[frameworks-ki18n] [Bug 341692] Need plural form format for real numbers

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=341692

Peter Wu  changed:

   What|Removed |Added

 Status|RESOLVED|REOPENED
 Ever confirmed|0   |1
 Resolution|WONTFIX |---
 CC||pe...@lekensteyn.nl

--- Comment #4 from Peter Wu  ---
The ki18n API appears to support real numbers (of the double type). I intended
to use it like this:

// double val;
setSuffix(i18np(" second", " seconds", val));

Corresponding autotests:

// Anything other than 1 is non-singular.
QCOMPARE(i18np("second", "seconds", 0.0),
 QString("seconds"));
QCOMPARE(i18np("second", "seconds", 1.0),
 QString("second"));
QCOMPARE(i18np("second", "seconds", 1.1),
 QString("seconds"));

The problem however is that the internal implementation only supports integer
numbers. Is it worth supporting this case? (feel free to close if you think it
is not, then the caller will just check val == 1.0 directly).

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 368939] Resize screenshot before copying/sending to another application

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=368939

--- Comment #3 from Peter Wu  ---
Hi-DPI can be detected per-screen, but maybe it is not a good idea to scale
down due to loss of details (hey, you *do* want to let your friends know of
your fancy screen right :p).

If you want to have the scaling option, do you have an idea how to add it
without making the interface too complex? Consider your workflows, what impact
would possible changes have?

The latest master has some other hi-dpi fixes btw, crop now works sanely and
the preview is not blurry anymore.

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 371017] Offer an option to upload to ipfs.pics

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=371017

Peter Wu  changed:

   What|Removed |Added

 CC||aleix...@gmail.com,
   ||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
The sharing option is part of the Purpose library, so adding Aleix for that. 

Aleix, do you think that this would be something to add?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 370945] 'Window under cursor' works more like 'all windows of an application'

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=370945

Peter Wu  changed:

   What|Removed |Added

 Status|UNCONFIRMED |CONFIRMED
 CC||pe...@lekensteyn.nl
 Ever confirmed|0   |1

--- Comment #2 from Peter Wu  ---
There is the "Capture the only pop-up" option that seems to do what you expect.
Changing the default value could probably be done, do you have suggestions for
improvement in this area?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 357022] Region selection screen does not take into account device pixel ratio

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=357022

--- Comment #17 from Peter Wu  ---
Git commit 1244980882382639f35925cef1d4cf3eeb3c240c by Peter Wu.
Committed on 20/10/2016 at 20:40.
Pushed by peterwu into branch 'master'.

Fix initial region location for High DPI screens

QML dimensions are measured in device-independent pixels, be sure to
make appropriate conversions. Also make sure that the new selection
size feature displays the correct dimensions.

Tested with: QT_SCALE_FACTOR=2 src/spectacle and dragging the region.

REVIEW: 129190

M  +1-1src/QuickEditor/EditorRoot.qml
M  +5-4src/QuickEditor/QuickEditor.cpp

http://commits.kde.org/spectacle/1244980882382639f35925cef1d4cf3eeb3c240c

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 358557] "Online services" menu is empty

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=358557

Peter Wu  changed:

   What|Removed |Added

  Latest Commit||http://commits.kde.org/spec
   ||tacle/47e1301926266c0597610
   ||9528671192468e35b9b
 Resolution|--- |FIXED
   Version Fixed In||16.12
 Status|REOPENED|RESOLVED

--- Comment #13 from Peter Wu  ---
Git commit 47e1301926266c05976109528671192468e35b9b by Peter Wu.
Committed on 20/10/2016 at 20:42.
Pushed by peterwu into branch 'master'.

Show message if kipi-plugins are not installed

If the kipi-plugins package is not installed as dependency of kipi (Arch
Linux) or if it failed to load due to an ABI mismatch, the Share menu
would be empty.

While at it, move setting of mKipiMenuLoaded to avoid a race condition
when the menu is opened too fast.

REVIEW: 129168
FIXED-IN: 16.12

M  +5-1src/Gui/ExportMenu.cpp

http://commits.kde.org/spectacle/47e1301926266c05976109528671192468e35b9b

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 357022] Region selection screen does not take into account device pixel ratio

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=357022

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #16 from Peter Wu  ---
I found one remaining edge case where the fullscreen picture is displayed at
the wrong location.

Consider a laptop screen on the left (primary) with an external, larger screen
on the right with resolutions 1920x1080 and 2560x1440 respectively and
QT_SCALE_FACTOR=2. The bottom sides are aligned such that you can move a
pointer between the bottoms without a jumping cursor. This means that the
origin of the laptop screen is no longer at (0, 0), but (0,360).

When the image is scaled by 2, the translated origin of the laptop screen
becomes (0,180). But Qt expects screen coordinates in native pixels and is
still painting from the (0,360), resulting in an upward shift of the image and
loss of the bottom half!

Adding a margin should work, but maybe you can think of other edge cases? (Two
different resolutions below each other? Others?)

(I have also reported the lack of documentation on the origin at
https://bugreports.qt.io/browse/QTBUG-56555)

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 363630] Crash: Spectacle hangs when using save as on a desktop where the default file dialog it gtk's

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=363630

--- Comment #5 from Peter Wu  ---
If it is reproducible, can you attach a gdb backtrace? Thanks :)

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 368524] Print Screen action opens KSnapshot instead of Spectacle

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=368524

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl
 Ever confirmed|0   |1
 Status|UNCONFIRMED |CONFIRMED
Summary|update "preset" shortcuts   |Print Screen action opens
   ||KSnapshot instead of
   ||Spectacle

--- Comment #1 from Peter Wu  ---
I remember having this issue when upgrading Plasma on Arch, but it was a matter
of changing the settings. It probably should have happened automatically.

Some other comments on this matter:
https://www.reddit.com/r/kde/comments/3yrv1e/plasma_55_default_printscreen_shortcut_with/

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 363630] Crash: Spectacle hangs when using save as on a desktop where the default file dialog it gtk's

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=363630

--- Comment #3 from Peter Wu  ---
Is this issue still reproducible or can the bug be closed?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 370529] "Active Window" area shows black box for Alt-Tab Compact switcher

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=370529

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
With the "Compact" switcher (System Settings -> Task Switcher -> Visualization
-> Compact), and "opening" the task switcher (via Alt-Tab), Spectacle 16.08
gives me the selected window in the list (not the compact switcher list).

If you experienced some missing pieces in the resulting screenshot, then it is
possibly fixed by bug 357223. Can you confirm this?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 370666] Core generated if no space left on disk

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=370666

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Is the core dump actually generated? If so, do you have a (gdb) backtrace?

What distro/version of Spectacle, Frameworks and Qt do you have?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 370303] Infinite recursion leading to stack overflow under XWayland

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=370303

Peter Wu  changed:

   What|Removed |Added

  Latest Commit||http://commits.kde.org/spec
   ||tacle/7a5603fda8f1e591d31db
   ||a1ad178c44ad6847cc2
   Version Fixed In||16.12
 Status|UNCONFIRMED |RESOLVED
 Resolution|--- |FIXED

--- Comment #1 from Peter Wu  ---
Git commit 7a5603fda8f1e591d31dba1ad178c44ad6847cc2 by Peter Wu.
Committed on 14/10/2016 at 09:08.
Pushed by peterwu into branch 'master'.

Refactor pixmap grabbing, add robustness

Do not go in an infinite loop when the root window cannot be captured
(bug 370303). Avoid crashing xcb_image_destroy by checking for NULL.

Split the getWindowPixmap into three logical methods depending on the
consumer: obtaining a pixmap from KWin, obtaining a pixmap from a single
window and obtaining a pixmap from a full screen. This enables some
optimizations, like obtaining a smaller area of the screen without
cropping. Additionally, it will make fixing HiDPI support easier.

Added robustness: if KWin somehow fails, now it will fallback to a
fullscreen capture instead of an null pixmap.

REVIEW: 129176
FIXED-IN: 16.12

M  +78   -30   src/PlatformBackends/X11ImageGrabber.cpp
M  +7-2src/PlatformBackends/X11ImageGrabber.h

http://commits.kde.org/spectacle/7a5603fda8f1e591d31dba1ad178c44ad6847cc2

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 361053] Center “click anywhere” message on screen instead of entire virtual desktop

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=361053

Peter Wu  changed:

   What|Removed |Added

 Status|CONFIRMED   |ASSIGNED
   Priority|VHI |NOR
   Assignee|m...@baloneygeek.com  |pe...@lekensteyn.nl
 CC||pe...@lekensteyn.nl

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 361053] Center “click anywhere” message on screen instead of entire virtual desktop

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=361053

Peter Wu  changed:

   What|Removed |Added

 CC||masonj...@gmail.com

--- Comment #1 from Peter Wu  ---
*** Bug 369287 has been marked as a duplicate of this bug. ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 369287] Help text for rectangular region mode appears in between dual monitors.

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=369287

Peter Wu  changed:

   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution|--- |DUPLICATE
 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---


*** This bug has been marked as a duplicate of bug 361053 ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 368939] Resize screenshot before copying/sending to another application

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=368939

Peter Wu  changed:

   What|Removed |Added

   Assignee|m...@baloneygeek.com  |pe...@lekensteyn.nl
 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Do you think that this scaling would be a useful default option? A preference
could then be added to override the behavior. Possible suggestion:

 [ ] Disable downscaling for HiDPI screens
 The default behavior for HiDPI screens is to save an image scaled down to a
nornal size. Enable this option to save the image in the original size instead.

Would this make sense?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 362450] Capturing small rectangular regions fails

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=362450

Peter Wu  changed:

   What|Removed |Added

 CC||mrb...@gmail.com

--- Comment #10 from Peter Wu  ---
*** Bug 365782 has been marked as a duplicate of this bug. ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 365782] Taken screenshot not always shown in dialog

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365782

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl
 Resolution|--- |DUPLICATE
 Status|UNCONFIRMED |RESOLVED

--- Comment #1 from Peter Wu  ---


*** This bug has been marked as a duplicate of bug 362450 ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 362061] Use popular shortcuts for Spectacle for sake of consistency with other environments

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=362061

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #5 from Peter Wu  ---
Print Screen currently open Spectacle.
Alt + Print Screen does nothing.
Shift + Print Screen currently saves the full screen to a file without showing
the dialog.

This does not really match the GNOME shortcuts, would you like to see it
differently?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 358557] "Online services" menu is empty

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=358557

Peter Wu  changed:

   What|Removed |Added

   Assignee|m...@baloneygeek.com  |pe...@lekensteyn.nl
 CC||pe...@lekensteyn.nl

--- Comment #12 from Peter Wu  ---
Would the message "No KIPI plugins available" be clear enough? Or should it
become "No KIPI plugins installed" (or something else)?

Proposed patch is at https://git.reviewboard.kde.org/r/129168/

@Boudhayan There is still activity in the libkipi and kipi-plugins repos, do
you still want to remove it (in favor of Purpose)?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 369330] 'Window under cursor' assume wrongly that dialogs are always in the upper left corner

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=369330

Peter Wu  changed:

   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
  Latest Commit||http://commits.kde.org/spec
   ||tacle/aab52c83ba70e9867632e
   ||83bd81715fc4405ffcf
 Resolution|--- |FIXED
   Version Fixed In||16.12

--- Comment #7 from Peter Wu  ---
Git commit aab52c83ba70e9867632e83bd81715fc4405ffcf by Peter Wu.
Committed on 13/10/2016 at 00:12.
Pushed by peterwu into branch 'master'.

Capture all windows for "Window under cursor"

Capture the correct area, not just the top-left one (bug 369330).
Capture all possible related windows, not just the top-level (or the
non-transient background window), this will correctly capture whole
applications including the open menus (bug 357223).

The last change also affects how windows with other overlapping ones are
captured, previously the hidden area is shown as if there was no popup.
For example, Kate would be captured and not its Open dialog.

Tested with Kate (Qt 5): Open File -> Add Places Entry -> Select Icon.
Verify that clicking all of these menus show the same screenshot and
that the non-overlapping parts are shown. Right-click to open a context
menu, this is also captured on the delayed screenshot.

Tested with heimdall-frontend and Hedgewars (Qt 4): open a menu, keep
the cursor on it. Verify that the menu and window are captured. Do the
same, but with the cursor outside the menu.

REVIEW: 129163
Related: bug 357223
FIXED-IN: 16.12

M  +37   -54   src/PlatformBackends/X11ImageGrabber.cpp
M  +1-2src/PlatformBackends/X11ImageGrabber.h

http://commits.kde.org/spectacle/aab52c83ba70e9867632e83bd81715fc4405ffcf

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 357223] spectacle fails to capture active window displaying a drop-down menu but xfce4-screenshooter succeeds

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=357223

Peter Wu  changed:

   What|Removed |Added

 Status|NEEDSINFO   |RESOLVED
  Latest Commit||http://commits.kde.org/spec
   ||tacle/aab52c83ba70e9867632e
   ||83bd81715fc4405ffcf
 Resolution|LATER   |FIXED
   Version Fixed In||16.12

--- Comment #15 from Peter Wu  ---
Git commit aab52c83ba70e9867632e83bd81715fc4405ffcf by Peter Wu.
Committed on 13/10/2016 at 00:12.
Pushed by peterwu into branch 'master'.

Capture all windows for "Window under cursor"

Capture the correct area, not just the top-left one (bug 369330).
Capture all possible related windows, not just the top-level (or the
non-transient background window), this will correctly capture whole
applications including the open menus (bug 357223).

The last change also affects how windows with other overlapping ones are
captured, previously the hidden area is shown as if there was no popup.
For example, Kate would be captured and not its Open dialog.

Tested with Kate (Qt 5): Open File -> Add Places Entry -> Select Icon.
Verify that clicking all of these menus show the same screenshot and
that the non-overlapping parts are shown. Right-click to open a context
menu, this is also captured on the delayed screenshot.

Tested with heimdall-frontend and Hedgewars (Qt 4): open a menu, keep
the cursor on it. Verify that the menu and window are captured. Do the
same, but with the cursor outside the menu.

REVIEW: 129163
Related: bug 369330
FIXED-IN: 16.12

M  +37   -54   src/PlatformBackends/X11ImageGrabber.cpp
M  +1-2src/PlatformBackends/X11ImageGrabber.h

http://commits.kde.org/spectacle/aab52c83ba70e9867632e83bd81715fc4405ffcf

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 365963] Try to take the screenshot ASAP on startup

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365963

Peter Wu  changed:

   What|Removed |Added

 Status|UNCONFIRMED |CONFIRMED
 Ever confirmed|0   |1
   Assignee|m...@baloneygeek.com  |pe...@lekensteyn.nl
 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Startup time should be halved (1100ms -> 500ms) with this patch:
https://git.reviewboard.kde.org/r/129167/

The picture should already have been taken before that though, is there any
evidence that the picture is taken too late? It would be interesting if we have
a tool that plays a fast timer which starts recording as soon as Print Screen
(or whatever start button) is pressed.

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 357022] Region selection screen does not take into account device pixel ratio

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=357022

Peter Wu  changed:

   What|Removed |Added

 CC||ishitatsuy...@gmail.com

--- Comment #15 from Peter Wu  ---
*** Bug 362653 has been marked as a duplicate of this bug. ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 362653] Incompatible with HiDPI settings

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=362653

Peter Wu  changed:

   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution|--- |DUPLICATE
 CC||pe...@lekensteyn.nl
   Assignee|m...@baloneygeek.com  |pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Can confirm, it is basically the same as the duplicate bug.

*** This bug has been marked as a duplicate of bug 357022 ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 370590] New: Spectacle opens itself on the wrong screen

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=370590

Bug ID: 370590
   Summary: Spectacle opens itself on the wrong screen
   Product: Spectacle
   Version: unspecified
  Platform: Archlinux Packages
OS: Linux
Status: UNCONFIRMED
  Severity: normal
  Priority: NOR
 Component: General
  Assignee: m...@baloneygeek.com
  Reporter: pe...@lekensteyn.nl

When Spectacle is started (Print Screen or via the program selection), it opens
on its last saved position. With multiple monitors, I would expect it on the
screen where my cursor is.

Reproducible: Always

Steps to Reproduce:
1. Start Spectacle

Actual Results:  
Spectacle opens on the other screen, I have to move the pointer to use it.

Expected Results:  
Spectacle should open in the screen where the pointer is located.

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 365657] Spactacle's selection mode works unreliable

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365657

Peter Wu  changed:

   What|Removed |Added

   Assignee|m...@baloneygeek.com  |pe...@lekensteyn.nl
 CC||pe...@lekensteyn.nl
 Resolution|--- |WAITINGFORINFO
 Status|UNCONFIRMED |NEEDSINFO

--- Comment #3 from Peter Wu  ---
The weird "window under cursor" issue is tracked by bug 357223 (a patch is in
progress).
(I can select and copy the version just fine (Arch Linux, KF 5.26, Qt 5.7.0).)

The default overlay is dark, have you changed this at Preferences? I cannot
reproduce the missing pan tool issue (with either bright/dark background
options). Can you still reproduce with the latest version in Arch?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 369174] Add image preview option for "Save As" dialog

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=369174

Peter Wu  changed:

   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution|--- |FIXED
   Version Fixed In||16.12
  Latest Commit||http://commits.kde.org/spec
   ||tacle/c9de3b0d09a41643b56e9
   ||62561f8187ac84e955e

--- Comment #1 from Peter Wu  ---
Git commit c9de3b0d09a41643b56e962561f8187ac84e955e by Peter Wu.
Committed on 12/10/2016 at 17:13.
Pushed by peterwu into branch 'master'.

Use native file dialog for the Save As option

The native file dialog has an image preview option which is lacking in
the widget-based Qt file dialog.

REVIEW: 129156
FIXED-IN: 16.12

M  +0-1src/ExportManager.cpp

http://commits.kde.org/spectacle/c9de3b0d09a41643b56e962561f8187ac84e955e

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 367907] spectacle segfault

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=367907

Peter Wu  changed:

   What|Removed |Added

 Resolution|WAITINGFORINFO  |WORKSFORME
 Status|NEEDSINFO   |RESOLVED

--- Comment #4 from Peter Wu  ---
Ok, hopefully some other component got fixed. I'll mark this as fixed, please
comment if you run into the same problem (and have a gdb backtrace).

Thanks!

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 363483] Problematic defaults for window decoration and default Save action

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=363483

Peter Wu  changed:

   What|Removed |Added

Summary|Problematic defaults, could |Problematic defaults for
   |be much better, thanks  |window decoration and
   ||default Save action
 Ever confirmed|0   |1
 Status|UNCONFIRMED |CONFIRMED
   Severity|normal  |wishlist
   Assignee|m...@baloneygeek.com  |pe...@lekensteyn.nl
 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Thank you for your feedback, some of your raised issues were already fixed
before in 16.04:
Remember save location: bug 357817
File overwrite confirmation: bug 357818

Note that the Save / Save As choice is remembered once you change the
selection, so if you change this option once you won't have to redo it.

A brief survey (n=2 :p) showed that people indeed seem to expect the window
decoration to be present by default. (This has technical implications, but this
should not influence the expected user behavior.)

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 363630] Crash: Spectacle hangs when using save as on a desktop where the default file dialog it gtk's

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=363630

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Do you have a gdb backtrace?

Can you point to the exact commit that fixed this for Krita? I see a bunch of
changes at
https://phabricator.kde.org/diffusion/KRITA/history/master/libs/widgetutils/KoFileDialog.cpp

Finally Krita removed the native dialog in
https://phabricator.kde.org/rKRITAcfc4efaf01cc17e795dfcac972563d4095cbd02f in
attempt to solve bug 364864 (which it did not according to the comments?)

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 363628] Usability bug: spectacle should by default ask for a filename on saving

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=363628

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #3 from Peter Wu  ---
Boudewijn, do you need additional changes? What is the behavior you expect?
If you had issues with the save location not being persisted (bug 357817), that
is fixed since 16.04.

Arne, I cannot reproduce your issue with 16.08.1, does it still occur after
restarting Spectacle?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 366282] I can't make a screen shot when spectacle window is opened

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=366282

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #4 from Peter Wu  ---
I cannot reproduce the issue with the steps in the original description. In
comment 2 you note a different problem where you press Print Screen after
having Spectacle open (bug 362616).

Could you clarify whether you are experiencing the first or second issue?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 368739] dimmed screen much too dark

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=368739

Peter Wu  changed:

   What|Removed |Added

 Ever confirmed|0   |1
 Status|UNCONFIRMED |CONFIRMED
 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
At Preferences (click the arrow next to the save button) you can select a
bright background. Does this help?

Personally I think that the dark overlay is too dark and that the bright
overlay is too bright (it is like a flashlight). Especially with the default
Plasma theme (Breeze) which has a dark cursor, the current choice is
suboptimal. Thoughts?

-- 
You are receiving this mail because:
You are watching all bug changes.

[kdebugsettings] [Bug 356700] Segmentation fault after saving to a file and pressing OK

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356700

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #2 from Peter Wu  ---
Could this be fixed by this fix in kbookmarks:
https://commits.kde.org/kbookmarks/68db76e9d8afce989d246768ebed7a12f9414b34

 I noticed that Kate and Spectacle would also crash when the file dialog was
opened once, breaking on QRegExp showed this:
#0  QRegExp::QRegExp (this=0x7fffa5d0, pattern="[/+]",
cs=Qt::CaseSensitive, syntax=QRegExp::RegExp) at tools/qregexp.cpp:3956
#1  0x7fffdbb8f646 in KBookmarkManager::findByAddress(QString const&) ()
from /usr/lib/libKF5Bookmarks.so.5
#2  0x7fffdc0bb82a in ?? () from /usr/lib/libKF5KIOFileWidgets.so.5
#3  0x7fffdc0bfab7 in KFilePlacesModel::Private::loadBookmarkList() () from
/usr/lib/libKF5KIOFileWidgets.so.5
#4  0x7fffdc0c0403 in KFilePlacesModel::Private::_k_reloadBookmarks() ()
from /usr/lib/libKF5KIOFileWidgets.so.5
#5  0x7fffdc0c1a3e in KFilePlacesModel::KFilePlacesModel(QObject*) () from
/usr/lib/libKF5KIOFileWidgets.so.5
#6  0x7fffdc0b706b in KFileWidget::KFileWidget(QUrl const&, QWidget*) ()
from /usr/lib/libKF5KIOFileWidgets.so.5

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 367907] spectacle segfault

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=367907

--- Comment #2 from Peter Wu  ---
The crash on exit issue after opening a file dialog is caused by a bug in
kbookmarks btw,
https://commits.kde.org/kbookmarks/68db76e9d8afce989d246768ebed7a12f9414b34

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 367907] spectacle segfault

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=367907

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl
 Resolution|--- |WAITINGFORINFO
 Status|UNCONFIRMED |NEEDSINFO

--- Comment #1 from Peter Wu  ---
Could you provide more details? What settings were you using? Have you done
other things like opening a file dialog prior to exiting?

(I managed to crash Kate and Spectacle after opening a file dialog once, but
that does not seem to be a bug in Spectacle.)

Can you also include what versions of Spectacle, Qt, Plasma and theme (like
Breeze) you have been using?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 365741] Rectangle selection screenshot: preview is missing on small sizes

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365741

Peter Wu  changed:

   What|Removed |Added

 CC||tipos...@tiscali.it

--- Comment #3 from Peter Wu  ---
*** Bug 369153 has been marked as a duplicate of this bug. ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 369153] No screenshot shown at times.

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=369153

Peter Wu  changed:

   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution|--- |DUPLICATE

--- Comment #2 from Peter Wu  ---
I'll assume that this bug is fixed by the other patch. If this is not the case,
please post a comment.

The problem does not occur with just a small region selection, but also under
these conditions:
 - the "Window under cursor" area option is in use
 - you select a transient window (e.g. a popup for some other window)
 - you have the "Capture the current pop-up only" checkbox disabled (this is
the default)

*** This bug has been marked as a duplicate of bug 365741 ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 362450] Capturing small rectangular regions fails

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=362450

Peter Wu  changed:

   What|Removed |Added

 CC||deve...@fuchsnet.ch

--- Comment #9 from Peter Wu  ---
*** Bug 365741 has been marked as a duplicate of this bug. ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 365741] Rectangle selection screenshot: preview is missing on small sizes

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365741

Peter Wu  changed:

   What|Removed |Added

 Status|UNCONFIRMED |RESOLVED
 Resolution|--- |DUPLICATE

--- Comment #2 from Peter Wu  ---


*** This bug has been marked as a duplicate of bug 362450 ***

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 362450] Capturing small rectangular regions fails

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=362450

Peter Wu  changed:

   What|Removed |Added

  Latest Commit||http://commits.kde.org/spec
   ||tacle/45751d7d8bc2a24abf756
   ||44d010018875c17daef
   Version Fixed In||16.12
 Resolution|--- |FIXED
 Status|UNCONFIRMED |RESOLVED

--- Comment #8 from Peter Wu  ---
Git commit 45751d7d8bc2a24abf75644d010018875c17daef by Peter Wu.
Committed on 11/10/2016 at 21:01.
Pushed by peterwu into branch 'master'.

Fix broken previews for small selection

The use of QGraphicsDropShadowEffect in KSImageWidget enables opacity.
When the previous format is RGB32, it can be converted directly to
ARGB32 with no data changes (assuming that the pixel data really has the
format 0xffRRGGBB).

Apparently the data returned by XCB has varying values (including zero)
for the alpha channel which violates the RGB32 format and results in
fully transparant pictures. Explicitly set the bits to fix this.

REVIEW: 129135
FIXED-IN: 16.12

M  +8-0src/PlatformBackends/X11ImageGrabber.cpp

http://commits.kde.org/spectacle/45751d7d8bc2a24abf75644d010018875c17daef

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 357080] Show area when capturing a rectangolar region

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=357080

Peter Wu  changed:

   What|Removed |Added

  Latest Commit||http://commits.kde.org/spec
   ||tacle/362cf1c7b5eca33c46de7
   ||c838776124f230e2eca
   Version Fixed In||16.12
 Status|UNCONFIRMED |RESOLVED
 Resolution|--- |FIXED

--- Comment #3 from Peter Wu  ---
Git commit 362cf1c7b5eca33c46de7c838776124f230e2eca by Peter Wu.
Committed on 11/10/2016 at 21:01.
Pushed by peterwu into branch 'master'.

Add selection size information

KSnapshot has this very nice feature where the selection size is
visible. This made it possible to use it as a measurement tool or create
screenshots with precise dimensions so restore this feature.

The 100x100 boundary was taken from KSnapshot while the positioning
policy, colors and boundaries were chosen by me.

Tested with light and dark background preference.

REVIEW: 129109
FIXED-IN: 16.12

M  +36   -0src/QuickEditor/EditorRoot.qml

http://commits.kde.org/spectacle/362cf1c7b5eca33c46de7c838776124f230e2eca

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 357223] spectacle fails to capture active window displaying a drop-down menu but xfce4-screenshooter succeeds

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=357223

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #14 from Peter Wu  ---
The reason why the menu is hidden in the Qt4 app is because this menu has the
WM_TRANSIENT_FOR property set (i.e. it is a transient "window"). The current
code does not correctly handle the case where a window has transient windows.

An attempt at fixing this:
https://github.com/Lekensteyn/spectacle/commit/cb4648021298c3eb1c775a111d680c83caa2

A possibly future patch could re-add the grabWindowUnderCursor fallback when it
can be sure that there are no child transient windows. That could however cause
some delay in obtaining the picture.

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 369153] No screenshot shown at times.

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=369153

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Does it depend on the selection size? This is possibly a duplicate of bug
365741

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 369330] 'Window under cursor' assume wrongly that dialogs are always in the upper left corner

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=369330

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #6 from Peter Wu  ---
Confirmed, when calculating the area, the translated window position was not
correctly taken into account. I've a local patch available (which depends on
other cleanup tasks). Will propose for review when the series is complete.

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 362241] Rectangular Region option does not preview and does not save an image after 15.12.3-2 -> 16.04.0-1

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=362241

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #16 from Peter Wu  ---
(In reply to Stuart K. Smith from comment #9)
> Rectanglar Region capture does not work below about 400x300 ish
> 
> Version 16.04.3 on KDEneon

Please note that bug 362450 tracks the issue of a blank preview for small
selections.

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 370303] New: Infinite recursion leading to stack overflow under XWayland

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=370303

Bug ID: 370303
   Summary: Infinite recursion leading to stack overflow under
XWayland
   Product: Spectacle
   Version: unspecified
  Platform: Compiled Sources
OS: Linux
Status: UNCONFIRMED
  Severity: normal
  Priority: NOR
 Component: General
  Assignee: m...@baloneygeek.com
  Reporter: pe...@lekensteyn.nl

While testing Spectacle in Weston+XWayland on top of Plasma (X11), it
segfaulted.

A stack trace showed recursive calls to getWindowPixmap with the same window
argument:

QPixmap X11ImageGrabber::getWindowPixmap(xcb_window_t window, bool
blendPointer)
{
...
if (xcbImage.isNull()) {
return getWindowPixmap(QX11Info::appRootWindow(), blendPointer)
.copy(geomReply->x, geomReply->y, geomReply->width,
geomReply->height);
}


Reproducible: Always

Steps to Reproduce:
1. Install xorg-server-xwayland and configure weston as described at
https://wiki.archlinux.org/index.php/wayland#XWayland
2. Launch weston
3. Inside weston, launch Spectacle

Actual Results:  
Segfault (stack overflow)

Expected Results:  
A screenshot (or failing to do so, an error message explaining why).

xorg-server 1.18.4-1
plasma-workspace 5.8.0-1
spectacle 16.08.1-1 (happens on this version and on git)

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 362450] Capturing small rectangular regions fails

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=362450

--- Comment #7 from Peter Wu  ---
9x340 is also visible, 9x339 (and some smaller) are not. What is special about
402x339? Looks like it is size of the preview area.

mPixmap.load("1.png") with the same pixmap that failed to load somehow DOES
show up in the preview. How does this work...? Strange Qt bug. (qt5-base
5.7.0-3)

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 362450] Capturing small rectangular regions fails

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=362450

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #6 from Peter Wu  ---
Can confirm this issue with Plasma 5.8 on Arch Linux (though the problem is
older). Using xf86-video-modesetting (xorg-server 1.18.4-1), mesa 12.0.3-2,
Linux v4.8 with i915, i7-6700HQ.

Anything with a width smaller than 403 (e.g. 402x21) results in an empty
preview (no idea if smaller heights cause issues, I could not make the height
smaller). Invoking this->pixmap()->save("1.png") in
KSImageWidget::setScreenshot does show the selected area though.

Previously reported as bug 359184, another duplicate is bug 365741.

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 359184] Preview not working for small screenshots

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=359184

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #6 from Peter Wu  ---
See also bug 362450

-- 
You are receiving this mail because:
You are watching all bug changes.

[Spectacle] [Bug 365741] Rectangle selection screenshot: preview is missing on small sizes

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=365741

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Duplicate of 362450

-- 
You are receiving this mail because:
You are watching all bug changes.

[plasmashell] [Bug 367685] Toggle application launcher on meta key

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=367685

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #4 from Peter Wu  ---
After an upgrade to Plasma 5.8 on Arch Linux, I noticed that pressing Meta
gives the Kickoff menu and indeed, pressing it again does not hide it.
Previously one could already use Alt+F1 to toggle the menu.

The new "Meta" shortcut gives:
method call time=1475665148.784542 sender=:1.18 ->
destination=org.kde.plasmashell serial=488 path=/PlasmaShell;
interface=org.kde.PlasmaShell; member=activateLauncherMenu

while the "Alt+F1" shortcut results in:
signal time=1475665160.428196 sender=:1.14 -> destination=(null destination)
serial=636 path=/component/plasmashell;
interface=org.kde.kglobalaccel.Component; member=globalShortcutPressed
   string "plasmashell"
   string "activate widget 196"
   int64 3600500

For some reason the Meta binding also tends to disable itself after some
activity. That is, after logging in, the Meta buttons shows the Kickoff menu.
After "some activity" (nothing special, I don't know the trigger), the Meta
button has no longer effect.

Personally I will probably be disabling this Meta shortcut since it is used as
modifier for many keybindings, just reporting it here in case it helps.

FYI, the commit adding this feature is:
https://quickgit.kde.org/?p=kwin.git=commitdiff=9b30cfef1d212ed2b40510a70d1f92c497f44398=eed4bf32ae04ff445d4c542366b44b7b12a915a5

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

--- Comment #15 from Peter Wu  ---
(In reply to Martin Sandsmark from comment #14)
> If I understand correctly, the test case is invalid:
> https://bugreports.qt.io/browse/QTBUG-48709

The linked issue is different, its root cause is heap-allocated QApplication
that is not deleted early enough. If I modify the ASAN testcase, then it will
indeed fail randomly:
make: Nothing to be done for 'first'.
ping

ASAN:DEADLYSIGNAL
=
==2657==ERROR: AddressSanitizer: SEGV on unknown address 0x7f5894fefaf9 (pc
0x7f5894fefaf9 bp 0x7f5892298d40 sp 0x7f5892298cf0 T1)
ASAN:DEADLYSIGNAL
AddressSanitizer: nested bug in the same thread, aborting.
TEST 1 FAILED

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

Peter Wu  changed:

   What|Removed |Added

 Attachment #100758|0   |1
is obsolete||

--- Comment #12 from Peter Wu  ---
Created attachment 100797
  --> https://bugs.kde.org/attachment.cgi?id=100797=edit
Testcase (ASAN) with normal QApplication::quit() and exit()

Extended the testcase to do normal QApplication::quit(), the proposed patches
would still result in a crash for that case. Better rip this delete hack
completely.

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

--- Comment #11 from Peter Wu  ---
breeze patch is at https://git.reviewboard.kde.org/r/128760/
oxygen patch is at https://git.reviewboard.kde.org/r/128761/

(sorry for the load of mails, just noting here for tracking purposes.)

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

Peter Wu  changed:

   What|Removed |Added

 CC||k...@davidedmundson.co.uk

--- Comment #10 from Peter Wu  ---
Adding David (Edmundson) since he did something similarin this area
(https://git.reviewboard.kde.org/r/122184/).

Can the title be changed to something like "Applications using QProxyStyle
crash on exit()".

Wireshark actually used setStyle(), passing an extended version of QProxyStyle
that follows this example: https://doc.qt.io/qt-5/qproxystyle.html#details.
With this patch, the crash no longer occurs:
https://git.reviewboard.kde.org/r/128760/

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

--- Comment #9 from Peter Wu  ---
Created attachment 100759
  --> https://bugs.kde.org/attachment.cgi?id=100759=edit
debugging patch

Adding David since he introduced the fix that is referenced in the breeze code:

commit 2ffe20e1bfe93c921c5372b4d21447b1de308d4b
Author: David Faure 
Date:   Sat Jun 7 14:57:29 2014 +0200

Fix crash on exit in all QCommandLineParser-based programs.

Example: kioclient5 ls  (= a syntax error).

The issue is that ::exit(1) is called, so ~QGuiApplication isn't called
(so the style isn't deleted), but global static objects are deleted,
which deletes the style-plugin-factory, which unloads the style plugin.

The crash happened because "AppEventFilter" in oxygen would still be
installed as an app event filter, but the plugin was unloaded, so any
calls to it would crash.

If I disable the delete hack or avoid calling delete for the second attempt
(likely wrong approach...), then the crash is gone. Was this a workaround for a
specific Qt version?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

--- Comment #8 from Peter Wu  ---
Sigh... wiped breeze before testing it, here is the stacktrace including
breeze:

...
#17 0x7fe4c3f9788f in QObject::~QObject() kernel/qobject.cpp:1048
#18 0x7fe4ab14d7d4 in Breeze::Style::~Style() ../kstyle/breezestyle.cpp:199
#19 0x7fe4ab14d7ef in Breeze::Style::~Style() ../kstyle/breezestyle.cpp:203
#20 0x7fe4ab1913ef in operator() ../kstyle/breezestyleplugin.cpp:45
#21 0x7fe4ab1919b5 in call /usr/include/qt/QtCore/qobjectdefs_impl.h:501
#22 0x7fe4ab191996 in call, void>
/usr/include/qt/QtCore/qobjectdefs_impl.h:558
#23 0x7fe4ab191927 in impl /usr/include/qt/QtCore/qobject_impl.h:198
#24 0x7fe4c3f8559a in QtPrivate::QSlotObjectBase::call(QObject*, void**)
src/corelib/kernel/qobject_impl.h:130
#25 0x7fe4c3f8559a in QMetaObject::activate(QObject*, int, int, void**)
kernel/qobject.cpp:3723
#26 0x7fe4c3f86323 in QObject::destroyed(QObject*) .moc/moc_qobject.cpp:213
#27 0x7fe4c3f97a18 in QObject::~QObject() kernel/qobject.cpp:920
#28 0x7fe4ab1915dd in Breeze::StylePlugin::~StylePlugin()
../kstyle/breezestyleplugin.cpp:54
#29 0x7fe4ab1915f9 in Breeze::StylePlugin::~StylePlugin()
../kstyle/breezestyleplugin.cpp:57
#30 0x7fe4c3f049f6 in QLibraryPrivate::unload(QLibraryPrivate::UnloadFlag)
plugin/qlibrary.cpp:557
#31 0x7fe4c3eedea0 in QFactoryLoaderPrivate::~QFactoryLoaderPrivate()
plugin/qfactoryloader.cpp:86
#32 0x7fe4c3eee1a0 in QFactoryLoaderPrivate::~QFactoryLoaderPrivate()
plugin/qfactoryloader.cpp:89
#33 0x7fe4c3f978e9 in
QScopedPointerDeleter::cleanup(QObjectData*)
src/corelib/tools/qscopedpointer.h:60
#34 0x7fe4c3f978e9 in QScopedPointer::~QScopedPointer()
src/corelib/tools/qscopedpointer.h:107
#35 0x7fe4c3f978e9 in QObject::~QObject() kernel/qobject.cpp:900
#36 0x7fe4c3eec2e3 in QFactoryLoader::~QFactoryLoader()
plugin/qfactoryloader.cpp:205
#37 0x7fe4c4889bf8 in ~Holder styles/qstylefactory.cpp:72
#38 0x7fe4c308a98f in __run_exit_handlers (/usr/lib/libc.so.6+0x3598f)
#39 0x7fe4c308a9e9 in __GI_exit (/usr/lib/libc.so.6+0x359e9)
#40 0x403337 in MainWindow::ping() (Trial+0x403337)
#41 0x7fe4c3f84fb6 in QMetaObject::activate(QObject*, int, int, void**)
kernel/qobject.cpp:3740
#42 0x7fe4c3fa65b3 in QTimer::timerEvent(QTimerEvent*)
kernel/qtimer.cpp:254
#43 0x7fe4c3f8770b in QObject::event(QEvent*) kernel/qobject.cpp:1285
#44 0x7fe4c473e417 in QApplicationPrivate::notify_helper(QObject*, QEvent*)
kernel/qapplication.cpp:3799
#45 0x7fe4c4750b8d in QApplication::notify(QObject*, QEvent*)
kernel/qapplication.cpp:3641
#46 0x7fe4c3f180a1 in QCoreApplication::notifyInternal2(QObject*, QEvent*)
kernel/qcoreapplication.cpp:988
#47 0x7fe4c3fef9b4 in QCoreApplication::sendEvent(QObject*, QEvent*)
src/corelib/kernel/qcoreapplication.h:231
#48 0x7fe4c3fef9b4 in QTimerInfoList::activateTimers()
kernel/qtimerinfo_unix.cpp:644
#49 0x7fe4c3ff0ac2 in timerSourceDispatch
kernel/qeventdispatcher_glib.cpp:182
#50 0x7fe4c0a18dd6 in g_main_context_dispatch
(/usr/lib/libglib-2.0.so.0+0x49dd6)
#51 0x7fe4c0a1903f  (/usr/lib/libglib-2.0.so.0+0x4a03f)
#52 0x7fe4c0a190eb in g_main_context_iteration
(/usr/lib/libglib-2.0.so.0+0x4a0eb)
#53 0x7fe4c3ff19b1 in
QEventDispatcherGlib::processEvents(QFlags)
kernel/qeventdispatcher_glib.cpp:425
#54 0x7fe4c3f146a4 in
QEventLoop::exec(QFlags)
kernel/qeventloop.cpp:210
#55 0x7fe4c3f26d1e in QCoreApplication::exec()
kernel/qcoreapplication.cpp:1261
#56 0x402ff6 in main (Trial+0x402ff6)
#57 0x7fe4c3075290 in __libc_start_main (/usr/lib/libc.so.6+0x20290)
#58 0x403179 in _start (Trial+0x403179)

0x60205930 is located 0 bytes inside of 16-byte region
[0x60205930,0x60205940)
freed by thread T0 here:
#0 0x7fe4c56415d0 in operator delete(void*)
/build/gcc-multilib/src/gcc/libsanitizer/asan/asan_new_delete.cc:92
#1 0x7fe4ab109cc6 in QtSharedPointer::ExternalRefCountData::operator
delete(void*) /usr/include/qt/QtCore/qsharedpointer_impl.h:171
#2 0x7fe4ab10a7de in QWeakPointer::~QWeakPointer()
/usr/include/qt/QtCore/qsharedpointer_impl.h:607
#3 0x7fe4ab114545 in
QWeakPointer::operator=(QWeakPointer&&)
/usr/include/qt/QtCore/qsharedpointer_impl.h:634
#4 0x7fe4ab11394b in QWeakPointer&
QWeakPointer::assign(QObject*)
/usr/include/qt/QtCore/qsharedpointer_impl.h:719
#5 0x7fe4ab1b3ccc in QPointer::operator=(QObject*)
/usr/include/qt/QtCore/qpointer.h:83
#6 0x7fe4ab1ab364 in qt_plugin_instance
kstyle/moc_breezestyleplugin.cpp:165
#7 0x7fe4c3eeca02 in QFactoryLoader::instance(int) const
plugin/qfactoryloader.cpp:283
#8 0x7fe4c4883e87 in QStyle* qLoadPlugin(QFactoryLoader const*, QString const&)
src/corelib/plugin/qfactoryloader_p.h:101
#9 0x7fe4c4883e87 in QStyleFactory::create(QString const&)
styles/qstylefactory.cpp:158
  

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

Peter Wu  changed:

   What|Removed |Added

 CC||fa...@kde.org

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

--- Comment #7 from Peter Wu  ---
Created attachment 100758
  --> https://bugs.kde.org/attachment.cgi?id=100758=edit
Testcase (with ASAN)

I bisected Wireshark and found that it started crashing after using
setStyle(new QProxyStyle). Fair enough, I can create a minimal testcase that
crashes when setStyle(new QProxyStyle) is used and exit() is called.

This problem does not occur when -style windows (or anything other than breeze
and oxygen) is in use.

Following test was done with:
qt5-base 5.7.0-2
breeze v5.7.4-17-gf79266d (also reproduced with 5.7.3-1 and many releases
before that...)
oxygen 5.7.3-1 (not in output, but also affected)

ASAN output:
=
==3262==ERROR: AddressSanitizer: heap-use-after-free on address 0x60205930
at pc 0x7f5644140df8 bp 0x7ffd4e1b4c90 sp 0x7ffd4e1b4c80
WRITE of size 4 at 0x60205930 thread T0
#0 0x7f5644140df7 in std::__atomic_base::operator--()
/usr/include/c++/6.1.1/bits/atomic_base.h:304
#1 0x7f5644140df7 in bool QAtomicOps::deref(std::atomic&)
src/corelib/arch/qatomic_cxx11.h:147
#2 0x7f5644140df7 in QBasicAtomicInteger::deref()
src/corelib/thread/qbasicatomic.h:111
#3 0x7f5644140df7 in QWeakPointer::~QWeakPointer()
src/corelib/tools/qsharedpointer_impl.h:607
#4 0x7f5644140df7 in
QWeakPointer::operator=(QWeakPointer&&)
src/corelib/tools/qsharedpointer_impl.h:634
#5 0x7f5644140df7 in QWeakPointer&
QWeakPointer::assign(QObject*)
src/corelib/tools/qsharedpointer_impl.h:719
#6 0x7f5644140df7 in QPointer::operator=(QObject*)
src/corelib/kernel/qpointer.h:83
#7 0x7f5644140df7 in QFactoryLoader::instance(int) const
plugin/qfactoryloader.cpp:283
#8 0x7f5644ad7e87 in QStyle* qLoadPlugin(QFactoryLoader const*, QString const&)
src/corelib/plugin/qfactoryloader_p.h:101
#9 0x7f5644ad7e87 in QStyleFactory::create(QString const&)
styles/qstylefactory.cpp:158
#10 0x7f5644b3351b in QProxyStylePrivate::ensureBaseStyle() const
styles/qproxystyle.cpp:99
#11 0x7f5644b3591d in QProxyStyle::event(QEvent*)
styles/qproxystyle.cpp:386
#12 0x7f5644992417 in QApplicationPrivate::notify_helper(QObject*, QEvent*)
kernel/qapplication.cpp:3799
#13 0x7f56449a4b8d in QApplication::notify(QObject*, QEvent*)
kernel/qapplication.cpp:3641
#14 0x7f564416c0a1 in QCoreApplication::notifyInternal2(QObject*, QEvent*)
kernel/qcoreapplication.cpp:988
#15 0x7f56441ea3e1 in QCoreApplication::sendEvent(QObject*, QEvent*)
src/corelib/kernel/qcoreapplication.h:231
#16 0x7f56441ea3e1 in QObjectPrivate::setParent_helper(QObject*)
kernel/qobject.cpp:1996
#17 0x7f56441eb88f in QObject::~QObject() kernel/qobject.cpp:1048
#18 0x7f562b413788  (/usr/lib/qt/plugins/styles/breeze.so+0x53788)
#19 0x7f56441d959a in QtPrivate::QSlotObjectBase::call(QObject*, void**)
src/corelib/kernel/qobject_impl.h:130
#20 0x7f56441d959a in QMetaObject::activate(QObject*, int, int, void**)
kernel/qobject.cpp:3723
#21 0x7f56441da323 in QObject::destroyed(QObject*) .moc/moc_qobject.cpp:213
#22 0x7f56441eba18 in QObject::~QObject() kernel/qobject.cpp:920
#23 0x7f562b427416  (/usr/lib/qt/plugins/styles/breeze.so+0x67416)
#24 0x7f56441589f6 in QLibraryPrivate::unload(QLibraryPrivate::UnloadFlag)
plugin/qlibrary.cpp:557
#25 0x7f5644141ea0 in QFactoryLoaderPrivate::~QFactoryLoaderPrivate()
plugin/qfactoryloader.cpp:86
#26 0x7f56441421a0 in QFactoryLoaderPrivate::~QFactoryLoaderPrivate()
plugin/qfactoryloader.cpp:89
#27 0x7f56441eb8e9 in
QScopedPointerDeleter::cleanup(QObjectData*)
src/corelib/tools/qscopedpointer.h:60
#28 0x7f56441eb8e9 in QScopedPointer::~QScopedPointer()
src/corelib/tools/qscopedpointer.h:107
#29 0x7f56441eb8e9 in QObject::~QObject() kernel/qobject.cpp:900
#30 0x7f56441402e3 in QFactoryLoader::~QFactoryLoader()
plugin/qfactoryloader.cpp:205
#31 0x7f5644addbf8 in ~Holder styles/qstylefactory.cpp:72
#32 0x7f56432de98f in __run_exit_handlers (/usr/lib/libc.so.6+0x3598f)
#33 0x7f56432de9e9 in __GI_exit (/usr/lib/libc.so.6+0x359e9)
#34 0x403337 in MainWindow::ping() (Trial+0x403337)
#35 0x7f56441d8fb6 in QMetaObject::activate(QObject*, int, int, void**)
kernel/qobject.cpp:3740
#36 0x7f56441fa5b3 in QTimer::timerEvent(QTimerEvent*)
kernel/qtimer.cpp:254
#37 0x7f56441db70b in QObject::event(QEvent*) kernel/qobject.cpp:1285
#38 0x7f5644992417 in QApplicationPrivate::notify_helper(QObject*, QEvent*)
kernel/qapplication.cpp:3799
#39 0x7f56449a4b8d in QApplication::notify(QObject*, QEvent*)
kernel/qapplication.cpp:3641
#40 0x7f564416c0a1 in QCoreApplication::notifyInternal2(QObject*, QEvent*)
kernel/qcoreapplication.cpp:988
#41 0x7f56442439b4 in QCoreApplication::sendEvent(QObject*, QEvent*)
src/corelib/kernel/qcoreapplication.h:231
#42 0x7f56442439b4 in QTimerInfoList::activateTimers()

[plasmashell] [Bug 356132] Starting KDE causes with power widget causes screen brightness to go to 0

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356132

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #5 from Peter Wu  ---
This was not fun debugging, jumping from kdeinit5/powerdevil via dbus to
plasmashell to a widget. Question: have you expanded the battery monitor
widget? That seems to trigger the bad condition in plasma-workspace 5.6.4.

Can you try this patch: https://git.reviewboard.kde.org/r/128098/ (and check
also if the analysis is sane)

-- 
You are receiving this mail because:
You are watching all bug changes.

[konsole] [Bug 361038] Terminal goes into no-echo mode after killing some programs with Ctrl+C

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=361038

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #1 from Peter Wu  ---
Cannot reproduce on Arch Linux with:
Qt: 5.6.0
KDE Frameworks: 5.21.0
Konsole: 16.04.0

After ^C, input is echoed as expected.

If you are still affected, have you tried the "stty echo" command to restore
echo?

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356940] kalarm crash on start up

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356940

Peter Wu  changed:

   What|Removed |Added

 CC||pe...@lekensteyn.nl

--- Comment #6 from Peter Wu  ---
With Breeze 5.6.1 + f56b8b9fbe97f83cf18523dde3a1f83fada9beab, I sitll manage to
crash an application on exit. To reproduce, start "wireshark -O foo" (this is
an invalid option, resulting in exit() after the QApplication is constructed,
but before exec() is called.

(Qt version 5.6.0, Wireshark from git, but also reproduced with any stable 2.0
version)

Backtrace:

#0 0x7fb493f8c01d in bool operator==(QPair const&, QPair const&)
src/corelib/tools/qpair.h:100
#1 0x7fb493f8c01d in QHash,
QPropertyAnimation*>::remove(QPair const&)
src/corelib/tools/qhash.h:779
#2 0x7fb493f8b6a5 in
QPropertyAnimation::updateState(QAbstractAnimation::State,
QAbstractAnimation::State) src/corelib/animation/qpropertyanimation.cpp:297
#3 0x7fb493f837e8 in
QAbstractAnimationPrivate::setState(QAbstractAnimation::State)
src/corelib/animation/qabstractanimation.cpp:990
#4 0x7fb493f837e8 in QAbstractAnimation::stop()
src/corelib/animation/qabstractanimation.cpp:1379
#5 0x7fb493f8a326 in QPropertyAnimation::~QPropertyAnimation()
src/corelib/animation/qpropertyanimation.cpp:169
#6 0x7fb479d7d712 in Breeze::Animation::~Animation()
/build/src/breeze-5.6.1/kstyle/animations/breezeanimation.h:47
#7 0x7fb479d7d712 in Breeze::Animation::~Animation()
/build/src/breeze-5.6.1/kstyle/animations/breezeanimation.h:47
#8 0x7fb4941adf52 in QObjectPrivate::deleteChildren()
src/corelib/kernel/qobject.cpp:1963
#9 0x7fb4941b7523 in QObject::~QObject()
src/corelib/kernel/qobject.cpp:1034
#10 0x7fb479d7e1d2 in Breeze::WidgetStateData::~WidgetStateData()
/build/src/breeze-5.6.1/kstyle/animations/breezewidgetstatedata.h:45
#11 0x7fb479d7e1d2 in Breeze::EnableData::~EnableData()
/build/src/breeze-5.6.1/kstyle/animations/breezeenabledata.h:43
#12 0x7fb479d7e1d2 in Breeze::EnableData::~EnableData()
/build/src/breeze-5.6.1/kstyle/animations/breezeenabledata.h:43
#13 0x7fb4941adf52 in QObjectPrivate::deleteChildren()
src/corelib/kernel/qobject.cpp:1963
#14 0x7fb4941b7523 in QObject::~QObject()
src/corelib/kernel/qobject.cpp:1034
#15 0x7fb479d7ee38 in Breeze::WidgetStateEngine::~WidgetStateEngine()
/build/src/breeze-5.6.1/kstyle/animations/breezewidgetstateengine.h:46
#16 0x7fb4941adf52 in QObjectPrivate::deleteChildren()
src/corelib/kernel/qobject.cpp:1963
#17 0x7fb4941b7523 in QObject::~QObject()
src/corelib/kernel/qobject.cpp:1034
#18 0x7fb479d7e98f in Breeze::Animations::~Animations()
/build/src/breeze-5.6.1/kstyle/animations/breezeanimations.h:52
#19 0x7fb479d7e98f in Breeze::Animations::~Animations()
/build/src/breeze-5.6.1/kstyle/animations/breezeanimations.h:52
#20 0x7fb4941adf52 in QObjectPrivate::deleteChildren()
src/corelib/kernel/qobject.cpp:1963
#21 0x7fb4941b7523 in QObject::~QObject()
src/corelib/kernel/qobject.cpp:1034
#22 0x7fb479d5cae8 in _init
/build/src/breeze-5.6.1/kstyle/breezestyle.cpp:203
#23 0x7fb4941afbad in QtPrivate::QSlotObjectBase::call(QObject*, void**)
src/corelib/kernel/qobject_impl.h:124
#24 0x7fb4941afbad in QMetaObject::activate(QObject*, int, int, void**)
src/corelib/kernel/qobject.cpp:3715
#25 0x7fb4941b01ee in QObject::destroyed(QObject*)
src/corelib/.moc/moc_qobject.cpp:213
#26 0x7fb4941b7184 in QObject::~QObject()
src/corelib/kernel/qobject.cpp:913
#27 0x7fb479d706e6 in Breeze::StylePlugin::~StylePlugin()
/build/src/breeze-5.6.1/kstyle/breezestyleplugin.cpp:52
#28 0x7fb479d706e6 in _init
/build/src/breeze-5.6.1/kstyle/breezestyleplugin.cpp:54
#29 0x7fb494178ad8 in QLibraryPrivate::unload(QLibraryPrivate::UnloadFlag)
src/corelib/plugin/qlibrary.cpp:551
#30 0x7fb49416eaa1 in QFactoryLoaderPrivate::~QFactoryLoaderPrivate()
src/corelib/plugin/qfactoryloader.cpp:88
#31 0x7fb49416ebd8 in QFactoryLoaderPrivate::~QFactoryLoaderPrivate()
src/corelib/plugin/qfactoryloader.cpp:91
#32 0x7fb4941b756b in
QScopedPointerDeleter::cleanup(QObjectData*)
src/corelib/tools/qscopedpointer.h:54
#33 0x7fb4941b756b in QScopedPointer::~QScopedPointer()
src/corelib/tools/qscopedpointer.h:101
#34 0x7fb4941b756b in QObject::~QObject()
src/corelib/kernel/qobject.cpp:893
#35 0x7fb49416dc99 in QFactoryLoader::~QFactoryLoader()
src/corelib/plugin/qfactoryloader.cpp:226
#36 0x7fb494f131e8 in ~Holder src/widgets/styles/qstylefactory.cpp:70
#37 0x7fb492673c37 in __run_exit_handlers (/usr/lib/libc.so.6+0x35c37)
#38 0x7fb492673c84 in __GI_exit (/usr/lib/libc.so.6+0x35c84)
#39 0x5613417cca8b in main 

[Breeze] [Bug 356561] New: [REGRESSION] [bisected] Rendering delays, slowdown

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356561

Bug ID: 356561
   Summary: [REGRESSION] [bisected] Rendering delays, slowdown
   Product: Breeze
   Version: 5.5.0
  Platform: Archlinux Packages
OS: Linux
Status: UNCONFIRMED
  Severity: normal
  Priority: NOR
 Component: QStyle
  Assignee: hugo.pere...@free.fr
  Reporter: pe...@lekensteyn.nl

Since upgrading from breeze 5.4.3-2 to 5.5.0-2 on Arch Linux, starting
Wireshark is 500% slower (5 seconds instead of 1).

Bisection points to 6d852f30a1f2c1988359d4e0cdb21e2f1714a6bd (bug 344425).
Reverting this on 5.5.0 (or master, c327915) restores the previous sane
behavior.

Reproducible: Always

Steps to Reproduce:
1. Install wireshark-qt (which uses Qt5) on Arch Linux
Start it with:

time echo 'os.exit()' | HOME=/tmp/wshome2 /usr/bin/wireshark -X
lua_script:/dev/stdin

Actual Results:  
The screen stays transparant for some seconds, then finally shows the progress
bar and then exits after five seconds.

Expected Results:  
It should immediately show the progress bar and then exit in one second.

Relevant package versions:
breeze5.5.0-2
wireshark-qt  2.0.0-2
qt5-base  5.5.1-5

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356561] [REGRESSION] [bisected] Rendering delays, slowdown

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356561

--- Comment #1 from Peter Wu  ---
Forgot to mention that this does not just impact the startup, it also affects
runtime. When scrolling through the packet list, the packet detail view is
updated after a delay (the packet byte view and packet list view instead update
immediately).

-- 
You are receiving this mail because:
You are watching all bug changes.

[Breeze] [Bug 356561] [REGRESSION] [bisected] Rendering delays, slowdown

[Peter Wu via KDE Bugzilla]

https://bugs.kde.org/show_bug.cgi?id=356561

--- Comment #3 from Peter Wu  ---
I cannot reproduce the persistent from-behind rendering as shown on the
screenshot, but I do notice a delay in rendering (the konsole window appears
with window decorations, but the contents of the window is empty for a second,
showing the image of the programs behind it). My Konsole theme has opacity
enabled, I don't know if that matters here.

Additional hardware/software details:
integrated Intel GPU from i5-560M using AccelMethod uxa (not sna)
xf86-video-intel 1:2.99.917+478+gdf72bc5-3
linux 4.3.0

-- 
You are receiving this mail because:
You are watching all bug changes.