Gentoo Archives: gentoo-dev

From: desultory <desultory@g.o>
To: gentoo-dev@l.g.o
Subject: Re: [gentoo-dev] [RFC] News Item: Desktop profile switching USE default to elogind
Date: Thu, 31 Oct 2019 03:22:07
Message-Id: 418917c2-4cbb-fae9-f095-076526daa663@gentoo.org
In Reply to: Re: [gentoo-dev] [RFC] News Item: Desktop profile switching USE default to elogind by Patrick McLean
1 On 10/29/19 19:08, Patrick McLean wrote:
2 > On Tue, 29 Oct 2019 23:03:15 +0000
3 > James Le Cuirot <chewi@g.o> wrote:
4 >
5 >> On Tue, 29 Oct 2019 11:23:20 -0700
6 >> Patrick McLean <chutzpah@g.o> wrote:
7 >>
8 >>> On Tue, 29 Oct 2019 16:10:37 +0100
9 >>> Andreas Sturmlechner <asturm@g.o> wrote:
10 >>>
11 >>>>
12 >>>> Anyone else who thinks this should not be restricted to just
13 >>>> desktop profiles?
14 >>>
15 >>> I am not aware of any use cases for elogind/consolekit on servers,
16 >>> it's really for machines where you have to distinguish between
17 >>> someone connecting remotely and someone physically using a
18 >>> keyboard/mouse connected to the machine.
19 >>
20 >> I switched from consolekit to elogind on my normally headless ARM box.
21 >> It runs XMMS2 and outputs through PulseAudio. Under consolekit, my
22 >> user's /var/run directory would disappear after a while, taking the
23 >> pulse socket with it. Under elogind, I was able to set my user to
24 >> "linger", which fixed the issue. It's not a typical use case but yeah,
25 >> they do have uses outside of desktop.
26 >
27 > Which sounds like perfect for a local setup, probably not something
28 > that should be in a generic profile.
29 >
30 >
31 >
32 The bulk of the proposed news item is apropos packages and the reason
33 that the profiles are changing, not actually profile specific
34 considerations; and running a desktop environment hardly requires
35 running a desktop profile. Given that the news is largely in regards to
36 specific packages it seems logical that it would more fully reach the
37 user base affected by the causal changes were it simply directed to
38 those with the package being deprecated installed (as indicated in the
39 proposed news item) without regard for what profile they are using.
40
41 Beyond that, I'm not aware of any advantages to limiting it to desktop
42 profiles despite the direct cause of the news item being a transition in
43 the profiles; while targeting it more generally does provide the benefit
44 of reminding users who are not using a desktop profile to at least
45 consider following suit prior to last rites on a package that they have
46 installed thereby providing a better user experience on the whole.