Firefox 80 To Support VA-API Acceleration On X11

[nanonyme]

When he says "display manager config file": Manjaro, and therefore probably arch, hard disables wayland in the gdm3 config file, so he is correct. On a default manjaro gnome install, the cog therefore does not show a wayland option.

Having spent a few days using Wayland gnome on Fedora (on a pure intel laptop), I think the manjaro default is poor. For me, Wayland is very close to being awesome. Zoom works (except for allowing a remote user to take control of the host), screen recording via the gnome extension EasyScreencast is great (choose a particular monitor or rectangular area), ksnip screen shots work (if installed via .rpm, the snap didn't work), and Firefox in native wayland is a revelation. Even the 'blurring' of non-Wayland apps is not too bad. If JetBrains made its products Wayland native, I would switch my main desktop (with AMD graphics). Note tlat screen shots, screen sharing ... these are things that Wayland has a reputation for breaking, but it's mostly not true any longer.

As for Firefox: it is very awesome again, the containers feature is so much better than Chromium profiles. A few days ago after I solved how to make bookmarks open in specific containers (meaning I can book mark gmail and personal gmail/calendar/gdrive for instant access to work and personal containers, thank you https://addons.mozilla.org/en-US/fir...l-in-container) , it's my default browser again, after five or six years of Chrome/Chromium.

Now if only Zoom used the portals properly intead of banging pipewire directly, it would also work from inside Flatpak sandbox.

[deusexmachina]

Now if only Zoom used the portals properly intead of banging pipewire directly, it would also work from inside Flatpak sandbox.

Would love to get some clarity on browser sandboxing... So many opinions and so little reliable data. I hear that firejail "increases the attack surface area," for example. There are also claims of chromium's sandboxing being better than Firefox's - and claims of them being equal...

[mpw1412]

I'm happy to see there's some progress. But there's little to no impact.

- Playing a Youtube-Video in Firefox: ~110-130% CPU
- Playing the exact same Youtube-Video via youtube-dl+mpv: 6-8% CPU

I don't know what they're doing there, but the acceleration is far from what Linux drivers provide.

[Azrael5]

It has not implemented as default both in 80 and 81 releases. In my opinion it's a false news.