On 2026-04-10 12:56, Rafael Sadowski wrote:
> I've been testing the following version for a hours now and am
> happy with it. I "simply" tried to replace timerfd with my own
> QTimer. It works great for the clocks. Since I've now received
> approval from upstream, I'd like to commit it. Does anyone see any
> showstoppers in the code?
> 
> I followed the cmake + create a new cpp file  pattern which is
> valid for upstream.
> 
> 
> Special: I only check for OpenBSD because Net and FreeBSD support
> fimerfd(2).
> 
> See the comments and the PR for an explanation:

I've been running the new plasma-workspace-6.6.4 snapshot package with 
this change for the past ~day and the clock has been working well. 
Nicely done, thanks!