Cleanup settings in override #138

worldofpeace · 2019-10-05T03:27:22Z

Breakdown of dropped settings:

auto-maximize=false
I believe its been discussed in the gala issue tracker
that it's desired for almost fullscreen windows to
be auto maximized. And notoriously I've noticed
this with applications like epiphany, where it's allmost
fullscreen, and it makes the panel look very strange
not being completely opaque and having a gap.
I haven't noticed issues with it being set.

Introduced in de894b7.

org.gnome-settings-daemon.plugins.cursor
This was introduced to workaround a bug where
cursors would be invisible with g-s-d. The report 0
was never actually closed, and I haven't noticed any
one's recently. I believe the issue is now gone in recent versions.
( reports mentioned gtk 3.10)
org.gnome.settings-daemon.plugins.power
ambient-enabled=false
idle-dim=false

These are just powersettings that are defaults
in g-s-d that appear to be completely sane.

I've been using this in NixOS and Soon™ a downstream patch like the rest of the PRs in this repo...

Breakdown of dropped settings: - auto-maximize=false I believe its been discussed in the gala issue tracker that it's desired for almost fullscreen windows to be auto maximized. And notoriously I've noticed this with applications like epiphany, where it's allmost fullscreen, and it makes the panel look very strange not being completely opaque and having a gap. I haven't noticed issues with it being set. Introduced in de894b7. - org.gnome-settings-daemon.plugins.cursor This was introduced to workaround a bug where cursors would be invisible with g-s-d. The report [0] was never actually closed, and I haven't noticed any one's recently. I believe the issue is now gone in recent versions. ( reports mentioned gtk 3.10) - org.gnome.settings-daemon.plugins.power - ambient-enabled=false - idle-dim=false These are just powersettings that are defaults in g-s-d that appear to be completely sane. [0]: https://bugzilla.gnome.org/show_bug.cgi?id=694758

cassidyjames · 2019-10-05T04:02:32Z

overrides/default-settings.gschema.override

 [org.gnome.settings-daemon.plugins.media-keys]
 screensaver='<Super>l'
 terminal='<Super>t'

-[org.gnome.settings-daemon.plugins.power]
-ambient-enabled=false
-idle-dim=false


This is intentional. Instead of dimming the display on idle, we turn the display off; it felt weird to have both settings here and there's no visual way to toggle this one. I would be open to a counter-argument and more input from @elementary/ux.

I can't recall if the idle-dim is happening at a smaller interval and then the display is turned off.
(dimming away)

I know at least, in Pantheon, the transition appears abrupt.
I'd have to investigate the default behavior and how it works in g-s-d though.

cassidyjames · 2019-10-05T04:03:22Z

overrides/default-settings.gschema.override

 [org.gnome.settings-daemon.plugins.media-keys]
 screensaver='<Super>l'
 terminal='<Super>t'

-[org.gnome.settings-daemon.plugins.power]
-ambient-enabled=false


I forget why this was disabled; I think because it seems dishonest to tell people it's on when it's unclear if it's actually supported? If we can actually detect if it's supported, that would be better. @danrabbit do you remember?

@cassidyjames It's likely there's no ui for this and g-s-d just disables it when appropriate.

@worldofpeace there is a UI; check System Settings → Power on a laptop, or the equivalent panel in GNOME. The problem may have also been that it's a poor experience by default; I think @danrabbit had made this call originally so I'm going to defer to him. I also do not have hardware with an ambient light sensor, so I can't test it.

I think at the time we disabled it, support was really bad. It could be better now. It's worth re-evaluating

overrides/default-settings.gschema.override

cassidyjames · 2019-10-05T04:04:52Z

overrides/default-settings.gschema.override

@@ -136,18 +135,10 @@ active=false
 [org.gnome.settings-daemon.plugins.color]
 night-light-temperature=4500

-[org.gnome.settings-daemon.plugins.cursor]
-# Workaround upstream gnome-settings-daemon bug (https://bugzilla.gnome.org/show_bug.cgi?id=694758) as tracked by elementary (https://bugs.launchpad.net/elementaryos/+bug/1248747)
-active=false


Ooh, this we can probably drop. I wonder if it has anything to do with the cursor size stuff, too?

I've tested this and it does seem to improve the cursor being the same size, and it looks like it also fixes support for the cursor size setting.

elementary/gala#300

cassidyjames

auto-maximize is intentionally false. The others are open to discussion.

cassidyjames · 2019-10-05T04:06:04Z

This also makes me think: we should probably have a comment with why on all overrides. There are reasons behind each but sometimes it's hard to remember.

worldofpeace · 2019-10-05T04:43:15Z

auto-maximize is intentionally false. The others are open to discussion.

Thanks, though how have you closed the discussion on auto-maximize? Is there some precedent in your tracker that I'm not aware off? I'd be more open to dropping this change if it didn't appear you decided this for me.

worldofpeace · 2019-10-05T04:43:39Z

This also makes me think: we should probably have a comment with why on all overrides. There are reasons behind each but sometimes it's hard to remember.

Very much so, git-blame here wasn't helpful for these types of things.

cassidyjames · 2019-10-05T04:49:14Z

@worldofpeace

I'd be more open to dropping this change if it didn't appear you decided this for me.

Well, you can do whatever you want downstream but we've decided this a long time ago for elementary OS. I've expanded on my code comment above, as well. I'm not sure where the original discussion took place (probably on IRC back when we used Launchpad, honestly), but this is still very intentionally the design of how Pantheon works.

worldofpeace · 2019-10-05T05:00:07Z

@worldofpeace

I'd be more open to dropping this change if it didn't appear you decided this for me.

Well, you can do whatever you want downstream but we've decided this a long time ago for elementary OS. I've expanded on my code comment above, as well. I'm not sure where the original discussion took place (probably on IRC back when we used Launchpad, honestly), but this is still very intentionally the design of how Pantheon works.

Thanks, I think this feeds into the point that default settings in elementary OS lack documentation.
And I prefer not to change defaults like this downstream if possible, especially when it it's a design decision you've made for the DE. It could reflect poorly on Pantheon and give confusion as to why there's differing behaviors. Usually only needed when there's a consistent user feedback something is degrading user experience for us specifically. And even then trying to avoid fragmentation.

cassidyjames · 2019-10-05T05:06:21Z

@worldofpeace as annoying as it might be… could you propose each of these as a PR? That way each can be discussed/tested/merged on their own without holding the others up.

worldofpeace · 2019-10-05T05:10:21Z

@worldofpeace as annoying as it might be… could you propose each of these as a PR? That way each can be discussed/tested/merged on their own without holding the others up.

Possibly. Would multiple commits be acceptable? I don't think much of the changes here are controversial, so testing a PR commit by commit could be similar.

cassidyjames · 2019-10-05T05:17:49Z

Eh, multiple PRs fits in with the GitHub and elementary workflow better and makes it more likely at least some of these changes will be accepted more quickly.

worldofpeace · 2019-10-05T05:24:29Z

Eh, multiple PRs fits in with the GitHub and elementary workflow better and makes it more likely at least some of these changes will be accepted more quickly.

Will do, there's an "elementary" workflow?
(semantically I'm expecting it to value simplicity 😄 )

cassidyjames · 2019-10-05T05:41:25Z

Yes, one change per PR for ease of testing and merging on its own merits. ;)

worldofpeace · 2019-10-05T10:23:58Z

Superseded by
#139
#140
#141

I rescinded the auto-maximize change.

worldofpeace force-pushed the correct-some-settings branch from aa91360 to 2a85781 Compare October 5, 2019 03:27

worldofpeace changed the title ~~Cleanup settings in override.~~ Cleanup settings in override Oct 5, 2019

cassidyjames reviewed Oct 5, 2019

View reviewed changes

overrides/default-settings.gschema.override Show resolved Hide resolved

cassidyjames reviewed Oct 5, 2019

View reviewed changes

cassidyjames suggested changes Oct 5, 2019

View reviewed changes

worldofpeace closed this Oct 5, 2019

worldofpeace deleted the correct-some-settings branch October 5, 2019 10:24

This was referenced Oct 5, 2019

Don't Inactivate g-s-d cursor plugin #139

Merged

Enable ambient and idle-dim in g-s-d power #140

Closed

Don't set draw-background #141

Merged

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Cleanup settings in override #138

Cleanup settings in override #138

worldofpeace commented Oct 5, 2019

cassidyjames Oct 5, 2019

worldofpeace Oct 5, 2019

cassidyjames Oct 5, 2019

worldofpeace Oct 5, 2019

cassidyjames Oct 5, 2019 •

edited

Loading

danirabbit Oct 7, 2019

cassidyjames Oct 5, 2019

cassidyjames Oct 5, 2019

cassidyjames left a comment

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

Cleanup settings in override #138

Cleanup settings in override #138

Conversation

worldofpeace commented Oct 5, 2019

cassidyjames Oct 5, 2019

Choose a reason for hiding this comment

worldofpeace Oct 5, 2019

Choose a reason for hiding this comment

cassidyjames Oct 5, 2019

Choose a reason for hiding this comment

worldofpeace Oct 5, 2019

Choose a reason for hiding this comment

cassidyjames Oct 5, 2019 • edited Loading

Choose a reason for hiding this comment

danirabbit Oct 7, 2019

Choose a reason for hiding this comment

cassidyjames Oct 5, 2019

Choose a reason for hiding this comment

cassidyjames Oct 5, 2019

Choose a reason for hiding this comment

cassidyjames left a comment

Choose a reason for hiding this comment

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames commented Oct 5, 2019

worldofpeace commented Oct 5, 2019

cassidyjames Oct 5, 2019 •

edited

Loading