Discussion
My "Future of Desktop UX (part 2)" talk at #KDE Akademy is now up and it's racking up lots of views. I'm excited people want to discuss this more.
Now I just need someone with Wayland experience to help me prototype this.
https://www.youtube.com/watch?v=V7AfAcQwLW0
@DanFessler sorry for the delay, I'm getting a bit slammed by the replies. This is very cool! This was clearly built with care and lots of attention to detail.
@scottjenson I can imagine! No worries at all!
@scottjenson
That's the google-free links: https://media.ccc.de/v/kde2026-7-are_we_really_going_to_use_the_same_desktop_ux_forever
@scottjenson scott, that was phenomenal! may I ask how you created these interactive prototypes? were they html/css?
@thibault They were all HTML/CSS but I was using the new Canvas-in-HTML API in the mor recent Chrome build. It allows me to build everything using standard HTML/CSS but then animate it as if it was a standard Canvas object. Very fun new API.
@scottjenson Thanks Scott! Will check it out
@scottjenson I think I had the pleasure of letting you know nepomuk exists so it was cool to see the reference to it actually at a KDE event.
With attention metadata as an episodic memory, is there a good way to talk about how it is distinct or interrelated to the semantic memory of users explicitly tagging files with colours and text tags? The mini, gradual virtual desktops approach resonates with me and provides a convergence opportunity to pick that activity back up selectively on a laptop size screen or other device.
I also find the centre area focus offers an interesting analogy to productivity as a directed graph of activity nodes where you might place inputs on the left and outputs on the right, with automation taking over steps in a pipeline when you ask it to.
Great talk. Can't wait to see more come of it, as I'm sure it will.
@octarine_wiggle So glad you liked it. I"m now trying to figure out how to make these HTML demos into something a bit more real. It doesn't have to be perfectly executed, but good enough to actually play with. I expect "making it real" will uncover more issues/insights.
Great talk!
I'm espectially intrigued with the idea of maintaining clipboard provenance, that immediately feels like it would be useful to me.
The spatial ultrawide concept is neat, though I immediately see the issue of it being kinda physically tedious to drag your cursor over to a window docked off to the side.
I also still think that scrolling tiling window managers are interesting! I remember annoying you a little bit about them after watching that first talk.
On an ultrawide they achieve an effect somewhat similar to your idea: they keep one window in the middle of the physical screen, then having a few secondaries to the sides, and then the less-used windows you can shuffle off-screen entirely.
@hjvt I haven't really explored keyboard shortcuts yet. The linux community LOVES their keyboard combos 😉 I have a few idea on how to do that. There is nothing wrong with coming up with "power user features" as long as they are layered on top of a solid and approachable base for more casual users.
@scottjenson
The big thing that's often bad about keyboard shortcuts is discoverability. It is pretty rare that applications have a sort of "you've pressed one key, here's a list of actions whose chords start with that key" pop-up, or even just a convinient list of all defined shortcuts.
While tiling window management is traditionally all about keyboard-only purism, one of the things that makes `niri` (which is my only window (pardon the pun) into the scrolling WM world) feel nice is that it actually utilises touchpad gestures.
@hjvt Good points. I think we can also take advantage of mouse events, e.g. a super-drag could park a window in that direction. But trackpad gestures are also worth exploring.
@scottjenson The Hookmark app for Mac could serve as a practical example of associative memory augmentation
As a very much non tech savvy user, only recently starting with Linux: Ooooh...
Watching your thought experiment of how to use a large screen reminded me strongly of the way I use my physical workspace. Every topic or task has it's own collection (pile) of materials, notes, highlights, scraps of paper and scribbles. Brought to the centre to work on. Very messy (I'm told) but my brain just works well that way. If I could replicate that, it'd be brilliant.
Thank you.
@scottjenson Thanks so much! I’m always so happy to see people that don’t just copy the status quo or make the borders more rounded and call it a progress. 🤜🤛
@scottjenson extremely interesting prototypes, thanks for your talk!
@scottjenson Care to explain why you included the Omarchy logo (26:04 in the video)?
They're literally a bunch of nazis.
@scottjenson thanks! Just rewatched the original from 9 months ago as prep for the new one. Honestly, the first one was great the first time, even better almost a year later.
@scottjenson holy hell, this is so great. How are you... sorry for this analogy in this context... pushing all the right buttons with me?
@scottjenson thanks, here is a link without Google and its tracking https://media.ccc.de/v/kde2026-7-are_we_really_going_to_use_the_same_desktop_ux_forever
@scottjenson I made a mastodon account just to say that was amazing and exactly the type of thinking I need for my current project. Where can I learn more about how to think creatively in HI space? It's such a battle of Jakob’s Law vs innovation.
@wojang Glad you enjoyed it. I'm afraid there's no single thing I can say to think more creatively other than just Racking up more experience. It's really just a matter of putting in the miles.
@scottjenson experience is the best teacher 100%. Also agreed with your assessment that large companies have a harder time innovating in this space. My thoughts are it's because they have core users that are the recurring customers (b2b or b2c) and are afraid of alienating their base. was surprised to learn you don't want to discuss AI though considering it seems to pollute the internet with rehashed tailwind + shadcn saas apps with threejs bolted on top on some.
@wojang actually I'm very willing to discuss AI. I just treat it as a second order effect. I'm very bullish on ethically trained small language models that can be run locally
Install bonfire.cafe
Get the full app experience
Hey Scott, this really inspired me so I put together a demo of an idea I've had for along time. Would love to know your thoughts!
https://youtu.be/Kd9AbKawwhg