- User Since
- Apr 17 2015, 10:32 PM (248 w, 4 h)
We've previously avoided adding any API to PC3 so it can be used as a Qqc2-style.
Is there a better way?
-1 I'd prefer if the config value was kept
Some tech notes:
Semi relevant: https://bugs.kde.org/show_bug.cgi?id=395247
Ack. Thanks for chasing up on indentifying causes.
What does enabled on a label do?
The continual forwarding of keys makes sense. I'm happy to accept that.
Thu, Jan 16
m_dropActions is conceptually linked to the menu we are showing
It shouldn't a direct member variable of ContainmentInterface
Remove unused destructor
Given timeframes and where we've ended up. Accepted.
I've gone through this again, and I'm somewhat confused.
Entry (via EntryWrapper) doesn't seem usable by QML. It doesn't have any properties.
Wed, Jan 15
I like red patches.
Ok, but why minimum?
We still have a few reports of people on weird tiny netbooks.
Quite a clever solution. I like the general idea.
Just to flesh that out more. KF5 will be presumably be co-installable with KF6 and Plasma 6 so we can't just move the file.
@guoyunhe can you propose a solution for KF6 that would make sure we don't have future conflicts
Tue, Jan 14
Roman preferred to merge the less good version as-is, then merge this as top instead of my suggestion of squashing.
Renamed commit appropriately.
Ok, then I withdraw my comment. That sounds like it's going to be good enough.
That error is about a missing include on flipswitch.h
It doesn't seem relevant to this patch?
Mon, Jan 13
Seems we got the go-ahead from kde-distro-packagers ML so ++
Sun, Jan 12
I'm still not totally clear on why some things require a domain to be added for translations, and other things don't.
Sat, Jan 11
There's a big difference between "I want to raise my window" and I "I want to raise $random window".
Fri, Jan 10
Just remove the whole thing
Sounds like 415699 which is unrelated to this change
As a more general comment, I think this entire debug context stuff can be deleted or at least opt-in when we want debug, rather than what seems to be a poor approach of detecting the nvidia resets.
Add extra guard, as there is a timer there's a hypothetical (but not real) chance that scene is destroyed
Extra clue that matches my thinking: