Plasma 6.4 review – A worrying trend(dedoimedo.com)
dedoimedo.com
Plasma 6.4 review – A worrying trend
https://www.dedoimedo.com/computers/plasma-6-4-review.html
20 comments
> Quite some emotions, also in the linked article about XLibre, from someone not emotionally invested in software.
An emotional impact from a piece of software’s failures is not the same as being emotionally invested in a piece of software.
I have not emotional investment in my NIC’s firmware: I honestly don’t care about it, I don’t think about it. But if it stops working, or starts flooding my LAN with random packets, then I will care.
Likewise, I am not personally invested in Wayland vs. X11. I am emotionally invested in the way I use my computer, in my editor, my terminal, my web browser and my window manager: as long as a Wayland implementation supports all of them without regressions, then I really don’t care if my distribution continues to support X.org.
But when no Wayland implementation supports my software stack, I care that my distribution continues to support X.org. And if folks associated with Wayland were to try to smuggle needless incompatibilities with X.org into software I use then I would be upset.
An emotional impact from a piece of software’s failures is not the same as being emotionally invested in a piece of software.
I have not emotional investment in my NIC’s firmware: I honestly don’t care about it, I don’t think about it. But if it stops working, or starts flooding my LAN with random packets, then I will care.
Likewise, I am not personally invested in Wayland vs. X11. I am emotionally invested in the way I use my computer, in my editor, my terminal, my web browser and my window manager: as long as a Wayland implementation supports all of them without regressions, then I really don’t care if my distribution continues to support X.org.
But when no Wayland implementation supports my software stack, I care that my distribution continues to support X.org. And if folks associated with Wayland were to try to smuggle needless incompatibilities with X.org into software I use then I would be upset.
A reply to the article by a KDE dev here [0].
[0] https://www.youtube.com/watch?v=vPJQnaKKKMc
[0] https://www.youtube.com/watch?v=vPJQnaKKKMc
Is there a text version? Asking because I don't feel like listening for one hour; I'd rather be reading.
Here's my attempt at a summary (likely biased, not complete, with mistakes).
Basically, aside the mentions of bugs and the hatred against Wayland, the author dislikes the trend everywhere to hide buttons behind menus in the name of a better organization, requiring more clicks to get things done, and to dumb down UIs for idiot mainstream users. The KDE dev disagrees with this premise, thinks that additional clicks don't matter much in the grand scheme of things, but making things more organized and less cluttered is a big win for making things easier for first time users while not alienating expert users that much, and rejects the idea that we should consider mainstream users idiots.
More specifically:
- KDE dev is happy this is a critical review: they are rare, usually reviews are just uncritical tours of new features which are not helpful feedback. He encourages the author and other people to keep doing this.
- Wayland vs X11 perf: surprising results, usually Wayland benchmarks are better
- The KDE dev doesn't see the differences wrt to font rendering from the screenshots but says we shouldn't trust him on this. Qt should be handling font scaling well. Again, this part is usually better on Wayland
- the 105% scaling could be a result of the complex logic that tries to figure out the correct scaling for your hardware (I agree with the blog post that 105% doesn't look good. Now, nothing prevents from reporting a bug)
- data shows that complaints from users are fewer with Wayland than with X11
- the gamma complaint is one of the niche issues that indeed Wayland doesn't address, but that doesn't mean Wayland is a big mess
- the blog post complains about Scrot, a screenshotting tool, not working on Wayland and that it's the kind of basic features Wayland messes up. It's false, on the contrary, it is a side effect of what Wayland tries to address (security), Wayland provides features to do screenshots, Scrot needs to adapt.
- blog complains about lack of borders, but KDE has been moving away from them to reduce clutter
- KDE dev says he actually likes the Gnome 3 / Windows 11 UIs
- KDE dev defends the idea that more clicks is sometimes better (if it allows better and clearer UI organisation), and can help provide more features, and can help new users figure things out by reducing the number of things appearing on the screen which is also important
- KDE dev rejects each suggestion to add buttons to the Dolphin toolbar (refresh: shouldn't be needed, Dolphin already autorefreshes; Home is already in Places, cut, copy and paste can be found on the context menu which is where users will look for these actions). He mentions that people not liking the hamburger menus are at a CTRL+M away from enabling back the regular menu.
- KDE dev says Spectacle is as efficient as possible like this, and doesn't get the complaint (as someone who followed Nate's blog posts, I followed evolutions of Spectacle, I do believe it indeed improved a lot over the time). He says that Plasma is often rightfully criticized for being overwhelming and they are trying to fix this
- KDE dev rejects the idea that we should consider the vast majority of users as idiots, but as people whose life doesn't revolve around computers and doesn't think that the improvements puts unreasonable burden on expert users (and as someone who has used KDE virtually forever, I agree)
- KDE dev says that saying Spectacle's UI being ruined is overly dramatic
- KDE dev is the one who designed the floating panel and feels being called out. He says that the floating panel was done to stand out from other desktop in screenshots, especially Windows. He says that expressly they made sure pushing the cursor to edges and corners work and added an option to revert the floating panel anyway. He says that the blog author should have tried and gets angry at the "malicious" accusation that the floating panel is there because Qt6 provided a shiny new widget that absolutely had to be used.
- KDE dev takes looking like Gnome 3 as a compliment because they have good designers
- KDE dev rejects the idea that displaying all the detailed packages to updates in a system upgrade is a good idea for the majority of users (everybody does this, it would be overwhelming)
- KDE dev argues in favor of robust, reliable system upgrades and rollbacks, also explains the advantages offline updates and how updates while the OS is running can break stuff because you end up with a system running mixed versions.
- KDE dev recognizes KDE Neon can have issues
The conclusion is a rehash of all this. He reaffirms that they are trying to build good software, not mimicking cheap competition. He rejects the idea that efforts working on atomicity and Wayland would have been better directed at improvements because they bring improvements. He rejects the comparison with Chromebooks, which are SaaS, while KDE remains local.
Basically, aside the mentions of bugs and the hatred against Wayland, the author dislikes the trend everywhere to hide buttons behind menus in the name of a better organization, requiring more clicks to get things done, and to dumb down UIs for idiot mainstream users. The KDE dev disagrees with this premise, thinks that additional clicks don't matter much in the grand scheme of things, but making things more organized and less cluttered is a big win for making things easier for first time users while not alienating expert users that much, and rejects the idea that we should consider mainstream users idiots.
More specifically:
- KDE dev is happy this is a critical review: they are rare, usually reviews are just uncritical tours of new features which are not helpful feedback. He encourages the author and other people to keep doing this.
- Wayland vs X11 perf: surprising results, usually Wayland benchmarks are better
- The KDE dev doesn't see the differences wrt to font rendering from the screenshots but says we shouldn't trust him on this. Qt should be handling font scaling well. Again, this part is usually better on Wayland
- the 105% scaling could be a result of the complex logic that tries to figure out the correct scaling for your hardware (I agree with the blog post that 105% doesn't look good. Now, nothing prevents from reporting a bug)
- data shows that complaints from users are fewer with Wayland than with X11
- the gamma complaint is one of the niche issues that indeed Wayland doesn't address, but that doesn't mean Wayland is a big mess
- the blog post complains about Scrot, a screenshotting tool, not working on Wayland and that it's the kind of basic features Wayland messes up. It's false, on the contrary, it is a side effect of what Wayland tries to address (security), Wayland provides features to do screenshots, Scrot needs to adapt.
- blog complains about lack of borders, but KDE has been moving away from them to reduce clutter
- KDE dev says he actually likes the Gnome 3 / Windows 11 UIs
- KDE dev defends the idea that more clicks is sometimes better (if it allows better and clearer UI organisation), and can help provide more features, and can help new users figure things out by reducing the number of things appearing on the screen which is also important
- KDE dev rejects each suggestion to add buttons to the Dolphin toolbar (refresh: shouldn't be needed, Dolphin already autorefreshes; Home is already in Places, cut, copy and paste can be found on the context menu which is where users will look for these actions). He mentions that people not liking the hamburger menus are at a CTRL+M away from enabling back the regular menu.
- KDE dev says Spectacle is as efficient as possible like this, and doesn't get the complaint (as someone who followed Nate's blog posts, I followed evolutions of Spectacle, I do believe it indeed improved a lot over the time). He says that Plasma is often rightfully criticized for being overwhelming and they are trying to fix this
- KDE dev rejects the idea that we should consider the vast majority of users as idiots, but as people whose life doesn't revolve around computers and doesn't think that the improvements puts unreasonable burden on expert users (and as someone who has used KDE virtually forever, I agree)
- KDE dev says that saying Spectacle's UI being ruined is overly dramatic
- KDE dev is the one who designed the floating panel and feels being called out. He says that the floating panel was done to stand out from other desktop in screenshots, especially Windows. He says that expressly they made sure pushing the cursor to edges and corners work and added an option to revert the floating panel anyway. He says that the blog author should have tried and gets angry at the "malicious" accusation that the floating panel is there because Qt6 provided a shiny new widget that absolutely had to be used.
- KDE dev takes looking like Gnome 3 as a compliment because they have good designers
- KDE dev rejects the idea that displaying all the detailed packages to updates in a system upgrade is a good idea for the majority of users (everybody does this, it would be overwhelming)
- KDE dev argues in favor of robust, reliable system upgrades and rollbacks, also explains the advantages offline updates and how updates while the OS is running can break stuff because you end up with a system running mixed versions.
- KDE dev recognizes KDE Neon can have issues
The conclusion is a rehash of all this. He reaffirms that they are trying to build good software, not mimicking cheap competition. He rejects the idea that efforts working on atomicity and Wayland would have been better directed at improvements because they bring improvements. He rejects the comparison with Chromebooks, which are SaaS, while KDE remains local.
The KDE dev disagrees with this premise, thinks that additional clicks don't matter much in the grand scheme of things,
I know I would.
The problem is not so much about "more mouse clicks". The problem is that by forcing the use of an hamburguer menu, you *must* use a mouse.
In a traditional WIMP GUI app, you press Alt and automatically there is a letter in the menu underlined. That means that pressing that letter in the keyboard, you get the menu pulled down. Then, you can navigate it using cursor keys + enter or other highlighted letters.
This is no longer possible with most hamburguer menu apps in Plasma. You MUST use the pointer device (mouse). That is my pet peeve. More or less clicks is not really of the essence.
In a traditional WIMP GUI app, you press Alt and automatically there is a letter in the menu underlined. That means that pressing that letter in the keyboard, you get the menu pulled down. Then, you can navigate it using cursor keys + enter or other highlighted letters.
This is no longer possible with most hamburguer menu apps in Plasma. You MUST use the pointer device (mouse). That is my pet peeve. More or less clicks is not really of the essence.
> The problem is that by forcing the use of an hamburguer menu, you must use a mouse [...] in Plasma
Apps doing this need fixing (and can be fixed). In Dolphin:
- F10 opens the hamburger menu
- CTRL+M restores the traditional toolbar
Apps doing this need fixing (and can be fixed). In Dolphin:
- F10 opens the hamburger menu
- CTRL+M restores the traditional toolbar
You have at least two options for this in KDE software:
- you add the button you often use to the toolbar
- you use the keyboard shortcut, that you can customize
- you add the button you often use to the toolbar
- you use the keyboard shortcut, that you can customize
Wow this is amazing!! Thanks for taking the time and effort to summarize it.
Running Plasma 6.4 with Slackware64-current.
I agree with TFA on the matter of the absolute atrocity that is the hamburger menu and the ergonomic regression it entails.
However, I have no issues with Wayland as described in the article. My GPU usage is normal, and the default scaling is 100%. Perhaps because I'm using the vanilla Intel onboard GPU of my laptop?
I agree with TFA on the matter of the absolute atrocity that is the hamburger menu and the ergonomic regression it entails.
However, I have no issues with Wayland as described in the article. My GPU usage is normal, and the default scaling is 100%. Perhaps because I'm using the vanilla Intel onboard GPU of my laptop?
At one point, default scaling on my Thinkpad T14s was 105% and it absolutely made everything look bad.
This has been fixed in a recent point release. It should no longer choose these low scaling values as a default. I'm always amazed by the speed small fixes are made to Plasma.
This has been fixed in a recent point release. It should no longer choose these low scaling values as a default. I'm always amazed by the speed small fixes are made to Plasma.
I still feel that Trinity was good than all these newer versions. Can't we have support for something (let it be Wayland, touchscreen, handheld, etc) with same UI than changing the UI again and again to support something (trend, flashy UI, etc).
I'm with you on it. KDE 3(Trinity) was the best.
> Wayland aside, which is going to be a disaster and neuter the Linux desktop
I can't recall where I said this some time ago but I feel like between all the FOSS "wars" (KDE vs GNOME, Vim vs Emacs, Ubuntu vs everyone else...) X11 vs Wayland is the most toxic of them all. It feels to me Wayland haters are the most noisy. For most people like me who are being using Wayland for years it is absolutely great right now so we can't complain. I'd go as far as to say people having issues with Wayland are having mostly a me problem rather than a Wayland problem - yes, it can't be absolutely perfect, but it is a mamooth of a project and if you want it to be better you can help in a lot of ways, not just coding.
Plus, blog posts like this are of little to no help at all about that "Wayland is going to be a disaster and neuter the Linux desktop", maybe even contributing to it - I can't help but feel for the KDE devs who nowadays seem to be really open to new ideas and feedback from users (everywhere from Reddit to the bugzilla tracker and their discourse instance), but some guy decides to make a blog post and torch the whole project.
I can't recall where I said this some time ago but I feel like between all the FOSS "wars" (KDE vs GNOME, Vim vs Emacs, Ubuntu vs everyone else...) X11 vs Wayland is the most toxic of them all. It feels to me Wayland haters are the most noisy. For most people like me who are being using Wayland for years it is absolutely great right now so we can't complain. I'd go as far as to say people having issues with Wayland are having mostly a me problem rather than a Wayland problem - yes, it can't be absolutely perfect, but it is a mamooth of a project and if you want it to be better you can help in a lot of ways, not just coding.
Plus, blog posts like this are of little to no help at all about that "Wayland is going to be a disaster and neuter the Linux desktop", maybe even contributing to it - I can't help but feel for the KDE devs who nowadays seem to be really open to new ideas and feedback from users (everywhere from Reddit to the bugzilla tracker and their discourse instance), but some guy decides to make a blog post and torch the whole project.
They typically make mountains out of minor gripes with Wayland while completely ignoring the huge gaping holes in X11 that made Wayland necessary to begin with. You cannot give normal users a DE that lets any random application have complete control over all of your desktop applications. Sure, it is fine to say, "I am not a normal user, so don't push Wayland on me until it fixes my gripes," but because the author is not a normal user, he should not be using distros intended for normal users. That's on him.
Adding a permission system was a huge improvement. Not actually bothering to implement useful features, even behind permission gates, is not an improvement.
What useful features aren't they bothering to implement? The author's complaint is about window positioning, but that is correctly a compositor feature, and his chosen compositor hasn't implemented it yet. https://bugs.kde.org/show_bug.cgi?id=15329
> For most people like me who are being using Wayland for years it is absolutely great right now so we can't complain. I'd go as far as to say people having issues with Wayland are having mostly a me problem rather than a Wayland problem
I mean, the article lists what appear to be bugs/regressions in the software. If the equivalent Wayland session is blurry and uses more CPU+GPU, that's not the user's fault.
I mean, the article lists what appear to be bugs/regressions in the software. If the equivalent Wayland session is blurry and uses more CPU+GPU, that's not the user's fault.
It’s a side effect of fractional scaling plus hinting
Quite some emotions, also in the linked article about XLibre, from someone not emotionally invested in software.
> And the floating panel. Nonsense. If I want to slam my mouse cursor into screen corners, I ought to do that without any major finesse. A floating panel means you need to aim like an idiot and waste your time trying to hit a "floating" target.
It would have taken 2 secs to check that actually you don't have to do this. I just checked, you can aim the corners or the under the panel and clicking still works. The floating panel is just esthetics and the KDE devs made sure it doesn't have ergonomic difference. It also "reverts" to regular panel whenever a window is close enough, so no space is lost. I do think that the floating panel was not a necessary change, but you can just not care and I do find it looks good.
I can't take such an article seriously when it contains unchecked rants like this, and confuses bugs with design choices like this.