Re: Review Request 124375: Fix potential endless recursion in PanelView::event() handler

2015-07-17 Thread David Edmundson

---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/124375/#review82604
---

Ship it!


Ship It!

- David Edmundson


On July 16, 2015, 1:48 p.m., Daniel Vrátil wrote:
 
 ---
 This is an automatically generated e-mail. To reply, visit:
 https://git.reviewboard.kde.org/r/124375/
 ---
 
 (Updated July 16, 2015, 1:48 p.m.)
 
 
 Review request for Plasma.
 
 
 Repository: plasma-workspace
 
 
 Description
 ---
 
 `PanelView::event()` generates new events for itself and dispatches them via 
 `QCoreApplication::sendEvent()`, which calls the handler directly. This can 
 lead to endless recursion. This patch changes the handler to use 
 `QCoreApplication::postEvent()` instead to enqueue the new event and dispatch 
 it from `QEventLoop` after the current `PanelView::event()` returns.
 
 Downstream bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1226644
 
 
 Diffs
 -
 
   shell/panelview.cpp b94673d 
 
 Diff: https://git.reviewboard.kde.org/r/124375/diff/
 
 
 Testing
 ---
 
 
 Thanks,
 
 Daniel Vrátil
 


___
Plasma-devel mailing list
Plasma-devel@kde.org
https://mail.kde.org/mailman/listinfo/plasma-devel


Re: Review Request 124375: Fix potential endless recursion in PanelView::event() handler

2015-07-17 Thread Daniel Vrátil


 On July 16, 2015, 4:34 p.m., David Edmundson wrote:
  Surely all this patch will do is change it from infinitely recursing to 
  100% CPU busy looping in each event loop?
  I don't see how that's much better?

With the event handling going through the event loop there is a high chance 
that you eventually move the mouse to a place that the code stops emitting the 
event further. With recursive event handling you get stack overflow and crash 
sooner or later. So yes, it's not perfect, but it solves a potential crash.


- Daniel


---
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/124375/#review82572
---


On July 16, 2015, 3:48 p.m., Daniel Vrátil wrote:
 
 ---
 This is an automatically generated e-mail. To reply, visit:
 https://git.reviewboard.kde.org/r/124375/
 ---
 
 (Updated July 16, 2015, 3:48 p.m.)
 
 
 Review request for Plasma.
 
 
 Repository: plasma-workspace
 
 
 Description
 ---
 
 `PanelView::event()` generates new events for itself and dispatches them via 
 `QCoreApplication::sendEvent()`, which calls the handler directly. This can 
 lead to endless recursion. This patch changes the handler to use 
 `QCoreApplication::postEvent()` instead to enqueue the new event and dispatch 
 it from `QEventLoop` after the current `PanelView::event()` returns.
 
 Downstream bug report: https://bugzilla.redhat.com/show_bug.cgi?id=1226644
 
 
 Diffs
 -
 
   shell/panelview.cpp b94673d 
 
 Diff: https://git.reviewboard.kde.org/r/124375/diff/
 
 
 Testing
 ---
 
 
 Thanks,
 
 Daniel Vrátil
 


___
Plasma-devel mailing list
Plasma-devel@kde.org
https://mail.kde.org/mailman/listinfo/plasma-devel


Re: baloo monitor visibility

2015-07-17 Thread Andrew Lake
It seems this desktop search monitor fits squarely into the powerful when
needed bucket. In that sense I don't think it is necessary to expose it
through any of the primary search interfaces (Alt-F2 or application menu).
That doesn't mean it should be completely hidden away since when it is
actually needed, it shouldn't be difficult to find.

I'm thinking perhaps we fold it into Search kcm as an indication of the
status of the search system. It could be helpful there if the user wants to
make a change to the search settings based on that information. I think we
should still do the plasmoid for the instances where the user might like to
do some long term monitoring. I'd suggest providing a way to access the
Search kcm from the plasmoid in case that user would like to make some
changes based on what they're seeing.

That's my 2c. Hope it helps,
Andrew

On Fri, Jul 17, 2015 at 1:33 AM Harald Sitter sit...@kde.org wrote:

 On Wed, Jul 15, 2015 at 10:39 AM, Pinak Ahuja pinak.ah...@gmail.com
 wrote:
  In fact I'm working on a plasmoid with the monitor's functionality, which
  will show up in the tray. Maybe we can hide the monitor from applications
  and add a button to plasmoid? i.e. if we do end up shipping the plasmoid.

 I think that would give it much more context, so that'd be cool.

 Mark and Martin do have a point in that it potentially being valuable
 without context though (such as when run from krunner by someone who
 knows what it does). I am not sure having it completely standalone is
 necessarily the way to go about this though. In particular I'd much
 rather have it be a KCM in ksysguard as it is a diagnostic tool for
 all intents and purposes (or kinfocenter... or fold it into the
 regular baloo KCM?).
 The one and only problem I have with the tool being in the menu is
 that it lacks any sort of context and at least to me file search is
 part of the workspace at large, so it being listed as a utility feels
 a lot like if for example the plasma wallpaper configuration was
 listed there.

 Perhaps you could ask the visual design group for some input on this?

 HS
 ___
 Plasma-devel mailing list
 Plasma-devel@kde.org
 https://mail.kde.org/mailman/listinfo/plasma-devel

___
Plasma-devel mailing list
Plasma-devel@kde.org
https://mail.kde.org/mailman/listinfo/plasma-devel