So Canonical just announced Project Myna, and honestly it’s about time Ubuntu got native speech-to-text. The idea of pressing a hotkey, speaking, and having text land wherever your cursor is - that’s genuinely useful, not just for accessibility but for everyday workflow too.
That said, I want to bring up something the announcement kind of glossed over: the initial release targets GNOME on Wayland only, delivered as a Snap.
For those of us running Kubuntu, Xubuntu, or Ubuntu MATE - what does that actually mean in practice? Will Myna work at all outside GNOME, or is it just untested? Is the Snap delivery going to cause friction on flavors where Snap integration is less seamless?
Canonical did say they’re keeping the architecture open enough to support additional desktop environments later, which is good. But “later” is doing a lot of work in that sentence.
2 Likes
My guess is that Snap delivery is not the hard part here. The tricky part is probably the desktop integration: global shortcut, Wayland input, and reliably inserting text into the focused app. GNOME on Wayland is one target to make that predictable. For flavours I would mainly want a clear “works/does not work” matrix: KDE Wayland, KDE X11, Xfce X11, MATE X11. Even if only GNOME is supported first, that would make it much easier for flavour users to test without guessing whether a failure is expected.
1 Like
My guess is, that it should be pretty agnostic to the desktop environment, as long as one can assign a keyboard shortcut for the speech-to-text action. I don’t see how Wayland, or not, should figure at all; it’s just audio capture that needs to work, and I believe that’s handled by snap connect myna:audio-record, or rather an automatic connection at install time.