Review Request 122400: Add timedated support into the clock KCM as an optional dependency

[David Edmundson]

> On Feb. 5, 2015, 5:16 p.m., Martin Gräßlin wrote:
> > how difficult would it be to get this with runtime detection? Just because the distro compiles with systemd support doesn't mean the user uses systemd (c.f. Debian).
> 
> David Edmundson wrote:
>     They don't need to use it, they just need to have it installed.
>     
>     It's possible; we'd just have to query some read property from timedated and see if it errors in the kcmodule ctor then switch accordingly.
>     
>     Downside is it means we still needs to provide our helper, even though it won't be used and the code gets even uglier.
> 
> Martin Gräßlin wrote:
>     if it works with having it just installed without needing e.g. systemd as init, that's good enough, I'd say.

I went for runtime. On reflection I think it makes sense.
As much as I like stirring up some controversy, this is both politically and practically safer for now.

It's not like people with timedated are going to lose much by having to install a tiny helper.


- David


-----------------------------------------------------------
This is an automatically generated e-mail. To reply, visit:
https://git.reviewboard.kde.org/r/122400/#review75487
-----------------------------------------------------------


On Feb. 8, 2015, 5:18 p.m., David Edmundson wrote:
> 
> -----------------------------------------------------------
> This is an automatically generated e-mail. To reply, visit:
> https://git.reviewboard.kde.org/r/122400/
> -----------------------------------------------------------
> 
> (Updated Feb. 8, 2015, 5:18 p.m.)
> 
> 
> Review request for Plasma.
> 
> 
> Repository: plasma-desktop
> 
> 
> Description
> -------
> 
> The current time setting helper is incredibly broken.
> 
> It manually tries to run a range of NTP utilities, all of which are
> deprecated.
> 
> We can just call timedated directly and cut out the middleman as it has
> uses polkit anyway.
> 
> This is currently an optional dependency, and the original helper still
> exists. It makes the code messy, but we have users to support for now.
> 
> Finding timedated is an cmake option rather than querying for systemd
> libs to make it easier for those deploying shims, such as BSD.
> 
> 
> (code is in two commits, first abstracting the saving from the dtime class; then adding in the second save mechanism) 
> 
> 
> Diffs
> -----
> 
>   kcms/dateandtime/timedated1.xml PRE-CREATION 
>   kcms/dateandtime/main.h c1e5234 
>   kcms/dateandtime/main.cpp 0041a9d 
>   kcms/dateandtime/dtime.h 1a90698 
>   kcms/dateandtime/dtime.cpp 482e483 
>   kcms/dateandtime/CMakeLists.txt 4a987ae 
>   kcms/dateandtime/dateandtime.ui c073b5e 
> 
> Diff: https://git.reviewboard.kde.org/r/122400/diff/
> 
> 
> Testing
> -------
> 
> 
> Thanks,
> 
> David Edmundson
> 
>

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/plasma-devel/attachments/20150208/779e26d4/attachment.html>