109 comments

[ 3.3 ms ] story [ 162 ms ] thread
I have always loved this effect and was sad when I had an OS thst didn't support compiz.

Yaye!

https://pointieststick.com/2023/10/27/these-past-2-weeks-in-...

KDE should pay attention to font families and sizes. This looks really cheap.

It does? It looks great to me: clear and very readable. What's cheap about it? Or, alternatively, what would not look cheap?
Fonts in screenshots do look a bit microsofty (cheap and low quality) to be honest. But fortunately fonts are trivial to install and change in kde.
What qualities about it make is seem cheap and low quality to you? I'm genuinely curious, because I'm not understanding this...
It has absolutely no personality. If you go to a font search engine and look for "plain" you probably get that one.
Sure, it's plain -- but that doesn't seem like a bad thing for a default font in a utilitarian UI. In any case, is it the lack of frills that makes it seem "cheap"?
Looks thin and enlarged. Somehow “weak” - i dont know how else to describe it. Not a major issue, and is subjective. When i dont like fonts i just install others, its trivial.
Yes, I agree it's a trivial "issue". It's not even really an issue, especially because it's just a default that is easily changed to suit anyone's taste. I'm just interested in getting a glimpse of how others view the world.
The fonts look ugly to me as well. As I read the news on the middle east, I was distracted by the beauty of this font.

https://www.tabletmag.com/sections/israel-middle-east/articl...

It is DISTRACTINGLY beautiful, especially the numbers. I'd rather have ugly fonts than forget everything but the way the numbers of people looked here.

Huh. I guess I must be aesthetically impaired. That font doesn't look any more or less beautiful to me than the KDE one. It's just different. They're both fine to my eyes.
It's just my opinion. Yours is just as valid. The KDE font looks like a basic font with AA. Both are fine to me too, but this font was distracting to me. As a font it fails to convey as much information to me but I LOVE the text. This site is very SEO written but has great readability although the font I'd consider ugly, it's very usable and functional. https://www.linearity.io/blog/old-english-font/
To me this one feels cluttered, way too much going on. Definitely distracting, but I wouldn't call it beautiful.
I think the kerning is slightly off, looking at the end of the word "Their".
random font tangent as an ignorant Westerner, but the Arabic writing system is incredibly beautiful, and there are many incredibly beautiful fonts, yet I often see fonts that look like they're welded together from PVC pipes. no slant at all, uniform weight, lots of circular bends. do these look ugly to native Arabic speakers too, or just to my Western aesthetics?
The font used in Plasma is Noto Sans and the reason we (I'm a kde developer) use it is because it has a very large support for all language in the world, which makes it an appropriate default.
Yeah, one of the first things I do on a KDE install is to set all the fonts to Inter[0]. It helps a lot but still doesn’t fix KDE’s selection of text sizes and weights which has always felt odd to me.

[0]: https://rsms.me/inter/

I like the the way KDE looks and operates, the fonts are clear and easy to read. Please KDE developers keep them as they are, KDE is great!
Yay! Haven't been able to use that since around 10 years ago! Welcome back, friend!
Oh boy that's a blast from the past.

I really liked the Compiz Fusion cylinder effect with windows 3D hovering, I used that effect for both alt-tab and virtual desktop switching.

It looked something like this, although I'm pretty sure mine was anti-aliased more effectively: http://4.bp.blogspot.com/-NPnXddu4LL4/UPRszBHQynI/AAAAAAAAAU...

Now I just use 3 monitors (| - - is a great layout, highly recommended) and switch windows using what was called Exposé and is now called Mission Control on macOS. I miss the experimentation; wobbly windows, flames, all sorts of zoom effects, and edge attraction effects.

Effects like Wobbly Windows and Magic Lamp are still in KDE waiting for you to come back :-) Those are built in, and more effects are available for download from the settings window at the click of a button.

I really appreciate the sense of whimsy that comes with stuff like this. Best of all, it is there if you want it and not if you don't.

How about the window going up in flames on close?
We have that too, it’s called BurnMyWindows in the KDE effects repo
That’s such a blast from the past. It may seem frivolous, but it’s funny what draws a kid in. I spent a lot of time in middle school fussing with drivers to get hardware acceleration to run on crappy secondhand laptops, all so I could play with the cube, wobbly windows, and the flame effects. I had no idea it was doing this at the time, but it helped familiarize me with POSIX concepts that later gave me a jumpstart on my career.
For anyone curious, all this is available on Gnome as well via extensions.
I used to use the “cube” but with six or seven sides (so whatever a hexagonal or septagonal cylinder is called), with the faces being transparent so you can see the windows floating in the background.

Nowadays I use an open laptop with two external monitors, with the laptop out to the left, my largest monitor in landscape in the center and a portrait monitor to the right: - — |

I have settled on exactly the same setup! It's pretty great. My right-hand monitor is a terminal 24/7. My center monitor is code—I used to think I'd want vertical for code, and it is nice sometimes, but at my current job I benefit from having several files open side-by-side more than I benefit from seeing more of one file. It's a 4k monitor and I have it fractionally scaled so that I can read comfortably, but I can also fit up to four files side-by-side on the screen at once. It is very useful. And then the laptop screen is for Slack, docs, AWS dashboard, whatever is useful at the time.
I use the right hand monitor for slack and documentation. Sometimes a terminal, very occasionally code that I’m reading (not writing).

I also used to think is want a portrait monitor for code but it turns out that the Jack of horizontal resolution actually bothers me for code. Maybe with a higher resolution monitor it would be better, but for now, I prefer to use a large but landscape monitor for code.

I will never understand people who have a landscape monitor as their main. The only thing landscape is good for is videos, everything else is portrait (including many videos).

The only semi-reasonable excuse I've heard is that people use a landscape screen as three or four portrait screens (so four windows side by side on it), but then just get a portrait screen and split it four ways also.

You can pry my | - from my dead fingers.

My monitor is primarily a gaming monitor. As such it's stand does not allow for rotation to portrait.
I use a 43" 4K TV as a monitor. It completely fills my visual field, and I cannot turn it sideways.

I hope this comes across as a semi-reasonable excuse.

I use landscape monitors because it effectively allows me to have two portrait arrangements side by side on one screen.

Also a vertical monitor is terrible on my neck. Humans necks are designed to look down but not up above eye level for extended periods.

I learned this after extended discussions with a neurosurgeon on what the most common neck and brain injuries are, and what I could do or not do to best avoid seeing her.

That said, we are all different. Use what you prefer!

You're right, I have the screen much lower than eye level (because I basically lie back on my chair), but this is generally good advice.
This idea sounds nice. How big is your monitor?

I just tried turning my monitor vertical. My initial impression is this left me with very little space with which to read text; and felt like it'd hurt my neck if I were to look up to the top of the monitor. I did get to see more content vertically. (A bigger monitor would alleviate the former but exasperate the latter).

It's 24", I found that anything larger than that was a bit too large.
I use a 24" 1440p screen. On a 1080p monitor (or 4k at 200% scaling) there isn't a lot of width left if you turn it, and if you go for more pixels than 2560x1440 (at 100% scaling) the text gets too small on a 24".
My main monitor is 31.5” and my portrait monitor is 27”. My main is raised maybe 10” above my desk and set up so its center is approx. my eye height, while the portrait monitor is on the desk (so it’s bottom is lower, but it’s too is about the same as the other monitor)

My main monitor is 1440p, I think the other is only 1080p and the lack of horizontal resolution is one reason I don’t use it for code, but I find it works well for slack and documentation. I occasionally use it for code too (eg unified diffs) but mainly I use the main monitor for code. Maybe one day I will replace it with a higher resolution one and then I might use it for code more.

Unless you have a super high DPI display (which in this day and age you probably do) rotating the screen breaks sub-pixel antialiasing, which makes text look blurrier.
Is that a given? I don't think there's an inherent reason why the font rendered couldn't anti-alias vertically instead of horizontally.
Only if something is broken in your configuration. You should be able to set the subpixel layout. Sure, the subpixels will be divided vertically instead of horizontally, but it should still help.
I don’t need an excuse. I just like it.
I find my laptop slides around too much when I'm trying to type with it on its side.
Also some ports get blocked. You have to drill a hole in your desk to use those ports. Very inconvenient.
There's a limit to how far up and down it's comfortable to look for long periods of time. That limit, in most desk setups, is pretty much the height of the top of a landscape monitor. Once you reach that you might as well just have a wider monitor to be able to see multiple threads of work at once.
With large 16:9 monitors in particular this is a problem. 16:10, 5:4, 4:3, 3:2, etc would all work better in portrait displays.
That’s actually a good point. My main monitor sits on top of a stand so it’s raised above my desk, set up that the center is roughly eye height. The portrait monitor is on the desk and its top is about the same height as the main monitor. The laptop is also on the desk so is much lower, but I use it mainly for calendar, jira board and things like that.
That sounds like a pretty good setup.

One thing I heard, which feels true but I don't actually have a way to verify tbh: you should aim to have the _top_ of your monitor at eye level, to avoid temptation to look up often. I've been told that that is bad for your neck.

I tend to use two split windows in my code editor and I’ve found this generally more pleasant than a separate window on another monitor (although part of that is that many editors don’t have great multi window support)
What do you like about split windows? Since I can only pay attention to one thing at a time, I've always found alt-tab superior.
I agree with you about landscape oriented displays, especially with 16:9 laptops. It's like looking at the world through a gun port. At home, I have two 24"displays, one portrait and one landscape which covers all bases.
> so whatever a hexagonal or septagonal cylinder is called

In case you're interested (probably not!), the word you're looking for is prism e.g. hexagonal prism or heptagonal prism.

Unfortunately experimentation, at least at organisations, has been optimised out of existence by a combination of shareholder interests and agile cultists.
kororaa gang. got me into gentoo.

i had a 21in viewsonic crt that was factory specced to 1600x1200, but kororaa would default to overdrive at 1856x1392. it had a bit of trouble with focus at that rez - the field was curved, and the screen was flat. i could have perfectly sharp edges, or a perfectly sharp center, but either way it was sharper than the factory max. i dumped the modelines and carried them with me in my heart. don't think i had that much vertical resolution again until i picked up my first 5k display in 2019.

I remember how much I loved compiz and beryl in high school, my classmates always enjoyed watching me do assignments. I remember it becoming increasingly difficult to make work with OS releases and it forced me to just adopt a "normal" workspace based workflow, a 3 by 3 flat grid instead of the many sided cubes and shapes I used for compiz. I stayed on gtk2 longer than most people via mate, but now I mostly use Mac os with a flat 11 workspaces and it feels so limiting. I wish I could use a 3x3 workspace layout instead of a flat row of workspaces.
I can only really use two these days, although one of them is cheating --it's a terminal running tmux with four windows. But I would replace tmux with four terminals which would go on two workspaces (2 each).

The other is my browser, of course.

I was a full time developer working like this. Tmux was my workspace killer, but of course not everyone works in a terminal.

But I still gotta ask, how are you making use of 11 workspaces?

I use 4-8 at a time by using one per task or activity type. So one for my main web browser, one per project I have going, one to throw tools like volume controls, another for each additional web browser profile.
I easily use 9 as a full stack dev.

1) Coding apps

2) Dev browsers

3) Terminals

4) Sql /DB apps

5) FTP/File handling apps

6) Non work browsing / leisure

7) Specs and text

8) Graphics / Figma / Ps

9) Music / podcast player

Once this is setup with shortcuts it becomes super fast to switch to DB-stuff, or terminal windows etc.

I use 12, though if I'm using them correctly many are usually empty. I assign custom visible labels, for which I use just a single letter. Also I am very familiar with the custom shortcuts to switch desktops or move a window between them; I use the wrap-between-rows shortcuts for the left-right ones. On occasion, I've found that many non-KDE environments are very primitive, often supporting only a single row and not even wrapping around the ends!

There's significant overlap compared to yours but I don't quite all of the dedicated app types you do. One thing I will note is that I don't have a desktop for music - I use VLC and collapse it to the systray (I've tried other players, but kept finding audio files that would break them, and also it was often tricky to make global keyboard controls work).

Column 1:

I = internet, for general-purpose browsing that doesn't fit elsewhere.

A = application, for programs (including) that don't fit elsewhere

R = reading, for blogs and stories, assuming they don't fit better elsewhere

Column 2:

C = code, D = debug (or was it documentation? I admit I'm sloppy within a column), E = ... I don't remember, but basically I use it if I need a third virtual desktop for code-related stuff

Column 3:

G = games, H = help (usually a wiki or walkthrough), J = journal (a place for me to record my own notes, since most games don't provide a builtin way to do that, and even if they do their primitive text editor usually sucks)

Column 4:

X = extra, Y, Z - in case I need another set.

I use 10, with i3 as window manager.

I keep the most frequent ones at 1-4, as these are easiest to hit with left hand via meta+numeric row.

Same, 10 split over 3 monitors using the keypad to switch them independently on each monitor.

I've gotten so used to it that muscle memory kicks in inadvertently for other desktop environments.

Terminal, Browser, Patchbay, Sequencer, Sampler, Synth, Mixer, Arranger, Text Editor, Blank Desktop. That's 10 (on Windowmaker) and I could easily add a couple more.
(comment deleted)
I remember having some mac app that could do this years ago but was abandoned as SIP destroyed many third party extensions.

I really want my 3x3 desktops back also!

Can it really be there hasn't been made an alternative yet?

It is just added latency for questionable benefit.

OTOH, it could actually be helpful to conceptualize the idea of virtual desktops.

Once already used to virtual desktops, the cube becomes cruft.

It’s fun. That’s benefit alone for a lot of people.
"It is just added latency for questionable benefit."

I guess it depends on how beneficial you consider whimsy. I'll trade a little efficiency for a little whimsy most days.

Fun is a valuable benefit.
Absolutely. If you love it, go ahead.

I remember I even used it without the switch animation, as the latency irritated me.

Just to fiddle by manually grabbing the cube and spinning it, while thinking.

Animations have a useful role - they provide often necessary visual cues on the context. For example, randomly jumping in a text page won’t tell you where you are now in relation to the previous position. Scrolling through does.
> For example, randomly jumping in a text page won’t tell you where you are now in relation to the previous position.

That's what scrollbars used to be good at, before they were dumbed down for dubious aesthetic reasons.

Right.

On Gnome, when I switch left/right between virtual workspaces, the animation provides some indication of the switch.

When I tried xfce, there was no animation, it didn't feel like an improvement.

Perhaps this is down to preferences.

It's useful the first few times, and annoying time wasting all the thousands of times after that. Animations should only be included if there's an easy way to turn them off.
How is that 20ms, or even less constitutes waiting?
I've never seen a UI animation that quick. Most UI animations are hundreds of milliseconds. And even if somebody did make one, it would still be annoying. Even a little latency is enough to break the sensation that the computer is part of my own body, and instead make it feel like a third party I'm commanding.
I can’t see I’ve ever felt like a computer is part of my own body. And I’ve been writing software for close to 40 years.
Either you’ve been using only crappy computers/software then or you’re not getting what they’re talking about. Do you feel like your car is an extension of your body when you’re driving? Because science showed that’s exactly how humans behave when driving. This is how our brain works with automatic motor functions.
Literally nothing happens instantly in real life, why should it be different with computers?
All hands tools (including common things like pencils and toothbrushes) react to your movement with zero added latency. Computers should be the same.
>Scrolling through does.

Yet "smooth scrolling" is latency, and to most who are already used to scrolling, it is extremely unpleasant.

When one does scroll (i.e. move the viewport of a larger screen that's being presented through the inside of a window), it is because one wishes to already be seeing the target position.

Any latency through added animation destroys the illusion.

I quickly close any webpage that forces that strong discomfort upon me. And switch away from any application that does it against my wishes.

Though “smooth scrolling” that interpolates between detents is a considerably different creature than “smooth scrolling” that tracks touch and has inertia resulting from it. The former is like nails on a chalkboard whereas I find tracking/inertial scrolling quite pleasing with how it lets one use muscle memory to “fling” the page to a particular point with decent accuracy.
I find it easier to maintain mental conceptions of three dimensional objects in my current working memory than I do a series of flat planes. I feel that we’re wired that way.

Take macOS as an example. The current space is shown as the largest view, which is silly because you’re already on it. The other spaces are shown as little thumbnails you can hardly see the contents of. You can either slide between spaces one at a time or click. In the cube, each face is the same size, you can see the backside of each window and you can view each window’s contents from an angle. It’s the same reason a book is easier to flip through than a Kindle.

I really wish we could shake this strange obsession with minimalism and return to embracing skeuomorphic representations. Computers should adapt to us, not us to them. We have insane transistor counts and memory density at our disposal. Why shouldn’t we take advantage of that to provide even questionable benefits, as you put it, to our user experience?

For virtual desktops I always liked the original animation OS X used for that back in 10.5 and 10.6, where the camera pulled back to show an overhead view of your grid of desktops, including miniaturized versions of all of your windows.

It did well to establish the idea that your current desktop was just a portion of a much larger surface, and by switching desktops you were changing the portion you were working in. It also leveraged spatial memory, as many Mac UI features over the years have.

The implementation in current versions of macOS, which is your typical linear setup similar that found in GNOME, is harder to conceptualize for users because its animations don’t communicate nearly as effectively, plus by default desktops move around based on frequency of usage which screws with spatial memory.

There’s very little in current versions of macOS that weren’t better in 10.6.
Eh, it places your windows in a 3D navigable space, which actually was sort of useful to quickly take in where you were at a glance. So not 100% cruft.
It depends on the implementation. If the actions don't wait for the animation to complete to do work, then you can move ahead of the animations (and maybe they could gracefully merge from one into the next animation).
When you already know what desktop you need to go to, then yes.

Otherwise, this is a great way to see where you put that window you are looking for.

I liked those effects back in the day, but have to admit I've set my macbook to "reduce motion" because it's faster to use with most effects disabled.

For example alt-tabbing between spreadsheets to look at something while typing in another, the effects just waste your time at every window switch.

Same goes for (Android) phones. If your phone ever starts to feel sluggish, just turn down animation durations.
The problem is that animations happen before the transition instead of after.

I would love to see more desktop animations that happen after a change. One idea I have is to make some waves around the window you alt-tabbed to.

Rotating cube!! I remember playing with this 15 years ago on KDE. Unfortunately I only used Linux for hobby reasons, so I never had many tabs or applications open often enough to really make use of it.

Does Windows finally have this rotating workspaces and maybe I'm just not aware?

The first place I saw effects like this was in OS X 10.4 Tiger, where fast user switching (switching users without logging out logged in users) utilized a rotating 3D cube transition a lot like the desktop cube (just without gaps on the cube edges and the camera looking straight on instead from an upper angle).

I thought it was a bit of a brilliant metaphor for multiple users, with each user’s desktop representing a different “face” of the computer, and switching users is an infrequent enough of a thing that the effect doesn’t become grating. It’s a bummer they removed the effect a few releases ago.

Full-on compiz effects with burning windows and all isn’t quite my thing, but I think that modern desktops would benefit a lot from judicious usage of “fancy” animations that go beyond the bog standard fade and scale effects. If designed well they can communicate a lot to users.

Wow! I thought it would never happen. Good to see some people still value fun in the community.
I always found those effects where you show on a 2D display a 3D view of a 2D UI to be like a weird category error for a GUI.
I'm looking forward to Plasma 6, which also fixes night color mode on Wayland. Have to wait a few months though.

> Plasma 6 is built on top of Qt 6 and is scheduled to be released in early February 2024. This page outlines important information and major changes coming in Plasma 6.

what a trip, when i was a young'n I learned a lot about linux by breaking my display trying to get compiz fusion to decorate my terminals and turn into a cylinder.
There used to be such a feature on Macs also. Either native in one of the old OS X versions are via a plug-in of some sort. Anyone knows if this is till possible on the latest OS X release?