Mir 2.5 Released With Wayland Extension Updates For Better Handling On-Screen Keyboards

[phoronix]

Phoronix: Mir 2.5 Released With Wayland Extension Updates For Better Handling On-Screen Keyboards

Canonical continues advancing their Wayland-based Mir stack for embedded and IoT use-cases. Out today is Mir 2.5 with the latest features as they work to provide better support for on-screen keyboards...

https://www.phoronix.com/scan.php?page=news_item&px=Mir-2.5-Released

[evasb]

Sad that Canonical basically became a smaller Red Hat.

[Britoid]

Sad that Canonical basically became a smaller Red Hat.

Nah that's SUSE.

[Vistaus]

/me *waits for the usual posts of people who still don't understand that Mir is a Wayland extension instead of a separate display server nowadays*

[SWY1985]

Wait, didn't this display server die in favor of Wayland? /s

[AlanGriffiths]

/me *waits for the usual posts of people who still don't understand that Mir is a Wayland extension instead of a separate display server nowadays*

Corrections:
1. Mir is a Wayland implementation (and implements various Wayland extensions).
2. Mir is still a display server.

[tildearrow]

Wait, didn't this display server die in favor of Wayland? /s

Wayland is not a display server.

What died is the Mir protocol, and Mir is now a Wayland implementation.

[evasb]

Tbh, I think Mir protocol at the time was a little better than Wayland. DEs could just use the Mir API and share all the infrastructure of the Mir display server. The fragmentation situation would be better than today. But instead of doing this case, Canonical was clearly going to the Open Core model, asking for Contributor Agreements instead of using a better license. Canonical bleed money until Mark shot down the desktop side of Canonical, closed the #1 bug and invested everything on IoT and Cloud business.

I sometimes feel bad for Mir developers (and Linux desktop) because I still think Mir was a better solution than Wayland and its multiple implementations. It took almost a decade to Wayland become acceptable.

[dragon321]

Tbh, I think Mir protocol at the time was a little better than Wayland. DEs could just use the Mir API and share all the infrastructure of the Mir display server. The fragmentation situation would be better than today. But instead of doing this case, Canonical was clearly going to the Open Core model, asking for Contributor Agreements instead of using a better license. Canonical bleed money until Mark shot down the desktop side of Canonical, closed the #1 bug and invested everything on IoT and Cloud business.

I sometimes feel bad for Mir developers (and Linux desktop) because I still think Mir was a better solution than Wayland and its multiple implementations. It took almost a decade to Wayland become acceptable.

DE can still just use Mir instead of making their own compositor from scratch. The fact Mir abandoned its own protocol and adopted Wayland protocol doesn't change this. As far I know MATE was interested in adopting Mir as their Wayland compositor. Even if you don't want to touch Mir for some reason then you don't have to write Wayland compositor from scratch - you can use existing compositor or library like wlroots.

Mir wasn't much better than Wayland in that regard. Mir and Wayland protocol and infrastructure were very similar to each other so Wayland problems were also Mir problems in many cases. Canonical was also pretty much alone in supporting Mir. To solve Mir problems they would need to push their own and incompatible solutions or adopt Wayland solutions. First case would increase fragmentation, second would make unnecessary code duplication.

[user1]

Sad that Canonical basically became a smaller Red Hat.

Canonical is anything but Red Hat.