graesslin added a comment.
Your commit message has a typo where X is written but Wayland is meant.
INLINE COMMENTS
> dialog.cpp:1160
>
> -if (ee->region().isNull()) {
> +if (ee->region().isNull() || KWindowSystem::isPlatformX11()) {
> return QQuickWindow::event(
graesslin created this revision.
Restricted Application added projects: Plasma on Wayland, Frameworks.
Restricted Application added a subscriber: plasma-devel.
REVISION SUMMARY
As https://phabricator.kde.org/R127:98e5d269a110dca9cae1e4ed2bb3051f02a69bff
just for pointer.
REPOSITORY
R127 KWay
This revision was automatically updated to reflect the committed changes.
Closed by commit R127:bea87753629e: [client] Properly track enteredSurface in
Keyboard (authored by graesslin).
REPOSITORY
R127 KWayland
CHANGES SINCE LAST UPDATE
https://phabricator.kde.org/D6741?vs=16786&id=16797
RE
graesslin created this revision.
Restricted Application added projects: Plasma on Wayland, Frameworks.
Restricted Application added a subscriber: plasma-devel.
REVISION SUMMARY
ASAN found a heap-use-after-free when deleting the focused keyboard
surface in the client library. Keyboard did not t
graesslin added a comment.
In https://phabricator.kde.org/D6727#125900, @davidedmundson wrote:
> We still need https://phabricator.kde.org/D6571, with my proposed change.
That's a special case where we delete the connection before the QPA.
It might be that we don't need it. Also K
This revision was automatically updated to reflect the committed changes.
Closed by commit R127:98e5d269a110: [server] Send keyboard leave when client
destroys the focused surface (authored by graesslin).
CHANGED PRIOR TO COMMIT
https://phabricator.kde.org/D6683?vs=16755&id=16775#toc
REPOSITOR
graesslin added a comment.
I like the idea! That would basically allow us to abandon
https://phabricator.kde.org/D6571 and would also solve the issues we see with
the kwayland-integration plugin which can crash applications on exit.
May I ask for a small unit test for the new functionali
graesslin updated this revision to Diff 16755.
graesslin added a comment.
Restricted Application edited projects, added Plasma; removed Plasma on Wayland.
Add nullptr check to prevent crash
REPOSITORY
R127 KWayland
CHANGES SINCE LAST UPDATE
https://phabricator.kde.org/D6683?vs=16661&id=16
graesslin planned changes to this revision.
graesslin added a comment.
Causes crash in KWayaland when closing a window.
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D6683
To: graesslin, #frameworks, #plasma
Cc: plasma-devel, ZrenBot, progwolff, lesliezhai, ali-moh
graesslin created this revision.
Restricted Application added projects: Plasma on Wayland, Frameworks.
Restricted Application added a subscriber: plasma-devel.
REVISION SUMMARY
This is a change inspired by https://bugreports.qt.io/browse/QTBUG-61930.
When Qt closes a window due to a key press
graesslin updated this revision to Diff 16660.
graesslin added a comment.
Restricted Application edited projects, added Plasma on Wayland; removed Plasma.
Make this compile. Looks like kdevelop fooled me and compiled KWin instead of
KWayland (that happens all the time to me)
REPOSITORY
R127
graesslin updated this revision to Diff 16406.
graesslin added a comment.
Restricted Application edited projects, added Plasma; removed Plasma on Wayland.
Use mutex to protect the potentially shared data structure
REPOSITORY
R127 KWayland
CHANGES SINCE LAST UPDATE
https://phabricator.kde.
graesslin added a comment.
@davidedmundson if you have another idea how to tackle the problem I'm
absolutely open to it.
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D6569
To: graesslin, #frameworks, #plasma, #kwin
Cc: davidedmundson, dfaure, plasma-devel, ZrenBo
graesslin added a comment.
In https://phabricator.kde.org/D6569#123081, @davidedmundson wrote:
> I don't think this is necessary.
>
> The only times we want need to delete the connection ourselves is when
client code is creating it via ConnectionThread::fromApplication. If they
creat
graesslin added a dependent revision: D6571: Delete all Wayland connections by
plugins prior to own Wayland connection.
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D6569
To: graesslin, #frameworks, #plasma, #kwin
Cc: dfaure, plasma-devel, ZrenBot, progwolff, lesliez
graesslin created this revision.
Restricted Application added projects: Plasma on Wayland, Frameworks.
Restricted Application added a subscriber: plasma-devel.
REVISION SUMMARY
This is change needed by KWin. KWin has the problem that it destroys its
internal Wayland connection (KWin as client
graesslin added a comment.
Erm why was this pushed although it was in state changes requested?
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D6523
To: mart, #plasma, davidedmundson, graesslin
Cc: graesslin, davidedmundson, plasma-devel, #frameworks, ZrenBot, progwo
graesslin requested changes to this revision.
graesslin added a comment.
This revision now requires changes to proceed.
please add a test case for the problem.
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D6523
To: mart, #plasma, davidedmundson, graesslin
Cc: grae
graesslin added a comment.
In https://phabricator.kde.org/D6047#119698, @mart wrote:
> trying to implement the kwin part of ping btw, will do in a branch of the
branch kwin/xdgv6
no need to hurry with it. IMHO the way ping works is completely broken and
won't work. At least Qt ho
graesslin added a comment.
could you please upload a version with context?
INLINE COMMENTS
> mart wrote in xdgshell_interface.h:80
> from kwin, client.cpp:
> i guess that's the kwin part that will have to use this.
> it uses a single timer, and tries for a couple of timeout, as soon as the
>
graesslin added inline comments.
INLINE COMMENTS
> kkeyserver.cpp:976-980
> +if (*keyQt != Qt::Key_Tab) { // KKeySequenceWidget does not map
> shift+tab to backtab
> +static const int FirstLevelShift = 1;
> +keySymX = xcb_key_symbols_get_keysym(symbols, e->detail,
graesslin added a comment.
> Man it's demotivating to contribute to KDE. Users say all sorts of bad
things about KDE, and then future-ex-maintainers reject your contribution.
Great.
Yes sure, and you can imagine how many angry mails and bug reports I get when
things break. I made the ex
graesslin added a comment.
> I need to point out that this creates a functional difference to Wayland
and according to the latest rules of Plasma such changes are no longer allowed
unless the implementation is done first for Wayland.
After reading through KWin code together with your KWi
graesslin requested changes to this revision.
graesslin added inline comments.
This revision now requires changes to proceed.
INLINE COMMENTS
> kkeyserver.cpp:160-180
> +{ Qt::Key_Space, XK_KP_Space },
> +{ Qt::Key_Tab,XK_KP_Tab },
> +{ Qt::Key_Enter, XK_KP_Enter },
graesslin added a comment.
I need to point out that this creates a functional difference to Wayland and
according to the latest rules of Plasma such changes are no longer allowed
unless the implementation is done first for Wayland.
Personally I am not really comfortable with such a large
graesslin accepted this revision.
graesslin added a comment.
This revision is now accepted and ready to land.
Eh what? That is not thread save?!? Oh sh*** I didn't expect that. Good catch
and a nice solution.
REPOSITORY
R127 KWayland
BRANCH
master
REVISION DETAIL
https://phabricator.k
graesslin accepted this revision.
graesslin added a comment.
This revision is now accepted and ready to land.
Thanks!
REPOSITORY
R127 KWayland
BRANCH
asdfasdf
REVISION DETAIL
https://phabricator.kde.org/D5887
To: davidedmundson, #plasma, graesslin
Cc: graesslin, plasma-devel, #framewo
graesslin added a comment.
In https://phabricator.kde.org/D5872#109962, @hein wrote:
> I'm fine with no change signal, but heads-up that libtaskmanager has code
connecting to it, so if you remove it from kwayland please also adapt
plasma-workspace or the build breaks.
I highly su
graesslin added a comment.
like David explained: we don't need the change signal and IMHO we shouldn't
expose "wrong API". I consider having a changed signal for a PID wrong API as
it indicates that the PID could change. But it won't ever change, so we
shouldn't expose it. I know that all ot
graesslin added a comment.
In https://phabricator.kde.org/D5851#109478, @davidedmundson wrote:
> It makes absolutely no sense to have a pidChanged signal and then say
clients shouldn't be using it.
>
> Either window has no reason to have a pidChanged signal or we should
connect to it
graesslin added inline comments.
INLINE COMMENTS
> plasmawindowmodel.cpp:84
>
> +QObject::connect(window, &PlasmaWindow::pidChanged, q,
> +[window, this] { this->dataChanged(window, PlasmaWindowModel::Pid); }
In the original report I critized that we connect to a pid changed signal
graesslin reopened this revision.
graesslin added a comment.
This revision is now accepted and ready to land.
Can reproduce the failing test locally. Please fix ASAP! Broken tests is not
acceptable in KWayland. This change has been on review so long, I find it quite
disappointing that it was
graesslin accepted this revision.
This revision is now accepted and ready to land.
REPOSITORY
R127 KWayland
BRANCH
sebas/processid
REVISION DETAIL
https://phabricator.kde.org/D5747
To: sebas, #plasma, hein, graesslin
Cc: apol, davidedmundson, plasma-devel, #frameworks, ZrenBot, spstarr,
graesslin requested changes to this revision.
graesslin added inline comments.
This revision now requires changes to proceed.
INLINE COMMENTS
> plasmawindowmodel.cpp:84
>
> +
> QObject::connect(window, &PlasmaWindow::activeChanged, q,
Unrelated newline
> plasma-window-management.xml:263
graesslin added a comment.
Your tests in plasma window management do not test the new requests.
INLINE COMMENTS
> test_wayland_windowmanagement.cpp:483
> +using namespace KWayland::Server;
> +// this test verifies setting the showing desktop state
> +QVERIFY(!m_window->isKeepAbove
graesslin added a comment.
In https://phabricator.kde.org/D5747#108916, @apol wrote:
> Who is going to be setting this pid? Because note that containerized
applications don't know their own pid (see kdbusaddons & knotifications patches
for big shrug).
It's KWin. Who gets the info
graesslin added a comment.
In https://phabricator.kde.org/D5757#108706, @hein wrote:
> Martin, did you mix things up here maybe? The patch has changes to
autotests/client/test_wayland_windowmanagement.cpp. The "newRow" bit isn't
related to QAIM. It's a row of test data.
no, I ans
graesslin added a comment.
Forget my last two comments - confused the review
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D5757
To: mart, #plasma, hein, graesslin, #plasma_on_wayland
Cc: graesslin, hein, plasma-devel, #frameworks, ZrenBot, spstarr, progwolff,
les
graesslin added inline comments.
INLINE COMMENTS
> test_plasma_window_model.cpp:215
>
> void PlasmaWindowModelTest::testRoleNames_data()
> {
Please extend this
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D5747
To: sebas, #plasma, hein, graesslin
Cc: davidedmun
graesslin added inline comments.
INLINE COMMENTS
> test_plasma_window_model.cpp:214
>
> void PlasmaWindowModelTest::testRoleNames_data()
> {
Please extend this
> test_plasma_window_model.cpp:306
>
> void PlasmaWindowModelTest::testDefaultData_data()
> {
Please extend this
REPOSITORY
graesslin added inline comments.
INLINE COMMENTS
> plasmawindowmodel.cpp:84-86
> +QObject::connect(window, &PlasmaWindow::pidChanged, q,
> +[window, this] { this->dataChanged(window, PlasmaWindowModel::Pid); }
> +);
I consider it as impossible that the PID changes.
REPOSITORY
graesslin accepted this revision.
graesslin added a comment.
This revision is now accepted and ready to land.
Are you sure the API is so compatible that you can run it like that?
REPOSITORY
R127 KWayland
BRANCH
davidedmundson/xdgv6
REVISION DETAIL
https://phabricator.kde.org/D5801
To:
graesslin requested changes to this revision.
This revision now requires changes to proceed.
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D5757
To: mart, #plasma, hein, graesslin, #plasma_on_wayland
Cc: graesslin, hein, plasma-devel, #frameworks, ZrenBot, spstarr, pro
graesslin added a comment.
In https://phabricator.kde.org/D5757#108115, @mart wrote:
> In https://phabricator.kde.org/D5757#108070, @graesslin wrote:
>
> > Please also extend the test in
autotests/client/test_wayland_windowmanagement.cpp
>
>
> isn't it covered by
no, tha
graesslin requested changes to this revision.
graesslin added a comment.
This revision now requires changes to proceed.
Please also extend the test in
autotests/client/test_wayland_windowmanagement.cpp
REPOSITORY
R127 KWayland
REVISION DETAIL
https://phabricator.kde.org/D5757
To: mart,
graesslin requested changes to this revision.
graesslin added a comment.
This revision now requires changes to proceed.
Please also extend the autotests/client/test_wayland_windowmanagement.cpp
INLINE COMMENTS
> plasmawindowmanagement.h:385
> + * @returns The process id this window belong
Am 2017-05-05 18:37, schrieb Detlef Eppers:
File indexing tools encounter all different types of files, both of
trusted and untrusted origin, and last November it was demonstrated by
Chris Evans how this can potentially be exploited*. In order to
provide a means for mitigation, I have written a F
201 - 247 of 247 matches
Mail list logo