Open damnkrat opened 1 year ago
@Aokromes Can you provide a link to the duplicate? I couldn't related issue before creating new one
I don't think this was reported before. Although there's no contributor who can implement that so it's unlikely to appear...
it was, i remember it reported, but hard to find xd
I see there were questions about why this is not implemented (#16730, #17163) but no bug report about that. Although I'd say this should be a feature request.
Also both are marked as closed, we need to open one of the questions, or without that, do this feature
I don't see a reason to open those questions, they're answered
I'm highly doubt the feature will be implemented, with an issue or without. But you can try to open a new feature request, hope @Aokromes won't close anymore.
I assume Telegram is using the XDG portal for screencasting? If so then sending audio with application/system capture is currently unsupported (which is why it doesn't work in Discord or OBS either). There has been some movement on this recently, I'd recommend keeping watch on this issue (and for the Telegram devs working on Linux it might be a good idea to contribute their own two-cents there).
I assume Telegram is using the XDG portal for screencasting?
WebRTC.
There has been some movement on this recently, I'd recommend keeping watch on this issue (and for the Telegram devs working on Linux it might be a good idea to contribute their own two-cents there).
I'm not sure that would help. WebRTC has no "capture system audio" API and if this would become a part of the screencasting API then it would become unimplementable for tdesktop as currently the portal session is managed by WebRTC and tdesktop doesn't have access to its handles so it would be unable to speak to the portal directly and WebRTC doesn't have an API for this.
Right now it's implementable as application has full access to PipeWire/PulseAudio/ALSA. The only problem is application uses OpenAL for audio and OpenAL doesn't have API for this feature either so this feature has to be implemented with native code while tdesktop project has no one with PipeWire/PulseAudio/ALSA experience to implement any backend (not even speaking about every) for Linux. I guess OBS and Discord having the same problem: they don't have experience with Linux audio APIs and they always have much more prioritized tasks to spend time on rather than learning the zoo of Linux audio APIs.
I know there's https://github.com/maltejur/discord-screenaudio/blob/master/src/virtmic.cpp which could be used as a base but still someone has to spend time to port the implementation to tdesktop...
some additional info about screen sharing with audio on wayland + pipewire
hope we get this feature in telegram soon
@gmankab someone interested should implement it and make a PR
This feature is still not present as of v4.16.6 , would be nice to have it implemented somehow
No one seem to be really interested doing such work :(
For those who want some workaround for a time, check out helvum. Just connect audio sources manually in it
For those who want some workaround for a time, check out helvum. Just connect audio sources manually in it
thanks
Steps to reproduce
Expected behaviour
A button to share audio
Actual behaviour
No ability to share audio at all. No matter X11 or Wayland
Operating system
Linux
Version of Telegram Desktop
4.8.7
Installation source
Other (unofficial) source -> From Manjaro repo
Crash ID
No response
Logs
No response