In the Performance section of Firefox's settings I have "Use hardware acceleration when available" turned on, and under Browsing "Use smooth scrolling" also turned on, and it's butter smooth under Plasma and bspwm with compton.

I prefer the single-click, though I imagine most people would prefer double. As long as it's an option to enable single-click, I think setting the default to double would be more inline with most users, especially new users coming from other DEs and platforms.

If you go into Dolphin's Control > Configure Dolphin > Services, there's a "Delete" service. Turning it on enables a right-click option to delete files instead of moving them to trash.

They can be used by fish shell, zsh, vim-airline and others that are powerline-styled @DataDrake. The nerd fonts have more, and arguably better glyphs than the powerline fonts.

Don't know if it helps @sunnyflunk, back when I was using Nvidia I always had to install libvdpau, then explicitly tell VLC and mpv to use it for rendering. In ~/.config/mpv/mpv.conf I'd set:

As changing the time setting system wide, yes. It's not just the one user's clock being changed, it's system time.

Well first, let me say thank you @feskyde. I've been able to build Awesome, but frankly have never been in a position were I could maintain it.

Disconnecting a display messes up list of applications in taskbar...

@dan_ , could you provide the output from inxi -G?

Works with LTS kernel, but as @Schyken said, it's intermittent. That's also a problem that was reported upstream back in 2013, with reports being added to the bug ticket right up to December of last year.

There seems to be a problem with the newer kernels and KDE's 'Solid' backend. There's a bug report here about it.

A picture is worth 1000 words :)

I don't have a battery so I can't test this @Brandywine, is this the "Battery Time" widget from the kde store, or the built-in one from the system tray? IIRC the default one in the system tray doesn't show time, just percentage.

Both of those issues are upstream bugs in Plasma @padraig. And sadly, they're both quite old ones at that.

This is really weird. Have you tried playing a different mp4 file? What happens if you try to play something like a .webm or other type of video file?

No no no! Sorry @blaz, I meant the one that VLC is using for rendering video. Under Tools > Preferences > Video > Output and Tools > Preferences > Input-Codecs > Hardware-accelerated decoding. Setting them to vdpau.

Have you tried changing the video driver backend?

What happens with just using the symlinks for lib64 sudo /usr/lib/vlc/vlc-cache-gen -f /usr/lib/vlc/plugins/? See https://dev.solus-project.com/T5903.

It seems like mpv is trying to use swrast for a video driver, which I don't think will work with mp4. Check if libvdpau is installed (it should be as a dependency). As with your VLC problem @blaz, it seems like they're having trouble detecting the right driver. You can set it manually in ~/.config/mpv/mpv.conf with:

Running sudo /usr/lib64/vlc/vlc-cache-gen -f /usr/lib64/vlc/plugins/ should clear up the plugin cache errors @blaz. Seems like VLC's install/update triggers aren't doing it as they should.

Try moving ~/.config/vlc, ~/.local/share/vlc and ~/.cache/vlc to a temporary location (or rename them in place to something else like "vlc-backup") so VLC will start with it's defaults. See if that gets it working. The old plugin caches may have corrupted something else in you user configs and cache.

Do you have tracker-miners installed? It's a separate package from tracker itself. If it is, try reinstalling it.

Try running sudo /usr/lib/vlc/vlc-cache-gen -f /usr/lib/vlc/plugins/ to force VLC to rebuild the plugin cache. That should fix it.

I've had similar trouble with tracker in the past. I suggest first running tracker status in a terminal to see if tracker is indeed running and all the miners are loaded. If not, start it with tracker daemon -s and then add that to your autostart programs.

I'm afraid this is beyond my knowledge now. The only thing I see and can think of that's different is the kernel, systemd and the rest of the startup should be the same. It's going to need someone that knows the kernel differences.

With things initializing in different orders, it's really hard to do an effective diff on them. Something I did notice is the monitor assignment seems to be different between the two.

@Vistaus, can you check the log after booting the 4.9 kernel and see if you find these lines anywhere in it:

I'm far from an expert at boot problems, but it seems to me from reading the log that the system is up, SDDM is loaded, you just can't see it.

I've not been able to reproduce the crash, but doing some testing in a VM have had baloo seem to go crazy and never stop indexing. Found that it seems to be choking on hidden folders. I believe it's trying to index things like .cache, .config, .local, .mozilla etc., and any .git folders a user might have. It hasn't happened on my daily-driver, and the only difference I found is I have that set to not index hidden folders. Adding the setting to baloofilerc and restarting it seems to cure it.

I don't know if it helps, but seeing I have an AMD POLARIS card on Intel, and have had no rendering issues, I thought it worth mentioning. Apparently it's not affecting all.

Okay, now I understand what's going on. I thought you meant having one window maximized, and another open, that you couldn't click on the maximized window when all the way on the left. Totally misunderstood what you meant.

No worries. I've tried it on all sides anyway, and can't reproduce it. Let's see what happens with the Acer.

Try opening a browser, then a terminal. In the terminal enter journalctl -f. That will show the last few lines of the system log, and give live feedback on any errors. Try clicking as you have been, see if anything is being reported to the log.

Were these installs all done from the same ISO image @Skullone? I'm not able to reproduce it either, not on my metal install or in VMs. Seeing you have the same problem on 3 different machines, it makes me wonder if the install ISO had some glitch in it.

Okay, I can confirm the missing icon problem. My daily driver is running the last ISO with updates, and it works there (as in screenshot above). Installed the latest ISO in a VM and the icons are missing. Checking the working one against the one on the latest ISO, I can't find any difference. Symlink to fontawesome is exactly the same, fontawesome is installed and visible in Font Manager. Tried rebuilding the font cache, still no luck. Checked links, permissions, locations...

Do you have Configure Cantata... > Interface > General > Style set to "Breeze"?

That's the new icon? I saw that in their GitHub, thought "No, that can't be the right icon, doesn't look anything like it."

Are you quitting and restarting VLC after switching outputs? Most settings won't take effect until it's restarted.

@whatever, I had the same problem with VLC, both full-screen and windowed. Looking for a reason, I found VLC isn't detecting the correct video decoder. The default "Automatic" doesn't seem to detect it. In my case (AMD GPU) setting it to "vdpau" in both Video > Output and Input/Codecs > Hardware-accelerated decoding fixed the problem. I believe with Intel one needs to use va-api or some such.

Hi @mrkucho, sorry I won't be able to help with your first issue, I don't have UEFI nor know much about it. Hopefully @sunnyflunk or someone else will have some ideas on that.

All original issues have been resolved. Any issues added later should have a new task created for that specific issue. I suggest waiting until after the next testing ISO is released, as some may be fixed upstream.

Resolved via this commit: https://dev.solus-project.com/R3433:21618bbd676980b41cb29534ce233c405a0dda7d

I thought you likely would have chosen the keyboard during install, had to ask to be sure.

I'm a bit surprised it wasn't set when you did the install. Did you choose the German keyboard then?

What does it say in /etc/locale.conf?

Don't know what to make of the user icon problem, I simply can't reproduce it. In my daily and the 3 different VMs I have for testing, they all work as expected. Guess we'll have to wait and see if anyone else can duplicate it.

On the user image issue, can you check permissions for your /home/<username> folder? When you create a user image, it gets stored in your home folder as a file named .face and a symlink to it named .face.icon. Perhaps the scm is having trouble writing to it?

@domacs, on your item 2, have you tried using any of the default images from the gallery? I'm not able to reproduce the problem here, so I'd like to know what steps you've tried. If you're using an image, have you selected an area in it to use? What format is the image, .jpg, .png?

After the updates synced this morning, I created a new VM to test. Installed the updates before attempting to change anything, then rebooted the VM to ensure the updates were applied.

@sunnyflunk, I can confirm the patched update is working. Created a new VM this morning when the latest update sync hit. Made no changes after install to the virtual machine, and ran updates, then rebooted. Went to System Settings > Regional Settings > Language and chose my language. After clicking "Apply" I opened Dolphin and checked ~/.config/plasma-locale-settings.sh and plasma-localrc, both showed the correct uft-8 setting. Rebooted again and checked env in the terminal, everything is correct.

I would think we do, if for no other reason than it's already packaged for nautilus and caja.

That could be. After some serious testing, about 2 days ago I felt it was stable enough to make my daily driver. So I backed-up my files (including ~/Dropbox) and blew away everything else.

Okay, now I think I know what the problem is. Right now the current ISO was bitten by an upstream bug that doesn't set the proper locale and language. If you open a terminal and enter env you will probably see several places where a variable is set to en_US.US-ASCII or something like it. All those environment variables should say en_US.UTF-8 (if you're using US English) instead. UTF-8 supports a lot of character encodings that ASCII doesn't, and anything trying to use those encodings will fail.

It probably does need some adjusting @robgriff444, but I'm not sure if dropbox does or Plasma. I looked at the file list that nautilus-dropbox installs, most of it just icons and such. The /usr/bin/dropbox executable it installs is really a Python script. The script checks if the user has a ~/.dropbox-dist folder, and if they do, launches the dropboxd daemon it finds there. If they don't have that folder, it goes and grabs it from the website and does an install. The python modules nautilus-dropbox pulls as dependencies are for the install dialogs, GTK ones in it's case. Once the daemon is installed, all /usr/bin/dropbox does is launch the daemon, then exit. The daemon is doing the rest.

Yes @sunnyflunk, nothing else installed except what came on the ISO, dropbox (via above), zsh and a few more CLI tools. Nothing GTK.

If you grab nautilus-dropbox from the repos, it doesn't pull in nautilus itself. It pulls:

The font is called "Ani". I have no idea why qOwnNotes would be trying to use it, but I'd change it to something more sane, Noto maybe?

@Vistaus, the problem seems to be (at least from what I'm finding on other forums) WPS Office is a GTK-only app. Works fine in any GTK-based desktop, but chokes on anything Qt.

Sunnyflunk may have already patched your desktop cube problem here. If you read the comments from the dev in the patch, it's sounds very much like what you're describing @whatever. Especially if you have icons on the desktop (the current default setting) and are using the mouse to move the cube. I would suggest to see if it's still a problem after the next sync.

This article by one of the Plasma devs is an honest look at what they known they need to fix in Plasma. A big one is GHNS, which suffers from the same problems Gnome's extensions do. Many are outdated, lots are badly made, etc. It's worthwhile reading, as it's gives just some of the bigger issues facing Plasma, and as far as Solus is concerned, shows what is upstream issues and nothing to do with Solus' spin.

Well, I have my main install on metal, on a spare drive, and three VM's installed for tests, and it's not set on any of them. I have no idea what would cause it to randomly come on either. Maybe something else being set/unset triggered it?

This is the task with the workaround and link in it: https://dev.solus-project.com/T5694

I haven't seen any splash screen show up in my install, nor in any of the VM's I spun up today. Kinda weird you're seeing it @whatever.

I created a couple of VM's to test on, and yes, the chosen language during installation is being set in /etc/locale.conf. However, once you change the settings for languages in System Settings, when you reload the session the default settings are no longer being honored. The user's settings take over, only they're not set properly.

Found this is an upstream bug which has already been fixed in master. Hopefully the patched version will be released soon.

I would think so. Seeing how all the other Solus desktops are GTK based, they have gnome-keyring by default. Plasma doesn't, so I suppose they'll have to add some dependencies to quite a few packages now to ensure they'll run under Plasma.

Okay @mmeinert, I found where the problem lies, but still not what's causing it. There are two files that are used when Plasma starts that are suppose to set the locale environment variables. For some reason they're not being set, and one of them is a shell script that isn't getting it's executable bit set. They both are in ~/.config. The first one is plasma-localerc, and is now set like this for mine:

I don't know if the problem is the environment setting isn't being read, or if the wrong location's being searched for it. Doesn't seem to be a problem on the other ISO's. Obviously something's different with Plasma or the ISO, but what and where is going to take some digging.

Hmm... I believe there should be two settings returned, one for the language, another for the character encoding. On my system it looks like this:

@mmeinert, can you run env | grep LANG in a terminal and paste the response here?

I found this on their Github page. Mailspring is basically hardwired to use libsecret on Linux, and libsecret doesn't work with kwallet.
To get it working it appears you'll have to install gnome-keyring, and add it to your startup applications. The syntax for the startup is: gnome-keyring-daemon -s

I don't have brightness keys, but the volume is working, I get the on-screen display when changing volume or muting.

I think I may have found the cause @sunnyflunk, though not a solution. I installed Keepassx, which immediately was set as the default for almost every filetype there is.

I can't reproduce as I don't have wifi. @jppmarujo, can you try running something from the terminal? Disconnect the wifi, open konsole, then connect the wifi. See if you can type into konsole, and if so, try to start dolphin or some other app. Do you get any error messages?

I just installed Kontact and Korganizer here, and they're working.

Now I get it. Some of Plasma settings only get written to config files that are read on startup, they don't always take immediate effect. When I changed it a few minutes ago it had no effect either, nothing changed. It could be one of those settings that needs a session restart.

After spending a while reading about mime handling, I've come to the conclusion it's a complete train wreck. Freedesktop tries to create and set standards, but everyone seems to just ignore them and do their own. XDG, dbus, mimeapp, *.desktop, etc., etc., there's just too many ways of activating file associations, and too many places the associations are can be stored/set.

Scrolling is working fine for me, but then I'm not sure what exactly is meant by "reverse scroll".

I really don't know if it's worth all the trouble to change the default editor. I'm guessing that's the way it's set by upstream, so perhaps it should be left as is until you get more user feedback @sunnyflunk.

Hadn't checked it, but yeah, mine's blank too. Nothing special during the install, and it was a clean disk, no earlier system or settings.

Actually, that should be setting Kate's to something lower. It seems the higher number takes precedence.

I'd hope there's an easier way. :D

I've always just set it in System Settings > Applications > File Associations @sunnyflunk, but honestly have no idea how one would set it as a system default. I've never built an ISO before. I'm looking now to see how it might be done.

They are, but Kwrite exposes less options and would IMHO be less intimidating than Kate to the average user. It could be set as the default interface instead of Kate.

I don't understand this one at all @Vistaus. It appears that SDDM is killing the X server before Plasma even starts.
https://s13.postimg.org/jbkzf0s07/IMG_20180127_142808.jpg

From a terminal run sudo journalctl -r. Normally you would just use journalctl to show your user's logs, but running as root will ensure you see all the logs. The "-r" switch tells journalctl to start from the last entries in the logs instead of the first. Depending on how much has gone on since you last tried, you may still have to scroll through a bit. You're mainly looking for errors from the DM, Plasma or Kwin, but any errors are worth noting.

I really can't say @Vistaus, I'm not familiar with SDDM. I use LightDM myself. I don't think the display manager would make a difference, but I don't know for sure. It shouldn't.