Re: Review Request 122206: [kio] Make tests optional

2015-03-16 Thread Albert Vaca Cintora 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122206/#review77603
---


I know this is merged already but this patch is being applied to every KDE 
package and I want to keep the discussion in a single place.

We already have a toggle option in CMake that is "BUILD_TESTING". If Gentoo 
wants to not build the tests (I'm not judging if they should, let them be free 
to do it), they can just set BUILD_TESTING to OFF. I understand that CMake will 
still try to find Qt5Test and fail, but here is where I think we got it wrong:

This patch does the following:
if (Qt5Test is not found) 
BUILD_TESTING = OFF

What I think this patch should be doing is this:
if (BUILD_TESTING == OFF) 
Don't look for Qt5Test

Did I miss something or this seems more reasonable to you guys as well?

- Albert Vaca Cintora


On Feb. 6, 2015, 4:14 p.m., Andreas Sturmlechner wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122206/
> ---
> 
> (Updated Feb. 6, 2015, 4:14 p.m.)
> 
> 
> Review request for KDE Frameworks.
> 
> 
> Repository: kio
> 
> 
> Description
> ---
> 
> [kio] Make tests optional
> This is a small patch to CMakeLists.txt to only depend on Qt5Test if 
> BUILD_TESTING.
> 
> 
> Diffs
> -
> 
>   CMakeLists.txt c1ed03f6cac648517828aec60e896baf9fbcfd9d 
> 
> Diff: https://git.reviewboard.kde.org/r/122206/diff/
> 
> 
> Testing
> ---
> 
> 
> Thanks,
> 
> Andreas Sturmlechner
> 
>

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122484: KCodecs: Fix i18n of character set names

2015-03-16 Thread Lasse Liehu 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122484/
---

(Updated maalis 17, 2015, 12:19 ap)


Review request for KDE Frameworks.


Changes
---

Removed outdated information in description


Repository: kcodecs


Description (updated)
---

Fix i18n of character set names

Now the tr calls match what is extracted into catalogs.


Diffs
-

  src/kcharsets.cpp 83a461b 

Diff: https://git.reviewboard.kde.org/r/122484/diff/


Testing
---

Manual testing with KWrite and Konversation that use these functions, using qm 
files generated with lrelease without -compress option.


Thanks,

Lasse Liehu

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122484: KCodecs: Fix i18n of character set names

2015-03-16 Thread Lasse Liehu 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122484/
---

(Updated maalis 17, 2015, 12:15 ap)


Review request for KDE Frameworks.


Changes
---

Specified "KCharsets" as context. Now tr calls don't need any changes.


Repository: kcodecs


Description
---

Fix i18n of character set names

Now the translate calls match what is extracted into catalogs.

For these translate calls to work with qm files generated by 
ecm_process_po_files_as_qm, the following Qt bug needs to be fixed: 
https://bugreports.qt.io/browse/QTBUG-44362


Diffs (updated)
-

  src/kcharsets.cpp 83a461b 

Diff: https://git.reviewboard.kde.org/r/122484/diff/


Testing
---

Manual testing with KWrite and Konversation that use these functions, using qm 
files generated with lrelease without -compress option.


Thanks,

Lasse Liehu

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: What's the point of tiers if we bend the definition of them?

2015-03-16 Thread Jeremy Whiting 
+1

On Mon, Mar 16, 2015 at 4:16 PM, Albert Astals Cid  wrote:

> We have KPackage depending on kdoctools but since it's optional we pretend
> it's not a real dependency and call KPackage tier2
>
> What's in it for us other than lying to ourselves? What's so bad about
> KPackage being tier 3?
>
> It'd still have only 3 dependencies and be totally usable, no?
>
> Cheers,
>   Albert
> ___
> Kde-frameworks-devel mailing list
> Kde-frameworks-devel@kde.org
> https://mail.kde.org/mailman/listinfo/kde-frameworks-devel
>
___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

What's the point of tiers if we bend the definition of them?

2015-03-16 Thread Albert Astals Cid 
We have KPackage depending on kdoctools but since it's optional we pretend 
it's not a real dependency and call KPackage tier2

What's in it for us other than lying to ourselves? What's so bad about 
KPackage being tier 3?

It'd still have only 3 dependencies and be totally usable, no?

Cheers,
  Albert
___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Review Request 122981: add KGlobalAccel::loadShortcutFromGlobalSettings

2015-03-16 Thread Gregor Mi 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122981/
---

Review request for KDE Frameworks, Martin Gräßlin and Thomas Lübking.


Repository: kglobalaccel


Description
---

In some cases you need to call loadShortcutFromGlobalSettings() in order not to 
get a an empty list when calling shortcut(). See discussion in 
https://git.reviewboard.kde.org/r/122249/.


Diffs
-

  src/kglobalaccel.h 3fe20ca8e4ec6ceb0bb9e54235aef7f1aeeb8c16 
  src/kglobalaccel.cpp 1b6b3f5cb6d42401d684e6a491d12a6e57248fd1 

Diff: https://git.reviewboard.kde.org/r/122981/diff/


Testing
---


Thanks,

Gregor Mi

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122978: Add unit test for crash of bug 343976

2015-03-16 Thread Marco Martin 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122978/#review77586
---

Ship it!


Ship It!

- Marco Martin


On March 16, 2015, 12:04 p.m., Martin Gräßlin wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122978/
> ---
> 
> (Updated March 16, 2015, 12:04 p.m.)
> 
> 
> Review request for KDE Frameworks.
> 
> 
> Bugs: 343976
> https://bugs.kde.org/show_bug.cgi?id=343976
> 
> 
> Repository: frameworkintegration
> 
> 
> Description
> ---
> 
> Use a QPointer for QMenu of SystemTrayMenu
> 
> We need better memory management as the QMenu the SystemTrayMenu creates
> might be deleted externally causing a double delete if the so-far
> QScopedPoiter cleans up.
> 
> Switch to QPointer and call deleteLater if the pointer is still valid
> during destroying the SystemTrayMenu.
> 
> BUG: 343976
> CHANGELOG: Fix possible crash when destroying a QSystemTrayIcon (triggered by 
> Trojita)
> 
> 
> Diffs
> -
> 
>   autotests/CMakeLists.txt e8ed6a99bb45969231ba21b8c5588e093bbbe3ee 
>   autotests/ksni_unittest.cpp PRE-CREATION 
>   src/platformtheme/kdeplatformsystemtrayicon.h 
> 8f13a36ae7cebed7499ce8a160f3533665b909ac 
>   src/platformtheme/kdeplatformsystemtrayicon.cpp 
> ce3d3de9f9cf479a5fe3af65ef7fb8ec2b9e0d29 
> 
> Diff: https://git.reviewboard.kde.org/r/122978/diff/
> 
> 
> Testing
> ---
> 
> 
> Thanks,
> 
> Martin Gräßlin
> 
>

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Review Request 122979: Add a unit test which performs testing a shortcut

2015-03-16 Thread Martin Gräßlin 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122979/
---

Review request for KDE Frameworks and Thomas Lübking.


Repository: kglobalaccel


Description
---

Uses X11's XTEST extension to simulate keypress/release.

One of the tested shortcuts needed to have the key changed as it seems
to be that there is no keycode for F28 by default.


Diffs
-

  CMakeLists.txt 91a235d824b4cddaf7737e35761fdad2d4f53c80 
  autotests/CMakeLists.txt b1a41ad19384589798ff6bdd512c5279003eac8a 
  autotests/kglobalshortcuttest.h 2a7b40d34875e16345a659fb9764e4f536ad79c6 
  autotests/kglobalshortcuttest.cpp 169bf92ffbff985cd4381e771c2fcecaff77156b 

Diff: https://git.reviewboard.kde.org/r/122979/diff/


Testing
---


Thanks,

Martin Gräßlin

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: New framework: ModemManagerQt

2015-03-16 Thread Jan Grulich 
On Thursday 12 of March 2015 14:12 David Edmundson wrote:
> Looks good to me.
> 2 minor comments.
> 
> All classes are namespaced, but generictypes.h is not.
> Given these names could easily clash with something else and are publicly
> included, it might be worth putting them in the same namespace.

Done.

> ModemManager::ModemMessaging::messages can be const?

Yes, fixed. 

> 
> David
> 
> ​

On Friday 13 of March 2015 00:06 Albert Astals Cid wrote:
> Kill the framework branch?

Removed.

> Move macros.h and mmdebug.h to _p.h?

Done.

> Can listBearers and findBearer be const?

Yep, done.

> In InterfaceType i'd say you can let the enums be there even if the
> MM_CHECK_VERSION doesn't match, makes sure in case some others are added
> later they always have the same "int" value

Removed check.

> Make BearerStruct, IpConfig and NetworkTimeZonea and the structs in
> generictypes.h classes with dptr in case you ever need more fields in them?

I made only BearerStruct, IpConfig and NetworkTimeZone as classes. Other 
structures are just pairs and won't need more fields in future.

> Add const & to params of ip4ConfigChanged, ip6ConfigChanged and
> networkTimeZoneChanged?

Done.

> Cheers,
>   Albert
> 

Thanks, is there anything else?

Regards,
Jan
-- 
Jan Grulich 
Red Hat Czech, s.r.o
jgrul...@redhat.com
___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122978: Add unit test for crash of bug 343976

2015-03-16 Thread Lukáš Tinkl 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122978/#review77578
---


+1 from me

- Lukáš Tinkl


On Bře. 16, 2015, 1:04 odp., Martin Gräßlin wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122978/
> ---
> 
> (Updated Bře. 16, 2015, 1:04 odp.)
> 
> 
> Review request for KDE Frameworks.
> 
> 
> Bugs: 343976
> https://bugs.kde.org/show_bug.cgi?id=343976
> 
> 
> Repository: frameworkintegration
> 
> 
> Description
> ---
> 
> Use a QPointer for QMenu of SystemTrayMenu
> 
> We need better memory management as the QMenu the SystemTrayMenu creates
> might be deleted externally causing a double delete if the so-far
> QScopedPoiter cleans up.
> 
> Switch to QPointer and call deleteLater if the pointer is still valid
> during destroying the SystemTrayMenu.
> 
> BUG: 343976
> CHANGELOG: Fix possible crash when destroying a QSystemTrayIcon (triggered by 
> Trojita)
> 
> 
> Diffs
> -
> 
>   autotests/CMakeLists.txt e8ed6a99bb45969231ba21b8c5588e093bbbe3ee 
>   autotests/ksni_unittest.cpp PRE-CREATION 
>   src/platformtheme/kdeplatformsystemtrayicon.h 
> 8f13a36ae7cebed7499ce8a160f3533665b909ac 
>   src/platformtheme/kdeplatformsystemtrayicon.cpp 
> ce3d3de9f9cf479a5fe3af65ef7fb8ec2b9e0d29 
> 
> Diff: https://git.reviewboard.kde.org/r/122978/diff/
> 
> 
> Testing
> ---
> 
> 
> Thanks,
> 
> Martin Gräßlin
> 
>

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122978: Add unit test for crash of bug 343976

2015-03-16 Thread Emmanuel Pescosta 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122978/#review77577
---



src/platformtheme/kdeplatformsystemtrayicon.cpp


unchecked m_menu


- Emmanuel Pescosta


On March 16, 2015, 1:04 p.m., Martin Gräßlin wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122978/
> ---
> 
> (Updated March 16, 2015, 1:04 p.m.)
> 
> 
> Review request for KDE Frameworks.
> 
> 
> Bugs: 343976
> https://bugs.kde.org/show_bug.cgi?id=343976
> 
> 
> Repository: frameworkintegration
> 
> 
> Description
> ---
> 
> Use a QPointer for QMenu of SystemTrayMenu
> 
> We need better memory management as the QMenu the SystemTrayMenu creates
> might be deleted externally causing a double delete if the so-far
> QScopedPoiter cleans up.
> 
> Switch to QPointer and call deleteLater if the pointer is still valid
> during destroying the SystemTrayMenu.
> 
> BUG: 343976
> CHANGELOG: Fix possible crash when destroying a QSystemTrayIcon (triggered by 
> Trojita)
> 
> 
> Diffs
> -
> 
>   autotests/CMakeLists.txt e8ed6a99bb45969231ba21b8c5588e093bbbe3ee 
>   autotests/ksni_unittest.cpp PRE-CREATION 
>   src/platformtheme/kdeplatformsystemtrayicon.h 
> 8f13a36ae7cebed7499ce8a160f3533665b909ac 
>   src/platformtheme/kdeplatformsystemtrayicon.cpp 
> ce3d3de9f9cf479a5fe3af65ef7fb8ec2b9e0d29 
> 
> Diff: https://git.reviewboard.kde.org/r/122978/diff/
> 
> 
> Testing
> ---
> 
> 
> Thanks,
> 
> Martin Gräßlin
> 
>

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122978: Add unit test for crash of bug 343976

2015-03-16 Thread Jan Kundrát 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122978/#review77574
---

Ship it!


+2 from me -- it fixes a crash for me.

+I don't know the code and API asumptions to know whether it's better to just 
Q_ASSERT in the other methods (`setText` etc.), but even if it was better to 
assert crash, this patch does provide a big improvement :).

- Jan Kundrát


On March 16, 2015, 12:04 p.m., Martin Gräßlin wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122978/
> ---
> 
> (Updated March 16, 2015, 12:04 p.m.)
> 
> 
> Review request for KDE Frameworks.
> 
> 
> Bugs: 343976
> https://bugs.kde.org/show_bug.cgi?id=343976
> 
> 
> Repository: frameworkintegration
> 
> 
> Description
> ---
> 
> Use a QPointer for QMenu of SystemTrayMenu
> 
> We need better memory management as the QMenu the SystemTrayMenu creates
> might be deleted externally causing a double delete if the so-far
> QScopedPoiter cleans up.
> 
> Switch to QPointer and call deleteLater if the pointer is still valid
> during destroying the SystemTrayMenu.
> 
> BUG: 343976
> CHANGELOG: Fix possible crash when destroying a QSystemTrayIcon (triggered by 
> Trojita)
> 
> 
> Diffs
> -
> 
>   autotests/CMakeLists.txt e8ed6a99bb45969231ba21b8c5588e093bbbe3ee 
>   autotests/ksni_unittest.cpp PRE-CREATION 
>   src/platformtheme/kdeplatformsystemtrayicon.h 
> 8f13a36ae7cebed7499ce8a160f3533665b909ac 
>   src/platformtheme/kdeplatformsystemtrayicon.cpp 
> ce3d3de9f9cf479a5fe3af65ef7fb8ec2b9e0d29 
> 
> Diff: https://git.reviewboard.kde.org/r/122978/diff/
> 
> 
> Testing
> ---
> 
> 
> Thanks,
> 
> Martin Gräßlin
> 
>

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122876: Delay notifications a bit on Plasma startup

2015-03-16 Thread Martin Klapetek 


> On March 12, 2015, 8:48 a.m., Martin Gräßlin wrote:
> > timer looks wrong to me, too. Not because it's ugly, but because Plasma 
> > notifications will not go above the splash screen. The fallback might be 
> > ugly, but as I have seen: it goes above the splash screen.
> > 
> > Thus I suggest that we implement missing pieces to figure out whether we 
> > are in startup and only show if startup is done.
> 
> Martin Klapetek wrote:
> I'm all for it, but note that splash screen is terminated once Plasma 
> starts, there should never be a case of Plasma running and being fully 
> covered by the splash screen.
> 
> Martin Gräßlin wrote:
> The splash screen gets faded out, so there is a certain chance that it 
> will be hidden before the animation finishes - especially if the required 
> DBus service is up before the desktop startup has fully finished.

So I thought about it and to make things as uncomplicated as possible, I could 
simply check if KSplash is running, it does export a dbus interface, but this 
means a dbus roundtrip each time a KNotification is send. Alternatively it 
could check if the process is running.

Comments or better ideas?


- Martin


---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122876/#review77324
---


On March 9, 2015, 8:03 p.m., Martin Klapetek wrote:
> 
> ---
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122876/
> ---
> 
> (Updated March 9, 2015, 8:03 p.m.)
> 
> 
> Review request for KDE Frameworks and Eike Hein.
> 
> 
> Bugs: 344903
> https://bugs.kde.org/show_bug.cgi?id=344903
> 
> 
> Repository: knotifications
> 
> 
> Description
> ---
> 
> Currently when something is started right after login and spawns a 
> notification, an ugly popup (KPassivePopup fallback) will appear over the 
> splash screen as the org.freedesktop.Notifications service is not ready yet. 
> 
> This patch delays the notifications by max 25 seconds if and only if 
> KDE_FULL_SESSION is set and there's no Plasma (org.kde.plasmashell) running. 
> If the org.freedesktop.Notifications does not appear within those 25 seconds, 
> the notifications will be put on screen using KPassivePopup.
> 
> Ideally this should also check if ksmserver is in the "starting phase", but I 
> haven't found a way to check for that. Suggestions welcome.
> 
> 
> Diffs
> -
> 
>   src/notifybypopup.h 416c533 
>   src/notifybypopup.cpp 316ff2b 
> 
> Diff: https://git.reviewboard.kde.org/r/122876/diff/
> 
> 
> Testing
> ---
> 
> Plasma not running, emitted KNotification, nothing, started Plasma, 
> notifications appeared when Plasma loaded.
> 
> 
> Thanks,
> 
> Martin Klapetek
> 
>

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122978: Add unit test for crash of bug 343976

2015-03-16 Thread Martin Gräßlin 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122978/
---

(Updated March 16, 2015, 1:04 p.m.)


Review request for KDE Frameworks.


Changes
---

Fix new issue


Bugs: 343976
https://bugs.kde.org/show_bug.cgi?id=343976


Repository: frameworkintegration


Description
---

Use a QPointer for QMenu of SystemTrayMenu

We need better memory management as the QMenu the SystemTrayMenu creates
might be deleted externally causing a double delete if the so-far
QScopedPoiter cleans up.

Switch to QPointer and call deleteLater if the pointer is still valid
during destroying the SystemTrayMenu.

BUG: 343976
CHANGELOG: Fix possible crash when destroying a QSystemTrayIcon (triggered by 
Trojita)


Diffs (updated)
-

  autotests/CMakeLists.txt e8ed6a99bb45969231ba21b8c5588e093bbbe3ee 
  autotests/ksni_unittest.cpp PRE-CREATION 
  src/platformtheme/kdeplatformsystemtrayicon.h 
8f13a36ae7cebed7499ce8a160f3533665b909ac 
  src/platformtheme/kdeplatformsystemtrayicon.cpp 
ce3d3de9f9cf479a5fe3af65ef7fb8ec2b9e0d29 

Diff: https://git.reviewboard.kde.org/r/122978/diff/


Testing
---


Thanks,

Martin Gräßlin

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122978: Add unit test for crash of bug 343976

2015-03-16 Thread Martin Gräßlin 


> On March 16, 2015, 12:26 p.m., Jan Kundrát wrote:
> > This doesn't fix the problem triggered by Trojita:
> > 
> > ```
> > Program received signal SIGSEGV, Segmentation fault.
> > QWidget::removeAction (this=0x0, action=0x55bf88a0) at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:3298
> > 3298
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:
> >  No such file or directory.
> > (gdb) bt
> > #0  QWidget::removeAction (this=0x0, action=0x55bf88a0) at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:3298
> > #1  0x7fffe792e1a0 in SystemTrayMenu::removeMenuItem 
> > (this=0x55cccd10, menuItem=)
> > at 
> > /var/tmp/portage/kde-frameworks/frameworkintegration-/work/frameworkintegration-/src/platformtheme/kdeplatformsystemtrayicon.cpp:96
> > #2  0x777a785f in QMenu::actionEvent (this=0x55ade210, 
> > e=)
> > at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/widgets/qmenu.cpp:3193
> > #3  0x7765f960 in QWidget::event (this=this@entry=0x55ade210, 
> > event=event@entry=0x7fffbc80)
> > at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:9042
> > #4  0x777adb8b in QMenu::event (this=0x55ade210, 
> > e=0x7fffbc80) at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/widgets/qmenu.cpp:2650
> > #5  0x776104dc in QApplicationPrivate::notify_helper 
> > (this=this@entry=0x55a72090, receiver=receiver@entry=0x55ade210, 
> > e=e@entry=0x7fffbc80)
> > at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qapplication.cpp:3716
> > #6  0x77615eba in QApplication::notify (this=0x7fffd4c0, 
> > receiver=0x55ade210, e=0x7fffbc80)
> > at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qapplication.cpp:3499
> > #7  0x7440b9cd in QCoreApplication::notifyInternal 
> > (this=0x7fffd4c0, receiver=receiver@entry=0x55ade210, 
> > event=event@entry=0x7fffbc80)
> > at 
> > /var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qcoreapplication.cpp:963
> > #8  0x77655ff9 in sendEvent (event=0x7fffbc80, 
> > receiver=0x55ade210) at 
> > ../../include/QtCore/../../../qtwidgets-5.5./src/corelib/kernel/qcoreapplication.h:228
> > #9  QWidget::removeAction (this=0x55ade210, 
> > action=action@entry=0x55c26fc0)
> > at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:3305
> > #10 0x77603b51 in QAction::~QAction (this=0x55c26fc0, 
> > __in_chrg=)
> > at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qaction.cpp:573
> > #11 0x77603cdf in QAction::~QAction (this=0x55c26fc0, 
> > __in_chrg=)
> > at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qaction.cpp:592
> > #12 0x74441feb in QObjectPrivate::deleteChildren 
> > (this=this@entry=0x55cb6a10)
> > at 
> > /var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qobject.cpp:1951
> > #13 0x7444d0dc in QObject::~QObject (this=, 
> > __in_chrg=)
> > at 
> > /var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qobject.cpp:1031
> > #14 0x779fd62f in QSystemTrayIcon::~QSystemTrayIcon 
> > (this=0x55cc90a0, __in_chrg=)
> > at 
> > /var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/util/qsystemtrayicon.cpp:150
> > #15 0x555c39ac in Gui::MainWindow::removeSysTray 
> > (this=this@entry=0x7fffd570) at 
> > /home/jkt/work/prog/trojita/src/Gui/Window.cpp:795
> > #16 0x555d5f88 in Gui::MainWindow::slotShowSettings 
> > (this=0x7fffd570) at /home/jkt/work/prog/trojita/src/Gui/Window.cpp:1150
> > #17 0x555e51ca in Gui::MainWindow::qt_static_metacall 
> > (_o=, _c=, _id=, _a= > out>) at moc_Window.cpp:397
> > #18 0x74443bf8 in QMetaObject::activate 
> > (sender=sender@entry=0x55bc61a0, signalOffset=, 
> > local_signal_index=local_signal_index@entry=1, 
> > argv=argv@entry=0x7fffc110) at 
> > /var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qobject.cpp:3718
> > #19 0x7315 in QMetaObject::activate 
> > (sender=sender@entry=0x55bc61a0, m=m@entry=0x77b96040 
> > , local_signal_index=local_signal_index@entry=1, 
> > argv=argv@entry=0x7fffc110) at 
> > /var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qobject.cpp:3583
> > #20 0x77603e1f in QAction::triggered 
> > (this=this@entry=0x55bc61a0, _t1=false) at .

Re: Review Request 122908: Fix segfaults due to missing screens

2015-03-16 Thread Jan Kundrát 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122908/
---

(Updated March 16, 2015, 11:28 a.m.)


Status
--

This change has been discarded.


Review request for KDE Frameworks.


Repository: kwindowsystem


Description
---

Fix segfaults due to missing screens

Qt 5.5 has some changes which allow returning of null QScreen*. This
file was OK with such a situation throughout a lot of other places, but
these two were missing.


Diffs
-

  src/kwindoweffects_x11.cpp 661a0af95a7a1757c5ae863270434b0c2d5c0a26 

Diff: https://git.reviewboard.kde.org/r/122908/diff/


Testing
---


Thanks,

Jan Kundrát

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122978: Add unit test for crash of bug 343976

2015-03-16 Thread Jan Kundrát 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122978/#review77566
---


This doesn't fix the problem triggered by Trojita:

```
Program received signal SIGSEGV, Segmentation fault.
QWidget::removeAction (this=0x0, action=0x55bf88a0) at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:3298
3298
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:
 No such file or directory.
(gdb) bt
#0  QWidget::removeAction (this=0x0, action=0x55bf88a0) at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:3298
#1  0x7fffe792e1a0 in SystemTrayMenu::removeMenuItem (this=0x55cccd10, 
menuItem=)
at 
/var/tmp/portage/kde-frameworks/frameworkintegration-/work/frameworkintegration-/src/platformtheme/kdeplatformsystemtrayicon.cpp:96
#2  0x777a785f in QMenu::actionEvent (this=0x55ade210, e=)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/widgets/qmenu.cpp:3193
#3  0x7765f960 in QWidget::event (this=this@entry=0x55ade210, 
event=event@entry=0x7fffbc80)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:9042
#4  0x777adb8b in QMenu::event (this=0x55ade210, e=0x7fffbc80) 
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/widgets/qmenu.cpp:2650
#5  0x776104dc in QApplicationPrivate::notify_helper 
(this=this@entry=0x55a72090, receiver=receiver@entry=0x55ade210, 
e=e@entry=0x7fffbc80)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qapplication.cpp:3716
#6  0x77615eba in QApplication::notify (this=0x7fffd4c0, 
receiver=0x55ade210, e=0x7fffbc80)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qapplication.cpp:3499
#7  0x7440b9cd in QCoreApplication::notifyInternal 
(this=0x7fffd4c0, receiver=receiver@entry=0x55ade210, 
event=event@entry=0x7fffbc80)
at 
/var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qcoreapplication.cpp:963
#8  0x77655ff9 in sendEvent (event=0x7fffbc80, 
receiver=0x55ade210) at 
../../include/QtCore/../../../qtwidgets-5.5./src/corelib/kernel/qcoreapplication.h:228
#9  QWidget::removeAction (this=0x55ade210, 
action=action@entry=0x55c26fc0)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qwidget.cpp:3305
#10 0x77603b51 in QAction::~QAction (this=0x55c26fc0, 
__in_chrg=)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qaction.cpp:573
#11 0x77603cdf in QAction::~QAction (this=0x55c26fc0, 
__in_chrg=)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/kernel/qaction.cpp:592
#12 0x74441feb in QObjectPrivate::deleteChildren 
(this=this@entry=0x55cb6a10)
at 
/var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qobject.cpp:1951
#13 0x7444d0dc in QObject::~QObject (this=, 
__in_chrg=)
at 
/var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qobject.cpp:1031
#14 0x779fd62f in QSystemTrayIcon::~QSystemTrayIcon 
(this=0x55cc90a0, __in_chrg=)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5./src/widgets/util/qsystemtrayicon.cpp:150
#15 0x555c39ac in Gui::MainWindow::removeSysTray 
(this=this@entry=0x7fffd570) at 
/home/jkt/work/prog/trojita/src/Gui/Window.cpp:795
#16 0x555d5f88 in Gui::MainWindow::slotShowSettings 
(this=0x7fffd570) at /home/jkt/work/prog/trojita/src/Gui/Window.cpp:1150
#17 0x555e51ca in Gui::MainWindow::qt_static_metacall (_o=, _c=, _id=, _a=) at 
moc_Window.cpp:397
#18 0x74443bf8 in QMetaObject::activate 
(sender=sender@entry=0x55bc61a0, signalOffset=, 
local_signal_index=local_signal_index@entry=1, 
argv=argv@entry=0x7fffc110) at 
/var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qobject.cpp:3718
#19 0x7315 in QMetaObject::activate 
(sender=sender@entry=0x55bc61a0, m=m@entry=0x77b96040 
, local_signal_index=local_signal_index@entry=1, 
argv=argv@entry=0x7fffc110) at 
/var/tmp/portage/dev-qt/qtcore-5.5./work/qtcore-5.5./src/corelib/kernel/qobject.cpp:3583
#20 0x77603e1f in QAction::triggered (this=this@entry=0x55bc61a0, 
_t1=false) at .moc/moc_qaction.cpp:363
#21 0x77607357 in QAction::activate (this=0x55bc61a0, 
event=)
at 
/var/tmp/portage/dev-qt/qtwidgets-5.5./work/qtwidgets-5.5.

Review Request 122978: Add unit test for crash of bug 343976

2015-03-16 Thread Martin Gräßlin 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122978/
---

Review request for KDE Frameworks.


Bugs: 343976
https://bugs.kde.org/show_bug.cgi?id=343976


Repository: frameworkintegration


Description
---

Use a QPointer for QMenu of SystemTrayMenu

We need better memory management as the QMenu the SystemTrayMenu creates
might be deleted externally causing a double delete if the so-far
QScopedPoiter cleans up.

Switch to QPointer and call deleteLater if the pointer is still valid
during destroying the SystemTrayMenu.

BUG: 343976
CHANGELOG: Fix possible crash when destroying a QSystemTrayIcon (triggered by 
Trojita)


Diffs
-

  autotests/CMakeLists.txt e8ed6a99bb45969231ba21b8c5588e093bbbe3ee 
  autotests/ksni_unittest.cpp PRE-CREATION 
  src/platformtheme/kdeplatformsystemtrayicon.h 
8f13a36ae7cebed7499ce8a160f3533665b909ac 
  src/platformtheme/kdeplatformsystemtrayicon.cpp 
ce3d3de9f9cf479a5fe3af65ef7fb8ec2b9e0d29 

Diff: https://git.reviewboard.kde.org/r/122978/diff/


Testing
---


Thanks,

Martin Gräßlin

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122770: Support KWindowSystem::icon with NETWinInfo on all platforms

2015-03-16 Thread Martin Gräßlin 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122770/
---

(Updated March 16, 2015, 9:42 a.m.)


Status
--

This change has been marked as submitted.


Review request for KDE Frameworks.


Changes
---

Submitted with commit 3306777fe098720238c2e28992b4b21bdea25511 by Martin 
Gräßlin to branch master.


Repository: kwindowsystem


Description
---

If we compile with X11 support we want this method to use the X11
implementation on all platforms and not just on platform xcb. The
NETWinInfo is already so X11 specific that the platform doesn't
matter.

To solve this the method it delegates to in KWindowSystemPrivateX11
is made static. It only operates through the NETWinInfo and doesn't
need anything else from KWindowSystemPrivateX11.


Diffs
-

  src/kwindowsystem.h 21b254b246753d6ee7805864e28284dfa169855e 
  src/kwindowsystem.cpp 97dd6072a77f864b1e08287e8546cc29c24474c6 
  src/kwindowsystem_p_x11.h a507922ba632eb54e9121a1420d83c26d3676c66 

Diff: https://git.reviewboard.kde.org/r/122770/diff/


Testing
---

kwin_wayland (Martin dev branch) shows icons for X11 applications again.


Thanks,

Martin Gräßlin

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel

Re: Review Request 122800: Add a Wayland implementation for KWindowSystemPrivate

2015-03-16 Thread Martin Gräßlin 

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122800/
---

(Updated March 16, 2015, 9:42 a.m.)


Status
--

This change has been marked as submitted.


Review request for KDE Frameworks and Plasma.


Changes
---

Submitted with commit 593056f1b8f194714256d0790d73a71f27a6c4a0 by Martin 
Gräßlin to branch master.


Repository: kwindowsystem


Description
---

This is the same with KWindowSystemPrivateDummy, just that
::compositingActive returns true instead of false as the dummy
implementation does.

This ensures that usages depending on compositing state work
correctly. E.g. Plasma picks the right SVGs.

CHANGELOG: Add a dummy Wayland implementation for KWindowSystemPrivate


Diffs
-

  src/kwindowsystem_p_wayland.h PRE-CREATION 
  src/kwindowsystem_wayland.cpp PRE-CREATION 
  src/CMakeLists.txt ef792e7c78b79a12870c875682c5ceb668ad5550 
  src/kwindowsystem.cpp 97dd6072a77f864b1e08287e8546cc29c24474c6 
  src/kwindowsystem_p.h fb7d90c6a0e650d2a90589b1a9687d1ffd6f43e3 

Diff: https://git.reviewboard.kde.org/r/122800/diff/


Testing
---


File Attachments


Without: Plasma uses wrong SVG
  
https://git.reviewboard.kde.org/media/uploaded/files/2015/03/04/1584791d-47fa-49b5-b766-0fc81728ad68__wayland-plasma-no-compositing.png
With: Plasma uses right SVG (black corners might be a KWin bug)
  
https://git.reviewboard.kde.org/media/uploaded/files/2015/03/04/b5bf7fcf-6913-43af-8bcf-e0bc30766783__wayland-plasma-compositing.png


Thanks,

Martin Gräßlin

___
Kde-frameworks-devel mailing list
Kde-frameworks-devel@kde.org
https://mail.kde.org/mailman/listinfo/kde-frameworks-devel