ddcutil in Plasma Powerdevil

[Dorian Vogel]

Rah that's dicky me trying to add the WITH_DDCUTIL symbol and propagate it to be used 
in a #ifdef in the actual code (ddcutilbrightness.{c, h}).... I spent a lot of time trying to get it 
to work although it's probably basic cmake stuff...

Hmm after a bit of research, is it possible that the jenkins uses a Cmake older than 3.12 ? 
It seems "add_compile_definitions" was introduced with this version.
Related discussion https://github.com/facebookresearch/habitat-sim/issues/137[1] 

The "old" command would be:
add_definitions(-DWITH_DDCUTIL)

Cheers,
*Dorian*


On Monday, 30 September 2019 22:27:51 CEST Jonathan Riddell wrote:
> https://build.neon.kde.org/job/bionic_unstable_kde_powerdevil_bin_amd64/127/
> console
> 
> *17:40:13* CMake Error at CMakeLists.txt:56
> (add_compile_definitions):*17:40:13*   Unknown CMake command
> "add_compile_definitions".
> 
> Is the cmake error I get for Powerdevil
> 
> 
> https://cgit.kde.org/powerdevil.git/
> 
> Jonathan
> 
> 
> 
> On Mon, 30 Sep 2019 at 21:17, Sanford Rockowitz <rockowitz at minsoft.com>
> 
> wrote:
> > You might want to look at the launchpad packaging (
> > https://launchpad.net/~rockowitz/+archive/ubuntu/ddcutil), which includes
> > the library version and development package.  It is at the current release
> > level (0.9.7).
> > 
> > If the cmake error is not obvious, let me know what it is.  The locus of
> > the problem may jump out at me.
> > 
> > Sanford
> > 
> > On 9/30/19 9:07 AM, Jonathan Riddell wrote:
> > 
> > I put ddcutils package in Neon unstable to be able to test it, using
> > Sanford's packaging from Debian but with the libraries turned on.  However
> > powerdevil doesn't compile with DDCUtils turned on, it fails on a cmake
> > error.
> > 
> > Jonathan
> > 
> > On Fri, 27 Sep 2019 at 07:14, Dorian Vogel <dorianvogel at gmail.com> wrote:
> >> Hey Jonathan,
> >> Unfortunately I did not test it for a long time (last time was around
> >> 5.15), and don't really have the possibility now because I'm away from
> >> home. *If I remember correctly,* back then I had a few instances when
> >> the startup of plasma (during the splashscreen) was quite longer due to
> >> retries of the DDC communication (of course this will not happen if the
> >> user does not have ddcutil itself installed). As I understand it, this
> >> is due to the architecture of Powerdevil, which assumes the monitors are
> >> a fixed configuration as far as brightness is concerned: backends are
> >> queried at startup and never refreshed. This is why enumeration has to
> >> happen during startup of plasma, and one would have to restart
> >> powerdevil when connecting a new DDC capable monitor
> >> 
> >> Thinking back I wonder if a plasmoid would have been a better way,
> >> enabling to have the enumeration done after startup is finished, and
> >> making
> >> it easier to have a "refresh" button, and would still allow to have both
> >> an
> >> integrated screen (laptop) and external (for now the backends in
> >> powerdevil
> >> are exclusive).
> >> 
> >> Considering this is my first contribution and I am not sure of the
> >> quality, I think some testing from more experienced devs would be
> >> required.
> >> I haven't heard much so far.
> >> 
> >> Cheers,
> >> *Dorian*
> >> 
> >> On Thu, Sep 26, 2019 at 11:13 AM Jonathan Riddell <jr at jriddell.org>
> >> 
> >> wrote:
> >>> The DDCUtil features of Powerdevil in KDE Plasma are currently disabled
> >>> by default
> >>> "DDCUtil library support is disabled by default as recomended by
> >>> authors, add -DHAVE_DDCUTIL=On to enable"
> >>> This is based on a conversation from 2017 where it was decided it was
> >>> unstable.
> >>> 
> >>> I see work has continued on DDCUtil and on support for it in
> >>> Powerdevil.  Is it ready to be set to on by default?
> >>> 
> >>> It is too late for Plasma 5.17 due for release in 2 weeks but we can
> >>> tell distros the message is old as appropriate.
> >>> 
> >>> Jonathan
> >> 
> >> --
> >> VOGEL Dorian



--------
[1] https://github.com/facebookresearch/habitat-sim/issues/137
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/plasma-devel/attachments/20190930/13644bf3/attachment-0001.html>