have we ever been hit by a Qt bug while developing the Octave GUI?

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

have we ever been hit by a Qt bug while developing the Octave GUI?

Carnë Draug
I am wondering if in the development of the Octave GUI we have been
hit by a Qt bug.  I don't follow the GUI related bugs with much
attention so I don't know but I don't remember seeing an Octave bug
that was actually a Qt bug.

So, have we ever been hit by a Qt bug while developing the Octave GUI?

Thanks
David

Reply | Threaded
Open this post in threaded view
|

Re: have we ever been hit by a Qt bug while developing the Octave GUI?

apjanke-floss

On 1/10/19 1:07 PM, Carnë Draug wrote:
> I am wondering if in the development of the Octave GUI we have been
> hit by a Qt bug.  I don't follow the GUI related bugs with much
> attention so I don't know but I don't remember seeing an Octave bug
> that was actually a Qt bug.
>
> So, have we ever been hit by a Qt bug while developing the Octave GUI?
>
> Thanks
> David

I've been hit by a minor one, but I don't know whether it's actually a
Qt bug or an underlying macOS bug.

https://github.com/octave-app/octave-app-bundler/issues/13

Sometimes, when first starting a Qt-enabled Octave on macOS, it issues
this scary-sounding warning in the command window:

 >> 2018-06-21 09:18 octave-gui[48030] (FSEvents.framework)
FSEventStreamFlushSync(): failed assertion '(SInt64)last_id > 0LL'

This is caused by Qt's QFilesystemWatcher doing an
FSEventStreamFlushSync(), which in some cases causes the underlying Mac
library to issue that warning. It's unclear to me whether this is a bug
in Qt's usage of the function, or the function itself.

I've been meaning to report this upstream to Qt, but haven't had time to
put together a minimal reproduction case yet.

Cheers,
Andrew

Reply | Threaded
Open this post in threaded view
|

Re: have we ever been hit by a Qt bug while developing the Octave GUI?

Daniel Sebald
In reply to this post by Carnë Draug
On 01/10/2019 01:07 PM, Carnë Draug wrote:

> I am wondering if in the development of the Octave GUI we have been
> hit by a Qt bug.  I don't follow the GUI related bugs with much
> attention so I don't know but I don't remember seeing an Octave bug
> that was actually a Qt bug.
>
> So, have we ever been hit by a Qt bug while developing the Octave GUI?
>
> Thanks
> David
>
>

Yes.  Early in 2018 a group of us tried to tackle an issue with floating
a subwindow which then froze on, was it, KDE?.  As a work-around there
was a message placed in an old Octave version describing how to use
control keys to move the window rather than the mouse.  This turned out
to be a mix of three or four issues, one of them being this Qt bug:

https://savannah.gnu.org/bugs/?53276#comment106
https://bugreports.qt.io/browse/QTBUG-44813
https://savannah.gnu.org/bugs/?53807

Furthermore, KDE Oxygen does not follow the convention that Qt calls out:

https://bugs.kde.org/show_bug.cgi?id=393076

KDE Oxygen is doing something a little strange in the fact they use Qt
to create an implementation of Qt.  That's possible to do, provide a
certain group of functions related to graphics are supplied.

In summary, that collection of Qt/KDE floated window issues was a major
effort--still with some workarounds until the bugs out of our control
are addressed.

Dan