It’s that time of the cycle we’re starting to look at what’s next. And to prioritize, we need your input.
What is it that you’re missing when building on Mir, or using Mir-based sessions? What functionality, what use cases are not covered?
Please treat this as a $holiday wish list, we might not know if you don’t tell us - even if it’s not something we can do immediately, it’s going to paint a road for us. And for you, too, as we’ll want to do better in terms of transparency of our current and future roadmap.
since it is a wishlist, i would be curious to see if the river window management protocol could be adopted. that way we could have a mir wayland compositor and have the ability to swap window manager, could help share compositors if the only difference the environment want is how the windows managed as well
Another protocol i’d like to see is the global menu protocol
and oh yeah I saw a lot of hype around the color management protocol a lil while ago, that might a good one to make mir more known as an option for wayland compositors to build on xd
there’s also the screen capture protocols, so don’t perse need portals for it— and then there’s xwayland shell…….but i may be getting greedy
I don’t see river window management protocol being adopted as part of Mir: Mir already has a way to customize window management (see my FOSDEM2026 talk).
For Miriway, I am interested in supporting alternative window management policies and would be interested in how closely RWMP is tied to the river compositor. Isaac didn’t seem to think the current version is ready for cross-compositor use. (PRs welcome!)
And Miracle is already working on a different approach to custom window management.
I came here to suggest this. I was at Isaac’s talk at FOSDEM and wrote about it on El Reg. I was very impressed by his talk.
The other most interesting project I’ve seen in the Wayland space in the last year or 2 is Wayback and I wrote about that, too – an intro followed by a status update. I suspect that the design of Wayback would be significantly more difficult to incorporate, though…
The main thing I want but can’t find is Mir currently is custom decorations. There’s a few issues referring to this but I’m not sure what the status is. Mostly I just want to add a border to windows, which Miracle does with a custom renderer, but it would be nice if there was an easier way to do this.
I also want middle click to paste, but I don’t know if there’s a way to implement this currently.
The status is that several things are in flight but need to complete for this to be the way we’d like it. Mir’s current SSD code could be exposed for tweaking, but is not flexible enough.
oh darn so there are 3 approaches to custom window management would be nice if we could standardize it one day
But yeah i don’t really expect this by 26.10, like i said “since it is a wishlist” tis more wish than anything else
if had to choose which protocols i really hope we can get sooner than later it would be the screen capture protocols and the global menu protocol
Thanks, I didn’t realise it just required enabling the extension. I guest the next request would be the data control protocol to have a unified clipboard and selection.
I’m sure you’re also already looking at it but another thing is output management, maybe wlr output management, although everyone seems to create their own protocol.
And I’m not sure if this is necessary but maybe events for when the mouse enters a client or when it enters an output. I’ve implemented focus follows mouse in my compositor but I have some performance issue, although I haven’t been able to look at it yet so it could be something I can fix.
That sounds like a compositor thing indeed, and should already be possible through the MirAL APIs? Maybe it would also fit a output management protocol, if one crystallizes.
That’s also there since 2.25 - but again, privileged, so not enabled by default (I just proposed an update to Wayland Explorer for 2.26).
Oh thanks, I can probably just enable all supported protocols. I think if wl-clipboard was updated to use primary-selection-unstable-v1 (they’re still using the old name gtk_primary_selection) I could use wl-paste --primary --watch wl-copy to sync the selection and clipboard.
That’s part of the core protocols: wl_pointer::enter and ::leave.
Thanks, I’ll have a look at that. Anyway, my performance issue wasn’t due to focus follows mouse, it’s something to do with NVIDIA hybrid graphics using an external monitor.
Hey I find Mir fascinating project. It is great that we have walk-throughs that help install applications and Kiosks etc. But I would like something like EGMDE to produce a mature desktop where developers can easily add their own customisations. Eventually I would like this to be able to support Gnome and KDE replacing their own backend….
I was at Isaac’s talk at FOSDEM and