If this message is sent to
clients, it will cause a nameless empty buffer to appear in the
desktop client and will cause Quasseldroid to crash (of course.)
Better behavior would be to simply not emit the message, since it
can't be displayed properly anyway.
Manuel Nickschas [Sun, 14 Sep 2014 20:16:31 +0000 (22:16 +0200)]
Update AboutDlg
Manuel Nickschas [Sun, 14 Sep 2014 15:22:37 +0000 (17:22 +0200)]
Merge pull request #88 from mamarley/psqllimitallfix
Use null QVariant instead of "ALL" for select limit on psql
Manuel Nickschas [Sun, 14 Sep 2014 15:16:39 +0000 (17:16 +0200)]
Merge pull request #85 from allanjude/patch-1
Update README re: manageusers.py
Manuel Nickschas [Sun, 14 Sep 2014 14:48:12 +0000 (16:48 +0200)]
Merge pull request #82 from jpnurmi/legacypeer
LegacyPeer: kill the dependency to Quassel::buildInfo()
ramz [Sun, 30 Mar 2014 11:07:25 +0000 (22:07 +1100)]
Retains Channel UnHiddens Status after client Restart.
Steps to Reproduce:
1. Hide channel temporarily
2. Part channel
3. Join channel (it will now be visible again, at the bottom of that network's buffer list)
Optional: talk in the channel, get other people to talk, etc but the buffer must not be blue (new message(s)) when you proceed next
4. Quit the client
5. When you'll restart it, the channel will be temporarily hidden again
Root Cause:
After step 3, we are not saving the bufferviewconfig synchronizable object right.
Fix:
Save the bufferviewconfig object correctly.
Fixes #1218.
Michael Marley [Thu, 4 Sep 2014 00:12:31 +0000 (20:12 -0400)]
Improve the odds of winning the oidentd race when using SSL for IRC
Previously, the .oidentd.conf file was not written until
socketInitialized(), which meant that when using SSL for the IRC
connection, the file would not be written until the SSL negotiation
was complete. This delay meant that, in practice, oidentd
integration almost never worked when using SSL IRC connections.
This patch adds a new socketConnected() signal that fires as soon
as the connection is open, regardless of encryption status, and
connects the call to the oidentd configuration generator to this
signal. This gives oidentd integration the same likelihood of
working with SSL connections as with non-SSL connections.
Michael Marley [Mon, 1 Sep 2014 22:42:22 +0000 (18:42 -0400)]
Use null QVariant instead of "ALL" for select limit on psql
Using "ALL" as it was before was making PostgreSQL error out when
attempting to return backlog for Quasseldroid when Quasseldroid is
in "Unread messages per chat" mode. According to
http://www.postgresql.org/docs/9.3/static/queries-limit.html,
LIMIT NULL is the same thing as having no LIMIT at all. According
to http://qt-project.org/doc/qt-4.8/qsqlquery.html, NULL values
are bound in Qt SQL prepared statements by using a null QVariant.