Reverting review 123473

[Marco Martin]

On Thursday 07 May 2015, David Edmundson wrote:
> We had an agreement that there would be no regressions when porting any
> KCMs to QML.
> 
> The following regressions remain in the newly ported KCM:
> 
>  - combo boxes load in wrong places
> https://codereview.qt-project.org/#/c/111326/ isn't merged, and Qt 5.5
> isn't a requirement for the Plasma 5.4 release.
> 
>  - combo boxes have wrong backgrounds

I don't see any of those being usable by an our release for almost ~6 months-
one year, that makes this work pretty much useless

>  - missing automatic accelerators
> (we need a QML version of the class in KXMLGui)

hmm keyboard shortcuts seem to work for the couple of buttons there?

> I am happy to help work on those problems, but I am not happy about seeing
> decisions being willfully ignored.

same as the decision of months ago of moving kcms to qml, if in the end the 
decision is reverted and is decided instead disadvantages outweigh advantages, 
fine

> Tomorrow I will be reverting that commit.
> 
> David Edmundson

on that reasonig, even the energy infocenter thing should be retracted, since 
has a combobox with the above 2 problems. and I don't think the fact of being 
new versus replacing something makes it more acceptable.

If others agree to revert it as well, go ahead (being a merge what is the best 
way? just pasting on top the old version or blacklisting the commit?).

I just fear that systemsettings and the modules will still stay stale for 
aloong time. I think they are pretty bad, (even just for showing their age) 
and need a lot of work, I don't think they can happen without regressions, but 
maybe that's just the grumpy me

-- 
Marco Martin