Hi all.
I've joined the list at Dan's suggestion, as I will be maintaining
waypipe[1] now, and admittedly it could be useful for sway users, so why
not.
[1] https://gitlab.freedesktop.org/mstoeckl/waypipe
Looking at the SIG page[2] there is a wishlist item for "redshift with
wayland support", and as a big user myself I would be all for it.
[2] https://fedoraproject.org/wiki/SIGs/Sway
The problem is that upstream[3] has not been very active for a while -
merged one PR in 2019, but that aside pretty much nothing since 2018.
Frankly major DEs (gnome, KDE) incorporate the feature natively so users
of redshift on wayland will basically only ever be wlroots-based
compositors, so we shouldn't expect to get much help from anywhere on
this... I imagine the dev uses gnome as he wrote the gnome-clock
location helper redshift-gtk service files, so he's made the switch and
hasn't looked back much.
[3] https://github.com/jonls/redshift
The wayland support issue[4] suggests arch uses minus7's wayland
branch[5] but that would appear to still be AUR and they are in the same
situation as us (some coprs have it alongside older versions of sway but
nothing official).
I would suggest the way forward would be to try to discuss with jonls
and minus7 if an official fork could start somewhere; either minus7 tree
if he is willing to do it or I suppose I could (I also have vested
interest, I am also using the stdin remote control patches discussed on
this issue[6], so "just" wayland wouldn't be enough for me, and whoever
takes over would need to be willing to deal with at least accepting PRs
:))
[4] https://github.com/jonls/redshift/issues/55
[5] https://github.com/minus7/redshift/tree/wayland
[6] https://github.com/jonls/redshift/issues/690
So, what do you folks think? Worth a shot?
Cheers,
--
Dominique
Hi all,
the mention on systemd integration in the recent Aleksei's PR [1]
and the related discussion in sway PR [2] got me thinking.
[1]: https://src.fedoraproject.org/rpms/sway/pull-request/11
[2]: https://github.com/swaywm/sway/pull/5622#issuecomment-740978176
TL;DR:
What about starting a semi-independent sway-systemd integration project?
Long version:
I get why emersion does not want any init system/display
manager/whatever integrations added to sway itself
– it's more of a maintenance burden,
and it can also be taken as "endorsement" of any given system(s).
On the other hand, leaving the integration instructions on wiki page
and telling distribution packagers to basically implement
any integration "downstream" does not sit well with me.
You would have to extract the snippets from wiki page for each distro
separately, then keep an eye on any changes, etc.
So, I got the idea of starting "unofficial" `sway-systemd` project –
basically extracting the stuff from sway wiki and keeping it in git,
allowing maintainers of other distro packages to pool the efforts.
Unfortunately, I'm not currently planning on actually *using* the
systemd integration in the near future, so I'm probably not the best
person to push that idea further – at least not on my own.
The question is: does this idea is something that appeals to other
people here, and if so, do we want to start the ball rolling?
Have a nice day!
--
Jan Staněk
Software Engineer, Red Hat
jstanek(a)redhat.com irc: jstanek
