Ivan MolodetskikhThis morning was my niri (-adjacent) talk! 
Went really well, got many interesting questions!
They promised us video recordings in a few days, so I'll post a link when I get it. Though of course the talk is in Russian.
unusual sights
Turns out, there's a lot of details to get right when implementing a floating window space. For example, dialog windows should always show above their parent window. Otherwise, it's easy to lose them under the (usually much bigger) parent.
The WIP floating branch in niri now handles this properly, even for xdg-desktop-portal dialogs (like file chooser) as long as the app correctly parents them via xdg-foreign.
Another piece of the floating puzzle: keeping windows on screen. When you change your monitor scale or resolution, you don't want your floating windows to suddenly go unreachable behind the monitor's new borders.
Here I'm resizing a nested niri with three windows, simulating resolution changes. No matter what I do, they always remain partially visible and reachable. Even for more unusual cases like trying to resize a window into out of bounds.
In the tiling layout, niri is constantly asking windows to assume their expected size. In contrast, floating windows should be able to freely change size as they see fit.
The logic turns out to be quite tricky. On the one hand we want a window to keep its latest size, but on the other we still want to be able to resize the window, which means asking it for a different size. The window can take a second to respond, or respond with a yet another size, and nothing must break.
While trying to make this work, I realized that this is the time when I *really really* want to be able to test this stuff. So I got on a sidetrack adventure to write testing infra for running real Wayland clients inside unit tests.
I've got it working! In these tests, I'm creating a new niri instance along with test clients, all on the same test-local event loop. No global state, no threads needed.
What's really cool is that this lets me test the weirdest client-server event timings.
This morning I worked on remembering the size for floating windows when they go to the tiling layout and back.
The whole sizing code must be at the top by logic complexity in niri. I have to juggle, all at once:
- new size I haven't sent to the window yet,
- size changes I sent, but window hasn't acked yet (0, 1, or more in-flight),
- size change window acked but hasn't committed for yet,
- size change window acked and responded to with a commit (maybe with a different size entirely).
The diff is 85 lines of change and 243 lines of new tests, and I already found a few weirder edge cases that I've missed. No way I could do this well without that client-server testing setup that I posted about yesterday.
Btw I pushed the testing setup if you're curious, along with the entirety of 1215 snapshot files for a powerset of new window workspace/output target settings: https://github.com/YaLTeR/niri/commit/771ea1e81557ffe7af9cbdbec161601575b64d81
The WIP floating branch caused them to update in several commits already.
The big 1215 snapshot test powerset (actually it already grew to 1695) continues to prove its worth. Just finished a big +495 -508 cleanup of the window opening code, and verified that not a single of those 1215 window opening configurations changed its outcome. I will be sleeping well tonight
After three weeks of hard work, I am undrafting the floating window PR in niri. Please give it thorough testing and report any bugs or issues!
I spent today figuring out the remaining layer-shell keyboard focus problems, and I've got it all working! Pop-ups now render above windows, and bottom/background layers can receive on-demand focus.
Effectively, this makes the desktop icons components from @LXQt or @xfce just work on niri!
Alright, I think I got all of the important things in for the next niri release. Today I updated Smithay for the DRM compositor changes, and added a workaround for a panic when you have two monitors with exactly matching make/model/serial.
I'll give it a week of testing (if you run niri-git, please report any problems) and if all goes well, tag next Saturday.
There are a few PRs I'll try to review in time, but they're fairly self contained.
After a full day of writing release notes (god how'd it take so long 😫), niri v25.01 is out with Floating Windows and Working Layer-Shell Desktop Icons and Layer-Shell Screencast Blocking Out and so many more improvements! Yes, you read that right, we finally escaped zerover! I feel that niri is now ready to graduate from v0.1
Read here and download when your distribution package updates: https://github.com/YaLTeR/niri/releases/tag/v25.01
something odd about these windows
Looking for testing and feedback for server-side shadows: https://github.com/YaLTeR/niri/pull/990
ppl from our niri matrix playing around with the new compositor-side shadows! These screenshots are from @r4hulrosh4n and calops (no fedi)
Added shadow support for layer-shell surfaces!
Though unfortunately layer-shell has no way to signal the visual geometry, so this only looks right if the layer surface doesn't have its own margins.
https://github.com/YaLTeR/niri/wiki/Configuration:-Layer-Rules#shadow
I'm adding tabs to niri. Instead of some separate mode, they're just changing how a column is displayed. This means all your hotkeys and everything works exactly the same with tabs. Which was a wonderful UX idea by @elkowar!
I've got a draft PR going with some design and UX questions, please feel free to try it and give feedback: https://github.com/YaLTeR/niri/pull/1085
Why would you even want tabs in niri? Occasionally it's quite useful. For example, when streaming programming, I increase the font size, so I can no longer fit editor + terminal on the screen at once. Scrolling back and forth gets annoying, and tabs feel just right for this.
A ton of changes in the tabs PR over the past few days. Mainly various options (tabbed display by default, tab indicator position, etc.).
Just finished with a big one: you can now place the tab indicator within the column rather than "outside". This is needed for thicker tab bars, since otherwise they overlap adjacent windows.
I merged tabbed columns into niri! Now you can play around with them using your nearest niri-git package. With working animations and all
Noticed tabs can sometimes be useful for comparing windows without taking screenshots. Here for example I'm running the Adw demo from F41 vs. nightly Flathub, showing the slight color difference and apparently a 1 px layout shift.
Another cool suggestion by @elkowar: you can now match windows recorded by an active window screencast. For example, to highlight the casted window with a different color.
You can now (finally) customize the important hotkeys list: https://github.com/YaLTeR/niri/wiki/Configuration:-Key-Bindings#custom-hotkey-overlay-titles
Today in *very long* overdue features (looks like I opened the issue even before v0.1.0): moving the mouse against a monitor edge now scrolls the view during drag-and-drop.
(before this change, you had to use the keyboard for this, and yes it was very awkward)
Another neat new thing: a bind to expand column to available width. Basically, "expand to fill empty space".
But, a bit smarter: with scrollable tiling we can have windows partially off-screen. This bind ignores such windows, making it easy to position things to exactly fill the screen, even in the middle of a scrolling layout.
Today I'm releasing niri v25.02 with tabs, shadows, DnD view scrolling, and a ton of other improvements! Read the release notes at https://github.com/YaLTeR/niri/releases/tag/v25.02 and @ your distro to update the package.
[[honestly wtf how did so much stuff happen, it's been barely over a month]]
This is not on main yet, but I've heavily reworked how offscreening works in niri.
Offscreening is when you render a group of elements to a separate texture, then draw that texture to the final render. In niri it's used for window opening and resize animations, and now also for full-tile opacity changes.
Before, it was very simple: every frame create a new texture and render it from scratch. The new code caches textures and does full damage tracking both "inside" and "outside" the offscreen.
In practice this means much less texture re-creations and much less redrawing.
The main reason to do this was to unlock using offscreens for more persistent visuals. E.g. it's fine to do some extra rendering for a 150 ms animation, but it's not fine to do it for something that can last seconds or minutes.
In particular, I can finally make windows semitransparent while dragging them in the tiling layout! Which needs an offscreen since it's a stack of border + window + subsurfaces.
Still testing this one: a window rule to set the xdg-toplevel Tiled state.
By default niri matches it to prefer-no-csd, because the Tiled state currently happens to be the best way to ask always-CSD windows to square their corners.
With this tiled-state window rule you can override it, for example to get title bars (for easy mouse-only gestures) together with square corners (if you're going for a square style with borders, like mine here). You can also e.g. limit it to tiled windows.
And another thing for today: finally made it so niri waits a bit for the lock screen to paint before locking the session. This fixes the "red flash" issue.
Please test this since it's security sensitive. I already caught one mistake before pushing where killing a lock screen, then spawning a new one, would briefly flash the session contents.
(Note that some lockers fade in from transparency, so you'll still briefly see red, that's just the niri locked session background.)
Since adding tabbed columns, just about everyone is asking for a way to open a new window directly in the focused tabbed column. This is rather complicated, but yesterday I came up with a 1 line change that should cover some workflows (covers how I use tabs at least).
I have logic that restores the view position when you open and close a window right away. Now I enabled it for consume-left, which means that you can open a new window, consume left (into a tab), and the view will go back.
Today I implemented a new niri screencasting feature: the dynamic cast target.
It shows up as a special "window" in the portal dialog. Once you select it, you can change what it casts with niri binds! Switch between windows, monitors, and in the future maybe workspaces and other stuff. Works with OBS, browsers, anything else.
I also wrote a wiki page with all our screencasting-related features: https://github.com/YaLTeR/niri/wiki/Screencasting
Thanks @elkowar for coming up with the dynamic cast idea!
Had to show slides a few times for uni lately, which motivated me to add this next feature.
Windowed fullscreen, also known in other WMs as fake or detached fullscreen, makes the window think that it went fullscreen, while in reality keeping it as a normal window. Useful in combination with apps like Google Slides to hide the browser UI without taking up the whole screen, especially on ultrawide monitors.
Wiki section: https://github.com/YaLTeR/niri/wiki/Screencasting#windowed-fakedetached-fullscreen
This one seems simple, but actually turned out quite tricky to implement.
I wrote niri code under the assumption that the fullscreen state does not apply immediately: we send a configure to the window, then only at a later point it commits in response. Windowed fullscreen breaks this because in some cases it can apply instantly (going from real to windowed fullscreen needs no state change from the window).
After trying a few approaches, I ended up doing it "properly" by associating the (entirely compositor-side) windowed fullscreen state to window configures and commits. This lets niri correctly track if a given fullscreen window commit was real fullscreen (so e.g. we need to draw a black backdrop), or if it was windowed fullscreen.
This even works across a chain of toggling windowed fullscreen, then the window slowly acking and committing them, and I have a test to verify that.
New niri users frequently come from other tiling WMs with static workspace systems. For them, dynamic workspaces is an unusual system, and it can be unclear how to use dynamic workspaces effectively. (Especially until we get an Overview type thing.)
I figured I'd write a wiki page describing niri's workspaces in more detail: https://github.com/YaLTeR/niri/wiki/Workspaces I also included an example of how I personally use workspaces on niri.
Hopefully this makes it easier to understand what's going on!
la ninpre@YaLTeR i came from hyprland and sway and then realized that i used workspaces there primarily to avoid having too many windows on one screen. with niri this is almost never an issue, since there's always some space on the right and on the left. niri's dynamic workspaces are more like actual workspaces, separating different tasks.