Review Request 119946: Wrap the desktop alternatives UI in a dialog

[Aaron J. Seigo]

> On Aug. 26, 2014, 5:49 p.m., David Edmundson wrote:
> > desktoppackage/contents/explorer/AppletAlternatives.qml, line 34
> > <https://git.reviewboard.kde.org/r/119946/diff/1/?file=307712#file307712line34>
> >
> >     Why redeclare the property?
> 
> Marco Martin wrote:
>     ah, yes, didn't see it, it shouldn't redefine it

indeed; a relic from a previous experiment.


> On Aug. 26, 2014, 5:49 p.m., David Edmundson wrote:
> > desktoppackage/contents/explorer/AppletAlternatives.qml, line 131
> > <https://git.reviewboard.kde.org/r/119946/diff/1/?file=307712#file307712line131>
> >
> >     Why do we need this?
> >     (and if we do need it, it should go on the Switch button too)
> 
> David Edmundson wrote:
>     Edit: I see why we're using it now. Should have read the other patch first.
>     It does need adding to the other button, but this whole thing isn't ideal. It means we have to rely on the Alternatives.qml having magic undocumented properties.
> 
> Marco Martin wrote:
>     the imperative show() and close() don't seem to be defined anywhere, it should just do visible = true; and visible = false; ?
> 
> David Edmundson wrote:
>     Personally I'd rather use an explicit dialog.destroy() rather than having a property that we monitor in code further away. 
>     
>     the code in the C++ then becomes
>     connect(qmlObj->rootObject(), SIGNAL(destroyed(QObject*)),
>             qmlObj, SLOT(deleteLater()));
>             
>     and we can drop the d->alternativesObjects, and the alternativesVisibilityChanged.
> 
> Marco Martin wrote:
>     ah, show()/hide() are from QWindow, nvm (if visible is not redeclared they shouldn't be required tough)
> 
> Marco Martin wrote:
>     "Personally I'd rather use an explicit dialog.destroy() rather than having a property that we monitor in code further away."
>     
>     the problem is that you can't use destroy() from javascript on object that have been declared, you can only for objects that have been dynamically created from js as well.
>     If you try to destroy from javascript the root object, destroy is only a noop and only gives a warning, so as far I know that one can be killed only from c++ unfortunately
> 
> David Edmundson wrote:
>     Fair point.
>     In that case I guess this is probably among the best solution we can find.
>     
>     Fix my first comment, and this is a ship it! from me.

also from the same old experiment.


- Aaron J.


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/119946/#review65293
-----------------------------------------------------------


On Aug. 26, 2014, 5:37 p.m., Aaron J. Seigo wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/119946/
> -----------------------------------------------------------
> 
> (Updated Aug. 26, 2014, 5:37 p.m.)
> 
> 
> Review request for Plasma.
> 
> 
> Repository: plasma-desktop
> 
> 
> Description
> -------
> 
> Following up on the changes to plasmashell, the dialog needs to used from the QML in the shell package.
> 
> 
> Diffs
> -----
> 
>   desktoppackage/contents/explorer/AppletAlternatives.qml de99b1ca370c819687230c1bcd54d2839b08dab9 
> 
> Diff: https://git.reviewboard.kde.org/r/119946/diff/
> 
> 
> Testing
> -------
> 
> Used Plasma Desktop with this shell package.
> 
> 
> Thanks,
> 
> Aaron J. Seigo
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/plasma-devel/attachments/20140826/675e6b91/attachment-0001.html>